Innovations in Smart Cities Applications Volume 5: The Proceedings of the 6th International Conference on Smart City Applications (Lecture Notes in Networks and Systems, 393) 3030941906, 9783030941901

Citation preview

Lecture Notes in Networks and Systems 393

Mohamed Ben Ahmed · Anouar Abdelhakim Boudhir · İsmail Rakıp Karaș · Vipul Jain · Sehl Mellouli   Editors

Innovations in Smart Cities Applications Volume 5 The Proceedings of the 6th International Conference on Smart City Applications

Lecture Notes in Networks and Systems Volume 393

Series Editor Janusz Kacprzyk, Systems Research Institute, Polish Academy of Sciences, Warsaw, Poland Advisory Editors Fernando Gomide, Department of Computer Engineering and Automation—DCA, School of Electrical and Computer Engineering—FEEC, University of Campinas— UNICAMP, São Paulo, Brazil Okyay Kaynak, Department of Electrical and Electronic Engineering, Bogazici University, Istanbul, Turkey Derong Liu, Department of Electrical and Computer Engineering, University of Illinois at Chicago, Chicago, USA Institute of Automation, Chinese Academy of Sciences, Beijing, China Witold Pedrycz, Department of Electrical and Computer Engineering, University of Alberta, Alberta, Canada Systems Research Institute, Polish Academy of Sciences, Warsaw, Poland Marios M. Polycarpou, Department of Electrical and Computer Engineering, KIOS Research Center for Intelligent Systems and Networks, University of Cyprus, Nicosia, Cyprus Imre J. Rudas, Óbuda University, Budapest, Hungary Jun Wang, Department of Computer Science, City University of Hong Kong, Kowloon, Hong Kong

The series “Lecture Notes in Networks and Systems” publishes the latest developments in Networks and Systems—quickly, informally and with high quality. Original research reported in proceedings and post-proceedings represents the core of LNNS. Volumes published in LNNS embrace all aspects and subfields of, as well as new challenges in, Networks and Systems. The series contains proceedings and edited volumes in systems and networks, spanning the areas of Cyber-Physical Systems, Autonomous Systems, Sensor Networks, Control Systems, Energy Systems, Automotive Systems, Biological Systems, Vehicular Networking and Connected Vehicles, Aerospace Systems, Automation, Manufacturing, Smart Grids, Nonlinear Systems, Power Systems, Robotics, Social Systems, Economic Systems and other. Of particular value to both the contributors and the readership are the short publication timeframe and the world-wide distribution and exposure which enable both a wide and rapid dissemination of research output. The series covers the theory, applications, and perspectives on the state of the art and future developments relevant to systems and networks, decision making, control, complex processes and related areas, as embedded in the fields of interdisciplinary and applied sciences, engineering, computer science, physics, economics, social, and life sciences, as well as the paradigms and methodologies behind them. Indexed by SCOPUS, INSPEC, WTI Frankfurt eG, zbMATH, SCImago. All books published in the series are submitted for consideration in Web of Science. For proposals from Asia please contact Aninda Bose ([email protected]).

More information about this series at https://link.springer.com/bookseries/15179

Mohamed Ben Ahmed Anouar Abdelhakim Boudhir İsmail Rakıp Karaș Vipul Jain Sehl Mellouli •

•

•

•

Editors

Innovations in Smart Cities Applications Volume 5 The Proceedings of the 6th International Conference on Smart City Applications

123

Editors Mohamed Ben Ahmed Faculty of Sciences and Techniques, Computer Engineering Department Abdelmalek Essaadi University Tangier, Morocco

Anouar Abdelhakim Boudhir Faculty of Sciences and Techniques, Computer Engineering Department Abdelmalek Essaadi University Tangier, Morocco

İsmail Rakıp Karaș Computer Engineering Department Karabük University Karabük, Turkey

Vipul Jain Operations and Supply Chain Management Wellington School of Business and Govern Wellington, Wellington, New Zealand

Sehl Mellouli Département Des Systèmes d’Information Organisationnels Faculty of Business Administration. Laval University Québec, QC, Canada

ISSN 2367-3370 ISSN 2367-3389 (electronic) Lecture Notes in Networks and Systems ISBN 978-3-030-94190-1 ISBN 978-3-030-94191-8 (eBook) https://doi.org/10.1007/978-3-030-94191-8 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2022, corrected publication 2022 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Preface

On the research area, the potential of the smart cities axis is much higher, as many cities are still at the start of their digital transformation. Because of this, the research field is known as a main driver of innovation and creativity, in order to create multiple applications in the future smart cities, ranging from safety to health, including mobility, the environment, agriculture, water, buildings, and infrastructure. In the same way, urban governance calls for research and development of more concerted and transversal approaches, closely involving citizens and businesses in the development of smart cities. In addition to the research development, the aims of this book are also to participate in developing new uses, whether internally or toward the citizen who will be in the next and future population of those smart cities. This must attract those uses to know how they will and they can be the smart people of the next generations. This edited book edition titled “Innovations in Smart Cities and Applications” continues providing the last researches and applications that involves the new technologies for smart cities. This edition is the fruit of the accepted and presented works in the fifth International Virtual Conference on Smart City Applications (SCA 2021) held on October 27–29, 2021, in Safranbolu, Türkiye. SCA 2021 conference regroups original research, achieved works, and proposed architectures on the main topics of the conference. The scope of SCA 2021 covers a variety of topics with an intersection with smart cities, geo-smart information systems, education, health care, economy and digital business, building and home automation, environment and agriculture, and information technologies and computer science. Many thanks are addressed to participants and researchers and to all keynote speakers for their valued and rich scientific talk at the conference session. We are deeply grateful to all organizing committee members from both sides, Moroccan and Turkish teams, to all program committee members and reviewers, to all chairs of sessions for their efforts and the time spent in order to evaluate the contributions and to the success of this event.

v

vi

Preface

We also would like to acknowledge and thank the Springer Nature Switzerland AG staff for their support, guidance, and for the edition of this book. To close, we hope to express our sincere thanks to Pr. Janusz kacprzyk, Dr. Thomas Ditzinger, and Ms. Viradasarani Natarajan for their kind support and help to promote the success of this book. Mohamed Ben Ahmed Anouar Abdelhakim Boudhir İsmail Rakıp Karaș Sehl Mellouli Vipul Jain SCA21 Chairs

Committees

Conference Chair İsmail Rakıp Karaș

Karabuk University, Türkiye

Conference General Chairs Mohamed Ben Ahmed Anouar Boudhir Abdelhakim Bernadetta Kwintiana Ane

FST, Tangier UAE University, Morocco FST, Tangier UAE University, Morocco University of Stuttgart, Germany

Conference Technical Program Committee Chair Vipul Jain

Victoria University of Wellington, New Zealand

Publication Chair Sehl Mellouli

Laval University, Laval, Canada

Special Issues Chair Parthasarathy Subashini

Avinashilingam University, India

Local Organizing Committee Idris Kahraman Emrullah Demiral Mustafa Aksin Kadriye Oz Hacer Kübra Köse Berna Gunes Umit Atila

Karabuk University, Türkiye Karabuk University, Türkiye Karabuk University, Türkiye Karabuk University, Türkiye Sinop University, Türkiye Karabuk University, Türkiye Karabuk University, Türkiye

vii

viii

Kasim Ozacar Yasin Ortakci Muhammed Kamil Turan Sohaib Abujayyab Emre Yücer

Committees

Karabuk Karabuk Karabuk Karabuk Karabuk

University, University, University, University, University,

Türkiye Türkiye Türkiye Türkiye Türkiye

Technical Program Committee Ismail Rakip Karas Abdel-Badeeh M. Salem Abdullah Elen Abdullah Emin Akay Abdurrahman Eymen Accorsi, Riccardo Adib Habbal Adnan Alajeeli Aftab Ahmed Khan Ahmad S. Almogren Ahmed Kadhim Hussein Alabdulkarim Lamya Alghamdi Jarallah Ali Jamali Alias Abdul Rahman Aliihsan Sekertekin Anabtawi Mahasen Anton Yudhana Arif Çağdaş Aydinoglu Arioua Mounir Assaghir Zainab Aydın Üstün Aziz Mahboub Bahadır Ergun Barış Kazar Bataev Vladimir Behnam Alizadehashrafi Behnam Atazadeh Ben Yahya Sadok Bessai-Mechmach Fatma Zohra Beyza Yaman Biswajeet Pradhan Berk Anbaroğlu Boutejdar Ahmed Burhan Selcuk

Karabuk University, Türkiye Ain Shams University, Egypt Bandirma Onyedi Eylül University, Türkiye Bursa Technical University, Türkiye Erciyes University, Türkiye Bologna University, Italy Karabuk University, Türkiye Karabuk University, Türkiye Karakoram International University, Pakistan King Saud University, Saudi Arabia Babylon University, Iraq King Saud University, Saudi Arabia Prince Sultan University, Saudi Arabia Universiti Teknologi Malaysia Universiti Teknologi Malaysia Cukurova University Al-Quds University, Palestine Universitas Ahmad Dahlan, Indonesia Gebze Technical University, Türkiye UAE, Morocco Lebanese University, Lebanon Kocaeli University, Türkiye FSTT UAE, Morocco Gebze Technical University, Türkiye Oracle, USA Zaz Ventures, Switzerland Tabriz Islamic Art University, Iran University of Melbourne, Australia Faculty of Sciences of Tunis, Tunisia CERIST, Algeria Dublin City University, Ireland University of Technology Sydney, Australia Hacettepe University, Türkiye German Research Foundation, Bonn, Germany Karabuk University, Türkiye

Committees

Bulent Bayram Caner Ozcan Caner Güney Chadli Lala Saadia Cumhur Şahin Damir Žarko Dominique Groux Dousset Bernard Edward Duncan Eehab Hamzi Hijazi El Kafhali Said Eftal Şehirli El Mhouti Abderrahim El Haddadi Anass El Hebeary Mohamed Rashad El Ouarghi Hossain Elif Sertel Emre Yücer Emrullah Sonuç En-Naimi El Mokhtar Enrique Arias Fatmagül Kılıç Gül Ferhat Atasoy Filip Biljecki Füsun Balık Şanlı Francesc Anton Castro Ghulam Ali Mallah Habibullah Abbasi Haddadi Kamel Iemn Hakan Kutucu Hande Demirel Hazim Tawfik Huseyin Bayraktar Hüseyin Pehlivan Huseyin Topan Huseyin Zahit Selvi İbrahim Baz İlhami Muharrem Orak Ilker Türker Iman Elawady Indubhushan Patnaikuni İsa Avcı Ismail Büyüksalih Ivin Amri Musliman

ix

Yildiz Technical University, Türkiye Karabuk University, Türkiye Istanbul Technical University, Türkiye University Sultan Moulay Slimane, Morocco Gebze Technical University, Türkiye Zagreb University, Croatia UPJV, France UPS, Toulouse, France The University of Mines & Technology, Ghana An-Najah University, Palestine Hassan 1st University, Settat, Morocco Karabuk University, Türkiye FST, Al-Hoceima, Morocco UAE University, Morocco Cairo University, Egypt ENSAH UAE University, Morocco Istanbul Technical University, Türkiye Karabuk University, Türkiye Karabuk University, Türkiye UAE, Morocco Castilla-La Mancha University, Spain Yıldız Technical University, Türkiye Karabuk University, Türkiye National University of Singapore Yıldız Technical University, Türkiye Technical University of Denmark Shah Abdullatif University, Pakistan University of Sindh, Pakistan Lille University, France Karabuk University, Türkiye İstanbul Technical University, Türkiye Cairo University, Egypt General Directorate of GIS, Türkiye Gebze Technical University, Türkiye Bulent Ecevit University, Türkiye Konya Necmettin Erbakan University İstanbul Ticaret University, Turkiye Karabuk University, Türkiye Karabuk University, Türkiye Ecole Nationale Polytechnique d’Oran, Algeria RMIT - Royal Melbourne Institute of Technology, Australia Karabuk University Bimtaş A.Ş., Türkiye Universiti Teknologi Malaysia

x

J. Amudhavel Jaime Lioret Mauri Jus Kocijan Kadir Ulutaş Kasım Ozacar Khoudeir Majdi Labib Arafeh Laila Moussaid Lalam Mustapha Loncaric Sven Lotfi Elaachak Mademlis Christos Mehmet Akbaba Mete Celik Miranda Serge Mohamed El Ghami Mohammad Sharifikia Mousannif Hajar Mufit Cetin Muhamad Uznir Ujang Mike Horhammer Muhammad Imzan Hassan Muhammed Kamil Turan Murat Yakar Murat Lüy Mustafa Akgul My Lahcen Hasnaoui Mykola Kozlenko Nesrin Aydin Atasoy Nusret Demir Oğuz Fındık Oğuzhan Menemencioğlu Omar Dakkak Omer Muhammet Soysal Ouederni Meriem Rachmad Andri Atmoko R. S. Ajin Rani El Meouche Raif Bayır Rafet Durgut Sagahyroon Assim Saied Pirasteh

Committees

VIT Bhopal University, Madhya Pradesh, India Polytechnic University of Valencia, Spain Nova Gorica University, Slovenia Karabuk University Karabuk University IUT, Poitiers University, France Al-Quds University, Palestine ENSEM, Casablanca, Morocco Mouloud Mammeri University of Tizi Ouzou, Algeria Zagreb University, Croatia FSTT, UAE, Morocco Aristotle University of Thessaloniki, Greece Karabuk University, Türkiye Erciyes University, Türkiye Nice University, France University of Bergen, Norway Tarbiat Modares University, Iran Cadi Ayyad University, Morocco Yalova University, Türkiye Universiti Teknologi Malaysia Oracle, USA Universiti Teknologi Malaysia Karabuk University, Türkiye Mersin University, Türkiye Kırıkkale University, Türkiye Istanbul University, Türkiye Moulay Ismail University, Morocco Vasyl Stefanyk Precarpathian National University, Ukraine Karabuk University, Türkiye Akdeniz University, Türkiye Karabuk University, Türkiye Karabuk University, Türkiye Karabuk University, Türkiye Southeastern Louisiana University, USA INP - ENSEEIHT Toulouse, France Universitas Brawijaya, Indonesia DEOC DDMA, Kerala, India Ecole Spéciale des Travaux Publics, France Karabuk University, Türkiye Karabuk University, Türkiye American University of Sharjah, United Arab Emirates University of Waterloo, Canada

Committees

Savas Durduran Sedat Bakici Senthil Kumar Serdar Bayburt Seyit Ali Kayış Sibel Senan Siddique Ullah Baig Sinasi Kaya Slimani Yahya Sohaib Abujayyab Sonja Grgić Sri Winiarti Suhaibah Azri Sunardi Sule Erten Ela Tarik Adnan Almohamad Tebibel Bouabana Thouraya Tolga Ensari Umit Atila Umit Isikdag Umran Koylu Xiaoguang Yue Yasin Ortakcı Yasyn Elyusufi Yüksel Çelik Youness Dehbi Yusuf Arayıcı Yusuf Yargı Baydilli Zafer Albayrak Zennure Uçar

xi

Konya Necmettin Erbakan University, Türkiye Turkish Cadastre Office, Türkiye Hindustan College of Arts and Science, India Bimtaş A.Ş., Türkiye Karabuk University, Türkiye Istanbul University, Türkiye COMSATS Institute of Information Technology, Pakistan İstanbul Technical University, Türkiye Manouba University, Tunisia Karabuk University, Türkiye Zagreb University, Croatia Universitas Ahmad Dahlan, Indonesia Universiti Teknologi Malaysia Universitas Ahmad Dahlan, Indonesia Ege University, Türkiye Karabuk University, Türkiye ESI, Alger, Algeria Istanbul University, Türkiye Karabuk University, Türkiye Mimar Sinan Fine Arts University, Türkiye Erciyes University, Türkiye International Engineering and Technology Institute, Hong Kong Karabuk University, Türkiye FSTT, UAE, Morocco Karabuk University, Türkiye University of Bonn, Germany Northumbria University, UK Karabuk University, Türkiye Karabuk University, Türkiye Düzce University, Türkiye

Keynotes

Zero-Touch Management and Orchestration of Network Slices in 5G and Beyond Networks Adlen Ksentini

Abstract. 6G systems are expected to serve a massive number of extremely heterogeneous network slices that cross multiple technological domains (i.e., RAN, edge, cloud, and core), posing significant challenges to classical centralized management and orchestration approaches in terms of scalability and sustainability. Within this context, a distributed and intelligent management and orchestration system is mandatory. This keynote presents the challenges related to the management and orchestration of network slices in 5G and beyond mobile networks. Based on these requirements, a hierarchical, distributed, and AI-driven management framework is introduced, featuring a zero-touch service management concept.

Biography: Prof. Adlen Ksentini is IEEE Communications Society Distinguished Lecturer. He received his Ph.D. degree in computer science from the University of Cergy-Pontoise in 2005. From 2006 to 2016, he worked at the University of Rennes 1 as Assistant Professor. Since March 2016, he has been Professor in the communication systems department of EURECOM, Sophia-Antipolis, France, where he leads the Network Softwarization group. He is involved in several European Union projects related to network slicing and 5G, such as the 5G!Drones and MonB5G projects

xv

Smart Cities Strategies: Critical Success Factors Domingos Santos

Biography: Domingos Santos holds a degree in Environmental Engineering (New University of Lisbon), a master’s degree in regional and urban planning (Technical University of Lisbon) and a Ph.D. in environmental applied sciences (University of Aveiro). He is Professor of the Polytechnic Institute of Castelo Branco (IPCB) where he has taught curricular units in the field of social development planning, development programs and projects, innovation and entrepreneurship, as well as sustainable development. He has developed teaching activities in the framework of cooperation and mobility programs, in University of Valladolid—Faculty of Economic and Business Sciences and University of Extremadura, both in Spain; Lithuania Business University of Applied Sciences; and, in Brazil, at Dom Bosco Catholic University and University of Santa Cruz do Sul.

xvii

Design for Energy: Prosumer Buildings Yusuf Arayici Northumbria University, UK

Biography: Yusuf Arayici is Professor in construction project management at Northumbria University. With an international outlook, he previously fulfilled an academic management role as Dean. He also successfully completed his research fellowship TUBITAK (Research Council of Türkiye) in Digital Construction. Since 2000, his research projects have ranged from building information modeling to sustainability. He has led substantial research groups over a prolonged period of time through continuous cycles of research with funded research projects, has graduated many Ph.D. and MSc students, and has published more than 100 papers and five books. Currently, he is researching on AI-supported heritage BIM.

xix

Smart Buildings to Smart Cities—The Role of BIM and GIS Integration Mohsen Kalantari University of Melbourne

Biography: Mohsen Kalantari is Associate Professor in geomatics and Associate Director at the Center for SDI and Land Administration in the Department of Infrastructure Engineering at the University of Melbourne, Australia. He teaches BIM, land administration systems, and spatial analysis. He is also Co-Founder of Faramoon, a start-up focusing on the automatic generation of building models from point clouds. His research covers the geospatial information value chain, including sourcing and capturing fit-forpurpose data, engineering data models for optimized storage and visualization, organizing, disseminating geospatial data, and leveraging geospatial data by integrating it with other data.

xxi

Smart Sustainable Urbanism Bakr Aly Ahmed North Dakota State University, USA

Biography: Bakr Aly Ahmed has nearly 27 years of academic and teaching experience in architecture and design criticism. Prior to joining the faculty of NDSU, he has completed his Ph.D. in environmental design and planning from Virginia Tech and a master’s in architecture focusing on the prefabrication of building and the mass production of housing projects. In parallel, he has maintained a professional practice of consultant and design work in numerous projects which included beach resorts, housing developments, and mix-use urban projects. His research interest focuses on sustainable design modeling, environmental capacity measurements, and simulation modeling for pedestrian movement in large buildings.

xxiii

Contents

Smart City Smart City Research Between 1997 and 2020: A Systematic Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Souad El Hilali and Ahmed Azougagh

3

Human Computation Based Platform for Citizen Services in Smart Cities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adnan Yahya, Yazan Yahya, Nibras Misk, and Hamzah Hijja

17

Smart Home Study Within the Scope of Urban Transformation Project: Case of MAtchUP Antalya Project . . . . . . . . . . . . . . . . . . . . . . Neşe Özçandır and Sevim Ateş Can

29

Spatial Analysis for Smart City Approach: The Case of Beşiktaş-Etiler Neighborhood . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anıl Çakir, Enver Murat Karababa, Furkan Talha Erdemir, Berfin Şenik, and Elif Kutay Karaçor Utilization of the Visiting Jogja Mobile Application as a Provider of Information Regarding Limitations of Tourism Activities During the COVID-19 Pandemic in the Special Region of Yogyakarta . . . . . . . Candra Triastutiningsih and Rini Rachmawati Intelligent Competitiveness of Logistics Companies Based on Benchmarking Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mohamed Achraf Laissaoui, Ouail El imrani, and Aziz Babounia An Integrated Human-AI Framework Towards Organizational Agility and Sustainable Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . Mohamed Amine Marhraoui, Mohammed Abdou Janati Idrissi, and Abdellah El Manouar

41

53

65

73

xxv

xxvi

Contents

The Place of Stock Photography as a Digital Commerce in Turkey . . . . İsa Avcı, Murat Koca, and Büşra Uysal

89

Fuzzy Classification of the Flow of Events for Decision-Making in Smart Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Anatolii Kargin and Tetyana Petrenko A Decision Tree-Based Model for Tender Evaluation . . . . . . . . . . . . . . 115 Samuel Kumbu Mandale and Bernard Shibwabo Kasamani A Centralized Credit Scoring Prototype for Microlending Institutions Using Neural Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Law Karingithi Maina and Bernard Shibwabo Kasamani Smart & Sustainable Rural Settlements Exam–The Plateau of Obruk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Aziz Cumhur Kocalar Smart Mobility and Intelligent Infrastructures Bayesian Regression Model Estimation: A Road Safety Aspect . . . . . . . 163 Magda Marek Improving Vehicle Localization with Two Low-Cost GPS Receivers . . . 177 Elnaz Namazi, Rudolf Mester, Chaoru Lu, Markus Metallinos Log, and Jingyue Li Current Trends in Smart Cities: Shared Micromobility . . . . . . . . . . . . . 187 Rukiye Gizem Öztaş Karlı and Selma Çelikyay A Resilient Smart Architecture for Road Surface Condition Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 Vincenzo Agate, Federico Concone, and Pierluca Ferraro A New Graph Method Based on Deep Learning for Smart Intersections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211 Erhan Turan, Beşir Dandıl, and Engin Avcı Smart Service Supply Chain and Just Walk Out Technology: A Netnographic Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 Badr Bentalha and Aziz Hmioui Anomaly Detection in Region Mobility Utilization Using Daily Taxi Trajectory Dataset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237 Yesim Dokuz and Ahmet Sakir Dokuz Collaborative Ant Colony Multi-agent Planning System for Autonomous Mobile Robots in a Static Environment . . . . . . . . . . . . . . . 249 Chaymaa Lamini, Said Benhlima, and Moulay Ali Bekri

Contents

xxvii

Smart Energie Management CFD Study of the Flow and Heat Transfer Through an Unvented Trombe Wall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269 Zouhair Charqui, Mohammed Boukendil, Lahcen El Moutaouakil, Rachid Hidki, and Abdelhalim Abdelbaki Energy Management Techniques in Off Grid Energy Systems: A Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281 Mohamed Elweddad, Muhammet Tahir Guneser, and Ziyodulla Yusupov Effect of Thermal Radiation on Natural Convection in an Air-Filled Cavity with an Inclined Heat-Generating Elliptical Body . . . . . . . . . . . . 293 Rachid Hidki, Lahcen El Moutaouakil, Mohammed Boukendil, Zouhair Charqui, and Abdelhalim Abdelbaki Numerical Simulation of Third-Generation Solar Cells Based on Kesterite CZTSSe Using SCAPS-1D . . . . . . . . . . . . . . . . . . . . . . . . . 305 Lhoussayne Et-taya, Touria Ouslimane, and Abdellah Benami Nonlinear Backstepping Control for Photovoltaic System Connected to the Grid Through Inverter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315 Fatim-Zahra Zaghar, El Mehdi Karami, Mohamed Rafi, and Abderraouf Ridah HBIM and Thermal Performance in Historical Buildings . . . . . . . . . . . 327 Ö. Özeren and M. Korumaz Thermal Modeling for Underground Cable Under the Effect of Thermal Resistivity and Burial Depth Using Finite Element Method . . . 339 Abdullah Ahmed Al-Dulaimi, Muhammet Tahir Guneser, and Alaa Ali Hameed Impact of Covid-19 Pandemic on Smart Natural Gas Grids and Infrastructure Companies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353 Cevat Özarpa, İsa Avci, Bahadir Furkan Kinaci, Hamza Yetik, and Suat Arapoğlu Smart Devices and Intelligent Softwares Programming Nao as an Educational Agent: A Comparison Between Choregraphe and Python SDK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367 Anushka Subedi, Dipesh Pandey, and Deepti Mishra Exploring and Extending Research in Multi-vendor Software Ecosystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379 Anshul Rani, Deepti Mishra, and Aida Omerovic

xxviii

Contents

Pre-planning Process Model in Agile Global Software Development . . . 393 Hajar Lamsellak, Houda Metthahri, Mohammed Ghaouth Belkasmi, and Mohammed Saber Software Quality Prediction Using Machine Learning . . . . . . . . . . . . . . 401 Bhoushika Desai and Roopesh Kevin Sungkur Phone Wallet for Mobile Payment in Algeria . . . . . . . . . . . . . . . . . . . . . 413 Abdelkader Belkhir, Maria Belkhir, and Fayçal Bouyakoub TV Recommendation for Multiple Users Based on Movie Ratings . . . . . 421 Wassila Guebli and Abdelkader Belkhir Establishment of a Watch Platform of Public Sustainable Purchase in Morocco . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435 Tarik El Haddadi, Mohamed Ben Ahmed, and Taoufik Mourabit Smart E-Healthcare Data Encryption for E-Health Service . . . . . . . . . . . . . . . . . . . . . . . . . . 447 Karima Djouadi and Abdelkader Belkhir A Deep Learning Approach for the Diabetic Retinopathy Detection . . . 459 Riad Sebti, Siham Zroug, Laid Kahloul, and Saber Benharzallah An Innovative Respiratory Rate Detection System Using Adaptive Filter with Speech Boundaries Detection Algorithm in Audio Signal . . . 471 Ahmet Reşit Kavsaoğlu and Mohamed Elhashmi Feature Extraction Methods for Predicting the Prevalence of Heart Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481 Ivoline C. Ngong and Nurdan Akhan Baykan Quality Attributes for Evaluating IoT Healthcare Systems . . . . . . . . . . 495 Loubna Chhiba, Abdelaziz Marzak, and Mustapha Sidqui Early Prediction of ICU Admission Within COVID-19 Patients Using Machine Learning Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507 Ikram Maouche, Sadek Labib Terrissa, Karima Benmohammed, Noureddine Zerhouni, and Safia Boudaira Agent-Based Model for Analyzing COVID-19 Infection in the Campus Using AnyLogic Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 519 W. X. Gan and S. Amerudin

Contents

xxix

Smart GIS and Earth Management Smart Prediction System for Territorial Resilience at the Large-Scale Level. Case Study of the Seasonal Forest Fires Risk in Northern Morocco . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533 Hicham Mharzi-Alaoui, Jean-Claude Thill, H. Bahi, H. Hajji, F. Assali, and S. Moukrim Mapping of the Study Area with GIS a Tool for the Description of Study Sites in Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549 Hajar El Omari, Abdelkader Chahlaoui, Fatima Zahra Talbi, Abdelkarim Taam, and Abdelhakim El Ouali Lalami Geodesign – a New Approach for Rapid Development of Planning and Carbon Sequestration Scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . 559 Fred Barış Ernst, Abdullah İzzeddin Karabulut, and Mehmet İrfan Yeşilnacar Assessment of Rapid Urbanization Effects with Remote Sensing Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 571 Nur Yagmur, Adalet Dervisoglu, and B. Baha Bilgilioglu Indexing Approach for the Evaluation of Heavy Metals in Drinking Water Produced by a Moroccan Water Treatment Plant . . . . . . . . . . . . 587 Abderrahman Achhar, Mohamed Najy, Driss Belghyti, and Almehdi Alibrahimi Envirolarm: A Mobile App to Manage Natural Hazards – Scenarios for a Small Island States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597 Rikeelesh Kumar Ramjattun, Mainkah Shicksha Rampersad, and Roopesh Kevin Sungkur Mechanical Characterization of a Geoconcrete Composite: Laterite with Addition of Peanut Shell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 607 Amadou Warore, Biram Dieng, Seydou Nourou Diop, and Senghane Mbodj Smart Water Management Characteristics and Assessment of Heavy Metals in the Water of Lake Sidi Boughaba (Kenitra, Morocco) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621 Mohamed Najy, Fatima Zahra Talbi, Hassan Ech-chafay, Omar Akkaoui, Nordine Nouayti, and Driss Belghyti Multiple Water Reservoirs in African Continent: Scarcity, Abundance and Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 629 Ahmed El Bakouri, Mourad Bouita, Fouad Dimane, Mohamed Tayebi, and Driss Belghyti

xxx

Contents

Seasonal Dynamics of Sandflies and Soil Texture of Breeding Sites, Aichoune Locality, Sefrou Province, Morocco . . . . . . . . . . . . . . . . . . . . 645 Fatima Zahra Talbi, Mohamed Najy, Hajar El Omari, Abdelkarim Taam, and Abdelhakim El Ouali Lalami Hydrogeochemical Study of the Hamma My Yacoube, Sidi Slimane – Morocco . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 657 Salah Aitsi, Jalal Ettaki, Khalid Doumi, Ahmed Chabli, and Driss Belghyti Hydrogen Production via Wastewater Electrolysis—An Integrated Approach Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671 M. Cartaxo, J. Fernandes, M. Gomes, H. Pinho, V. Nunes, and P. Coelho Flood Aleas Diagnostic and Assessment Case of the Jebha Zone . . . . . . 681 Mohammed Benessayyad, Soufiane Saber, Driss Belghytı, and Kacem Naımı Assessment of the Intensity of Floods and Study of Their Impact on the Ourika Area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 691 S. Saber, M. Benessayyad, M. S. Elyoubi, and D. Belghity Smart Education and Intelligent Learning Systems Teaching and Learning in a Virtual Environment: The Case of a Regulated Access Institution in Morocco . . . . . . . . . . . . . 705 Nadia El Ouesdadi and Sara Rochdi Authoring Systems in Computer-Based Education: Learning Efficacy and Opportunities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 719 Oussama Hamal, Housseine Bachiri, Nour-eddine El Faddouli, and Samir Bennani Toward Using Cloud Computing at Universities in Developing Countries Considering the Covid-19 Crisis . . . . . . . . . . . . . . . . . . . . . . . 733 M’rhaouarh Ibtissam, Chafiq Nadia, and Namir Abdelwahed The Influence of Mathematics on Students’ Performance in Computer Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745 Mayowa A. Sofowora, Seraphin D. Eyono Obono, and Abdultaofeek Abayomi Comparison of the Availability of Online Platforms for Distance Instrument Training According to Various Variables . . . . . . . . . . . . . . 757 Mert Ergül and Şevval Satıcı A Technological Transformation in Music Talent Exams: A Start for Smart Technology with the BILSEM Music Diagnostic Exam Post-2018 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767 Ahmet Serkan Ece, Hasan Hakan Okay, Sefa Zeybel, and Şevval Satıcı

Contents

xxxi

Data Science Technologies and Social Media Analysis Toward a Smart Approach of Migration from Relational System DataBase to NoSQL System: Transformation Rules of Structure . . . . . . 783 Abdelhak Erraji, Abderrahim Maizate, and Mohammed Ouzzif A New Algorithm for Data Migration from a Relational to a NoSQL Oriented Column Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 795 Ahmed Dourhri, Mohamed Hanine, and Hassan Ouahmane Improving a New Data Lake Architecture Design Based on Data Ponds and Multi-Agent Paradigms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 815 Jabrane Kachaoui and Abdessamad Belangour Data Lakes: A Survey Paper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 823 Mohamed Cherradi and Anass EL Haddadi Automatic Sarcasm Detection in Dialectal Arabic Using BERT and TF-IDF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 837 Soukaina Mihi, Brahim Ait Ben Ali, Ismail El Bazi, Sara Arezki, and Nabil Laachfoubi MAC: An Open and Free Moroccan Arabic Corpus for Sentiment Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 849 Moncef Garouani and Jamal Kharroubi Sentiment Analysis Using Machine Learning and Deep Learning on Covid 19 Vaccine Twitter Data with Hadoop MapReduce . . . . . . . . 859 Seda Kul and Ahmet Sayar Image Processing, Recognition Systems and 3D Modelling Classification of RASAT Satellite Images Using Machine Learning Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 871 Sohaib K. M. Abujayyab, Emre Yücer, I. R. Karas, I. H. Gultekin, O. Abali, and A. G. Bektas Study the Effect of Noise on Compressed Images Used in Smart Application Based on JPEG Standard . . . . . . . . . . . . . . . . . . . . . . . . . . 883 Elawady Iman and İsmail Rakıp Karaș Comparative Study Between the Rectangular and Trapeze Design of Plasmonic Nanoparticles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 893 H. Oubeniz, Z. Oumekloul, Y. Achaoui, A. Mir, and A. Bouzid Impact of Standard Image Compression on the Performance of Image Classification with Deep Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901 Tajeddine Benbarrad, Marouane Salhaoui, Hatim Anas, and Mounir Arioua

xxxii

Contents

Ship Detection in Optical Remote Sensing Images Using YOLOv4 and Tiny YOLOv4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 913 Esra Yildirim and Taskin Kavzoglu Wall Size Prediction from 2D Images with the Help of Reference Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 925 Seda Kul and Ahmet Sayar Improvements on Road Centerline Extraction by Combining Voronoi Diagram and Intensity Feature from 3D UAV-Based Point Cloud . . . . . 935 Serkan Biçici and Mustafa Zeybek Towards a 3D Real Estate Valuation Model Using BIM and GIS . . . . . 945 Muhammed Oguzhan Mete, Dogus Guler, and Tahsin Yomralioglu Classification of Mobile Laser Scanning Point Cloud in an Urban Environment Using kNN and Random Forest . . . . . . . . . . . . . . . . . . . . 963 Semanur Seyfeli and Ali Ozgun Ok Deep Learning Models Performance Evaluation of Transfer Learning for Surface Defect Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 977 Tajeddine Benbarrad, Mounir Arioua, and Hatim Anas Generic Automated Implementation of Deep Neural Networks on Field Programmable Gate Arrays . . . . . . . . . . . . . . . . . . . . . . . . . . . 989 El Hadrami Cheikh Tourad and Mohsine Eleuldj A Deep Convolutional Neural Networks for the Detection of Cervical Cancer Using MRIs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1001 Ichrak Khoulqi and Najlae Idrissi Product Quality Prediction of 95% Naphtha Cut Point in Crude Distillation Unit Using Artificial Neural Networks . . . . . . . . . . . . . . . . . 1011 Filiz Al-Shanableh A Smart Recipe Recommendation System Based on Image Processing and Deep Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1023 Seda Kul and Ahmet Sayar IoT Technologies and Connectivity Architectures 5G Implementation in Ibn Tofail University . . . . . . . . . . . . . . . . . . . . . 1037 Hafida Amgoune and Tomader Mazri Designing a LoRa Network Using Dijkstra’s Algorithm . . . . . . . . . . . . . 1047 Ali Semih Yılmaz and Özlem Öztürk

Contents

xxxiii

Android Application Test for GPS Geolocation Using CN . . . . . . . . . . . 1057 S. M. H. Irid, M. H. Hachemi, H. E. Adardour, and M. Hadjila A Predictive and Scalable Architecture Based on IoT and Fog Computing for Smart City Applications . . . . . . . . . . . . . . . . . . . . . . . . . 1071 Boudanga Zineb, Benhadou Siham, and Leroy Jean-Philippe Smart Security Conceptual Model for Crowd-Sourcing Digital Forensic Evidence . . . . . 1085 Stacey O. Baror, H. S. Venter, and Victor R. Kebande The Proposed Self-defense Mechanism Against Security Attacks for Autonomous Vehicles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1101 Tomader Mazri and Siham Tibari Chaotic Light Weight Authentication Protocol for Vehicular Adhoc Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1113 G. Kothai and E. Poovammal Security Classification of Smart Devices Connected to LTE Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1125 Samatar Mohamed Ali, Muhammet Çakmak, and Zafer Albayrak Performance of Ad-Hoc Networks Using Smart Technology Under DDoS Attacks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1133 Aden Ali Said, Muhammet Çakmak, and Zafer Albayrak A Comprehensive Evaluation of Cryptographic Ciphers on Secure Publish/Subscribe Communications for IoT Devices . . . . . . . . . . . . . . . . 1141 Seda Kul and Ahmet Sayar Correction to: Multiple Water Reservoirs in African Continent: Scarcity, Abundance and Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . Ahmed El Bakouri, Mourad Bouita, Fouad Dimane, Mohamed Tayebi, and Driss Belghyti

C1

Smart City

Smart City Research Between 1997 and 2020: A Systematic Literature Review Souad El Hilali and Ahmed Azougagh

Abstract Smart city has been a subject of great interest in research and practice. Since its first appearance in early 1998, the Smart City concept is still unclear in terms of context and perspective. The aim of this article is to track the evolution of this emergent field of research between 1997 and 2020 through a systematic literature review based on the Theory method.







Keywords Smart City research Systematic literature review Grounded theory Evolutionary perspective




1 Introduction The economic development and the technological advances of the second half of the twentieth century have contributed to the promotion of urban development. Thus, it has led to a rural migration of the population to cities that can offer their inhabitants more opportunities for work, education, quality of life, etc. [1–3]. Similarly, according to Florida [4] and Kourtit [5], current trends indicate the third revolution in urban development. Cities are no longer only clusters of inhabitants but generators of creative and innovative potential. However, this rapid growth of the last thirty years has resulted in many challenges related to limited resources, pollution and, social inequalities. Therefore, there is a need for more innovative management of cities [6, 7]. Several approaches can be considered to address these challenges. Some approaches rely on the use of information and communication technologies [8, 9]. Other approaches rely on human capital including learning, creativity, cooperation among relevant actors, and the generation of new knowledge [10, 11]. Cities that have succeeded in addressing these challenges in a smart and innovative way have achieved the label of “Smart City”. This concept is gaining S. El Hilali (&)
A. Azougagh LIREFMO Laboratory, Faculty of Economics and Management, Sidi Mohamed Benabdellah University, Fez, Morocco e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_1

3

4

S. El Hilali and A. Azougagh

growing attention from academics and practitioners becoming the new paradigm of smart and innovative urban development [12]. This article presents a systematic literature review on the smart city. The systematic literature review is based on Grounded Theory method. The objective of our literature review is to analyze the existing literature and to see how it has evolved over the last decades to contribute to the advancement of knowledge in this field. We expect that this literature review—which aims to cover a large part of the literature—will bring out relevant concepts and eventually research questions.

2 Method 2.1

A Systematic Literature Review

The art and science of gathering information from primary data is a very critical but often unaddressed area of research. Systematic literature reviews are an indispensable tool in today’s research practice. According to Tranfield [13], they differ from narrative reviews by their use of a scientific, transparent, and replicable process. These literature reviews make it possible to reduce bias through an in-depth search of the various studies undertaken on a particular subject. These types of reviews, being well structured and probative, give users confidence and are a source of information about the evolution of knowledge on a given subject [14]. Systematic literature reviews do not provide answers but they do record what is known and what is not known concerning the research question [15]. In this article, we will apply a systematic and rigorous approach to conduct a literature review using Grounded Theory [16] as an analysis process.

2.2

Grounded Theory Approach for Reviewing Literature

Grounded Theory. Grounded theory was founded by the two sociologists Glaser and Strauss in 1996 in American hospitals as a result of their experience with near-death patients [13]. The main question the two authors asked was: how to generate a theory from data collected and analyzed in a rigorous manner. The goal of a grounded theory approach is to generate theories. In other words, a researcher who adopts a grounded theory approach is guided by data. Grounded theory is based on data. Data may be empirical or theoretical. In fact, grounded theory can be used in conducting literature reviews as a method of analysis. The use of this method in literature reviews consists of searching, selecting, analyzing, and presenting data from the literature in a rigorous manner in order to highlight a concept or develop new theories. There are two schools of thought in grounded theory, the Glaserian school [14] and the Straussian school [15] which gave rise to the Charmazian grounded theory [16] after an epistemological break from positivism to constructivism. We will adopt the Straussian school of thought to evoke grounded theory in systematic literature reviews.

Smart City Research Between 1997 and 2020 ...

5

Table 1 Five stage approach for reviewing literature Stages

Steps

1. Definition

• • • • • • • • • •

2. Research 3. Selection 4. Analysis

5. Presentation

Defining inclusion and exclusion criteria Identifying research areas Finding the relevant sources Choosing the keywords Searching the bibliographic references in the different sources Refine the document sample Open coding Axial coding Selective coding Present and structure the content

Five Stage Approach for Reviewing Literature. The purpose of using a Grounded Theory approach to the literature review is to achieve a substantive and theoretically relevant analysis of the topic at hand. Many good reviews have used a similar approach, largely in a tacit way. The Grounded theory method of literature analysis has five steps and is iterative in nature. In the first “Definition” step, four steps are taken to identify the most appropriate data set. It is only in the second “Research” step that the research studies are actually conducted. The third “Selection” step refines the sample of studies to be examined. The fourth “Analysis” step shows how qualitative research methods, based on Grounded Theory, obtain real value from the selected studies. The fifth step, “Presentation”, includes the two key steps of writing a coherent synthesis document, which should show not only the results and insights obtained but also the main decisions made during the literature review process (Table 1).

3 Results 3.1

Definition

Inclusion and Exclusion Criteria. Since the concept of “smart city” is an emerging concept, we are interested in all existing literature on the subject. The articles included in the review must contain the word “Smart City” or “Smart Cities” in their abstract and/or in their full text. Both conceptual and empirical researches related to the smart city were included. Hence, articles that mention the word smart city without treating it in their content are excluded. Furthermore, the literature analyzed is essentially Anglo-Saxon with a limited number of French references given the abundance of Anglo-Saxon literature compared to French literature. The literature also includes grey literature sub-subject to intellectual property rules. The articles taken into account must contain the definition of the smart city. The period considered is between 1997, the year in which the concept

6

S. El Hilali and A. Azougagh

Table 2 Stages of the systematic literature review based on Grounded Theory Inclusion criteria

Exclusion criteria

Topic

Documents containing the word “Smart City” or “Smart Cities”

Type of study Language Document types Content of the documents Publication period

Conceptual and empirical studies English and French Peer-reviewed articles and grey literature Definition of “Smart City”

Other documents not containing “Smart City” or “Smart Cities” in their text – Other languages –

1997–2020

Any other document that does not address the subject in its content –

was born, and 2020. The choice of this period aims to integrate all the existing literature about the smart city to track its evolution. Table 2 summarizes the inclusion and exclusion criteria considered. Identification of the Research Area. The concept of the smart city is a multidisciplinary one and is of interest to researchers from different disciplines (Architecture, Urban Planning, Geography, Data Science, Economic and Management Sciences). We are mainly interested in research that falls within the framework of economic and management sciences. Based on the results of the Web of Science database 2020, we have identified all the disciplines and sub-disciplines in which Smart City research is being conducted. Smart city research is dominated by the Information and Communication Technology (ICT) fields.

Fig. 1 The number of researches on Smart City per domain

Smart City Research Between 1997 and 2020 ...

7

Indeed, the majority of smart city research is in the computer sciences (leading discipline) followed by engineering, social sciences and energy and environment. Figure 1 shows the distribution of research by discipline. Finding the Relevant Sources. The first step in a literature review study is to locate relevant literature by targeting certain international journals and conferences. In general, this approach is appropriate for topics that have been studied for a long period of time and have become a well-developed area of research [17]. However, for a contemporary concept like “smart city”, relevant literature is collected by searching online databases, and this has become an upcoming culture among researchers [18–20]. Thus, we chose the scientific databases made available to researchers by the National Center for Scientific and Technical Research (CNRST) (Fig. 2). Among these different databases, we excluded the databases specialized in natural sciences such as MathScinet, Aluka, Jstore and Global plants. As a result, we have integrated the following databases: Scopus, ScienDirect, Jstore, Web of Science, Springer for the Anglo-Saxon literature and CAIRN.info for the French literature. In addition, we conducted an additional search of the journals included in the 2019 Journal Citation Reports database (Social Science Editions). This literature also included news articles. Additional relevant articles were identified through the review of reference lists of all articles were also included but not limited to peer-reviewed journals. This step is considered by the authors to be imperative in

Fig. 2 List of scientific databases made available to researchers by CNRST

8

S. El Hilali and A. Azougagh

order to take into account relevant work that was not identified by the selected databases. Among these different databases, we excluded the databases specialized in natural sciences such as MathScinet, Aluka, Jstore and Global plants. As a result, we have integrated the following databases: Scopus, ScienDirect, Jstore, Web of Science, Springer for the Anglo-Saxon literature and CAIRN.info for the French literature. In addition, we conducted an additional search of the journals included in the 2019 Journal Citation Reports database (Social Science Editions). This literature also included news articles. Additional relevant articles were identified through the review of reference lists of all articles were also included but not limited to peer-reviewed journals. This step is considered by the authors to be imperative in order to take into account relevant work that was not identified by the selected databases. Choosing Keywords. The queries used to search the major databases contain the term “Smart City” as well as other terms or labels relevant to research in this area. Figure 3 shows the frequency of occurrence of these specific labels. The results presented in the figure above indicate that the “Smart City” label is the most frequently used among other city terms or labels. These are more or less similar to “Smart City”. The “Smart City” label is followed by the “Sustainable City” and “Creative City” labels. This wave of labels can be explained by the technological development, which is largely about computer science. The

Fig. 3 Frequency of occurrence of similar labels used in the literature until 2020

Smart City Research Between 1997 and 2020 ...

9

Fig. 4 Results of using the Boolean operator “OR” in the search

“Sustainable City” label is more related to environmental sciences, while “Creative City” is related to urban studies. In addition, the “Smart City” and “Digital City” labels seem to have relatively more technical connotations, unlike the other labels, which have relatively more environmental and urban connotations. The keywords “Smart city” and “Smart Cities” were used to conduct the search since other researchers used them in their search. To do this the phrase “Smart City” OR “Smart Cities” was entered on the search bar of all selected databases. The objective of using the Boolean operator “OR” is to extract all documents containing both the word “Smart city” and the word “Smart Cities” in their titles, abstracts or texts (Fig. 4). The choice of these keywords was also used by the other researchers.

3.2

Research

The search was conducted simultaneously in the six selected databases. The total number of results obtained is about 140,000. It differs from one database to another. The Scopus database contains the most results (a total of 86,476 documents) which shows that researchers interested in the Smart City field tend to publish their work in Scopus indexed journals (Fig. 5). Moreover, there is a very notable gap between the number of works found on the CAIRN.info reference database (only 427 articles) and on other databases. This proves that this field of research is not yet well developed in the French literature. Figure 6 shows this gap. Furthermore, a time analysis was conducted, the objective of this time study is to analyze the temporal trend and distribution of Smart City research and to understand what the main factors of this temporal trend are. To achieve these objectives, the articles identified in the databases were organized in chronological order and classified according to the year of publication to list them. Figure 7 shows the number of articles on Smart City over the past twenty-three years. As the trend line highlights, the first study regarding this topic was in 1997. This year was marked by

10

S. El Hilali and A. Azougagh

Fig. 5 Number of results obtained per database

Fig. 6 Language used in the research on Smart City

the adoption of the Kyoto Protocol, which aims to reduce greenhouse gas emissions in order to protect the environment. China has become the most productive country in recent decades in terms of Smart City research. In the first decade, the main countries producing research on the topic were mainly European countries. In the second decade, China, Japan and the United States topped the list while European countries were displaced to the bottom. China remains the predominant country. China, Japan, and the United States are at the top of the list (in dark blue in Fig. 8) followed by Italy and the United States (in dark blue in the figure) and other countries with modest research in the subject.

Smart City Research Between 1997 and 2020 ...

11

Fig. 7 Number of papers about smart city

Fig. 8 Smart city research world map (the more intense blue, the higher the productivity). WOS. 2020

3.3

Selection

In accordance with the eligibility criteria, a total of 180 papers were identified from the six selected databases and 191 articles from a direct review of the journals listed in the Journal Citation Reports (JCR). A total of 371 articles were eligible for selection after excluding duplicates. Figure 9 shows the selection process.

12

S. El Hilali and A. Azougagh

Fig. 9 The selection process

The first sorting was based on the title and the abstract. 221 papers were excluded either because they did not deal with the topic in their content. The second

Smart City Research Between 1997 and 2020 ...

13

filtering was based on the reading of the introduction and the conclusion. We supposed that if the results were not included in the introduction or the conclusion, then the topic was not relevant for our analysis. A total of 62 papers were excluded. The final selection of 88 articles was eligible for full-text analysis.

3.4

Analysis

The analysis was conducted document by document by reading each document in order to retain ideas and conclusions relevant to our research. The open coding consists of a re-reading of the selected documents in order to bring out concepts. Our main category is the concept of “Smart City”. A priori we have retained three main approaches to explain the Smart City. A technological approach, a human resource approach and a collaborative approach. These three approaches constitute categories for our main category for which there are sub-categories (Fig. 7). A linkage of the different documents was done in order to extract these categories (Fig. 10).

Fig. 10 The coding process

14

S. El Hilali and A. Azougagh

4 Conclusion It is crystal clear that the smart city concept is still evolving and gaining much interest among researchers from various disciplines and from all over the world. Our literature analysis based on grounded theory method resulted in three main approaches to address the smart city subject. The first one called technological approach, highlights the opportunities provided by Information and Communication Technologies to enhance the urban fabric, stressing technology as the distinctive attribute of a Smart City. Second, the human resource approach through the continual updating and development of social knowledge, this approach emphasizes human and social capital as the driving force of urban transformation. Finally, the collaborative approach, the concept of collaboration is central to this approach, which focuses on the constructive connections between various urban actor networks. These three schools of thoughts are dominant in the literature. In this regard, some comprehensive perspectives of smart city do mix these three approaches putting people in the center of the city’s projects by making technology closer and connecting them to divers stakeholders in a holistic approach in order to respond to the city’s problems. Nonetheless, more theoretical understanding of the smart city concept is needed in order to generate smart city theories. The grounded theory that we used to analyze the literature can also be used to analyze empirical data in order to contribute to the advance of knowledge in the field.

References 1. S. Sassen, The Global City: Introducing a Concept, vol. 11 (2004), p. 27 2. C.R. Berry, E.L. Glaeser, The divergence of human capital levels across cities. Pap. Reg. Sci. 84(3), 407–444 (2005) 3. M. Castells, The Rise of the Network Society, vol. 12 (Wiley, Hoboken, 2011) 4. R. Florida et al., Cities, skills and wages. J. Econ. Geogr. 12(2), 355–377 (2012) 5. K. Kourtit, P. Nijkamp, D. Arribas, Smart cities in perspective–a comparative European study by means of self-organizing maps. Innov.: Eur. J. Soc. Sci. Res. 25(2), 229–246 (2012) 6. T. Nam, T.A. Pardo, Conceptualizing smart city with dimensions of technology, people, and institutions, in Proceedings of the 12th Annual International Digital Government Research Conference: Digital Government Innovation in Challenging Times (2011) 7. J.H. Lee et al., An integrated service-device-technology roadmap for smart city development. Technol. Forecast. Soc. Change 80(2), 286–306 (2013) 8. M. Batty et al., Smart cities of the future. Eur. Phys. J. Spec. Top. 214(1), 481–518 (2012) 9. M.J.C. Angelidou, Smart city policies: a spatial approach. Cities 41, S3–S11 (2014) 10. R. Giffinger, N. Pichler-Milanović, Smart cities: ranking of European medium-sized cities. Centre of Regional Science, Vienna University of Technology (2007) 11. A. Caragliu, C. Del Bo, P. Nijkamp, Smart cities in Europe. J. Urban Technol. 18(2), 65–82 (2011) 12. M. De Jong et al., Sustainable–smart–resilient–low carbon–eco–knowledge cities; making sense of a multitude of concepts promoting sustainable urbanization. J. Clean. Prod. 109, 25– 38 (2015) 13. B.G. Glaser, A.L. Strauss, Awareness of Dying (Transaction Publishers, 1966)

Smart City Research Between 1997 and 2020 ...

15

14. B.G. Glaser, Conceptualization: on theory and theorizing using grounded theory. Int. J. Qual. Methods 1(2), 23–38 (2002) 15. A. Strauss, J. Corbin, Basics of Qualitative Research (Sage Publications, 1990) 16. K. Charmaz, Grounded theory: objectivist and constructivist methods. Handb. Qualit. Res. 2, 509–535 (2000) 17. E.W. Ngai, F. Wat, A literature review and classification of electronic commerce research. Inf. Manag. 39(5), 415–429 (2002) 18. M.I. Hwang, R.G. Thorn, The effect of user engagement on system success: a meta-analytical integration of research findings. Inf. Manag. 35(4), 229–236 (1999) 19. R. Sabherwal, A. Jeyaraj, C. Chowa, Information system success: Individual and organizational determinants. Manag. Sci. 52(12), 1849–1864 (2006) 20. S. Petter, E.R. McLean, A meta-analytic assessment of the DeLone and McLean IS success model: an examination of IS success at the individual level. Inf. Manag. 46(3), 159–166 (2009)

Human Computation Based Platform for Citizen Services in Smart Cities Adnan Yahya , Yazan Yahya, Nibras Misk, and Hamzah Hijja

Abstract We describe a platform that utilizes the power of human computation to provide improved services to citizens in a smart city context. The platform enables users to report problems to authorities and facilitates communication between parties to work on solutions to encountered problems. An important aim is to involve citizens in the city decision-making process. System users can be registered who provide necessary information directly or through their social networking accounts. Alternatively, users can proceed as guests and use the system anonymously, without registration. For registered users the profile information and use history can be used to strengthen the validity/trust of their contributions. Users can perform several actions on the platform including: reporting a problem using username or anonymously, voting on reports submitted by others, reviewing the status of their reports and responding to requests to perform specific tasks on their travel route. We describe the system and focus on human computation aspects, including: user privacy issues, data collection methods, data reliability and data quality assessment, user engagement incentives and reward system, models of interaction between the user and the system (Push/Pull) and mention frameworks used for system implementation and testing. Although focused on citizen services, the platform is flexible and extensible to tasks in other areas. Keywords Smart City services Data-driven services


Human computation
Citizens services


A. Yahya (&)
Y. Yahya
N. Misk
H. Hijja Electrical and Computer Engineering Department, Birzeit University, Birzeit, Palestine e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_2

17

18

A. Yahya et al.

1 Introduction and Background 1.1

Motivation

The huge technological advances are affecting all aspects of life and many daily activities are increasingly employing modern technological solutions for solving problems. Several factors encouraged us to investigate the use of human computations for better citizens services in a smart city context. Human Computation (HC) is defined as the approach to solve computational tasks, particularly those tasks that are hard to automate, through the contribution of humans [14]. Technology supported city services are still weak in our country. The main aim of the HC platform is to strengthen communications between citizens and authorities to enhance collaboration on reporting and solving different kinds of problems affecting citizens life and essential for smooth and efficient functioning in terms of time and cost. Another aspect is raising the trust level between citizens and responsible authorities and allowing citizens to be informed about city activities in a timely and transparent manner. A solid relationship between citizens and authorities makes users important and active link in the services– governance chain. We also wanted to employ the concept of CrowdSourcing which allows the integration of human effort into the decision making processes directed at solving critical problems. Combined with HC elements, CrowdSourcing is essential when dealing with problems that even computer systems cannot solve easily.

1.2

Related Systems

Human Computations for improved citizen services are becoming an important approach to turn to. Here are some examples: • Youknow: a social networking platform for problem solving and deployment solutions. It allows one to communicate with decision makers, based on the principle: “You have fair rights and you have duties”. In Youknow users specify to whom the problem is directed. Decision makers look at these problems, reply to them and make a decision. The platform was launched March 1, 2015 [12]. • City Smart Services: a cross-platform application, designed for municipalities. Citizens use the application to submit public issues to the municipality. The municipality receives the reports and assigns cases to available technicians to solve the issues raised. The technician receives a notification about the case, accepts the job and fixes the problem on location and declares the problem solved [13].

1.3

Distinguishing Features of our Platform

After reviewing the existing systems, we strived to have our system posses features that make it appealing to potential users: individuals and institutional. Here is a summary of our system features:

Human Computation Based Platform for Citizen Services …

19

• Supports the concept of human computation, things computer technologies cannot do well, and a single user cannot adequately report the problem so we let users do it as community and their aggregate input is the main system input [1, 2]. • Outsources work by encouraging the contribution of members of the public (community) of tasks that are usually handled by institutions. • Takes data trustworthiness in high consideration and has rules and acceptance criteria for user data that affects and is affected by user trustworthiness. • Is cross-platform: the resulting app runs on both iOS and Android devices. • It works in multiple modes: people volunteer to report but are also asked to report. The system uses two-way communications between users and authorities: users report to authorities and authorities ask users for their input on problems, which makes our approach more engaging and democratic. Users can also monitor the progress of their reports (status of the report: submitted, in progress, solved). • The system is extensible: it is easily tailored for other applications of similar nature, there is no need to rewrite the code for additional applications of similar nature. Potential applications are by no means limited to municipalities: one can easily think of applications like monitoring wireless connectivity in a smart city setting, monitoring water availability and leakage problems in a town. • Care for the user: user information is in safe hands. The passwords are saved in encrypted form and the user is always aware of the current privacy settings and have the option of anonymous reporting.

2 System Description and Design In this section we describe the system we implemented to utilize HC for improved municipal services; its components, functionality and distinguishing features.

2.1

User Profiling

User Profiling is the process of collecting information about the user. This information can be used to understand users and appreciate their interests/capabilities for enhancing the user input and assessing its quality and for providing better user experience/satisfaction[4]. The main approaches of user profiling are: • Explicit User Profiling: also called Static Profiling. In this approach, the user profile is constructed from user’s data one gets by filling forms. There are some problems when we only depend on explicit profiling as users are reluctant to share their actual information out of concern about privacy or because the form

20

A. Yahya et al.

filling process might be boring so the user may try to avoid it or provide incorrect information [4]. • Implicit User Profiling: also called Profile Extraction. Here one extracts the needed user information from different sources, such as web pages or social media platforms like Twitter, Google plus and Facebook[4]. Another source is user behavior-based techniques which help user profiling systems gather information of interest about users, for example by observing user web navigation patterns. For our platform, we allow both profiling methods. We give the option to the user to register using explicit profiling, either by filling forms or we extract user information from social media (Facebook or Google) accounts. Furthermore, we allow Guest login to the platform, which means using functionalities without registration. However, we decided to restrict the number of daily logins for guest users to discourage spammers. For this purpose, we used Universally Unique IDentifier (UUID), a.k.a. Globally Unique IDentifier (GUID): a 128-bit number used to identify information in computer systems. When generated according to the standard methods, UUIDs are unique for practical purposes, without depending for their uniqueness on a central registration authority or coordination between the parties generating them. While the probability of duplicate UUIDs is not zero, it is close enough to zero to be negligible [3]. Version 1 UUIDs, the most common, combines a MAC address and a timestamp to produce sufficient uniqueness. In the event of multiple UUIDs being generated fast enough that the timestamp doesn’t increment before the next generation, the timestamp is manually incremented by 1. If no MAC address is available, or if its presence would be undesirable for privacy reasons, 6 random bytes sourced from a cryptographically secure random number generator may be used instead. Each mobile device is assigned a UUID and we extract the UUID of the guest user’s mobile device, independent of the operating system (iOS, Android) and save it in our database allowing us to track guest users, logging into the platform repeatedly to prevent spam or suspicious logins.

2.2

Data Collection Methods

Data collection refers to collecting information from participants to help solving problems. When collecting data from devices, there are two choices, either push or pull. Push is the process of sending data from devices to server without prompting, while pull is the process of receiving data from the devices to the server by polling them periodically. There is also another choice called hybrid that uses push and pull together depending on the application context and requirements [5]. Pushing is preferred when the server should send a subset of data to different devices in real time. A disadvantage is that the sent data is limited, so sending a variety of data subsets may bring the model down [5].

Human Computation Based Platform for Citizen Services …

21

Pulling is preferred when the devices will send different types of data to the system. This is a great choice, but also has some disadvantages; first, it may not be in real time, and if it was in real time, it will consume a lot of resources [5]. We settled on a hybrid system, that uses pushing and pulling together. Pulling is used to send user data to the platform; this can be done in real time or delayed if the user is not online, while pushing is used by the platform to send data/requests to users in real time, though reading requests by devices can be delayed if not online. So, in our system pulling is used in the following cases/scenarios: 1. 2. 3. 4.

The The The The

user reports his problem on the platform. user provides his route and waits for system requests of nearby problems. users can vote on the reported problems on the platform. users can respond to the system requests to check on nearby problems.

Pushing, on the other hand, is used in the following cases: 1. The platform notifies the users of a reported problem nearby her/his home location or along her/his route with a request to confirm that the problem really exists/solved. 2. The platform notifies the user when the status of his report changes either to solved or partially solved or unsolvable. The System Data Pipeline given the Information flow is as depicted here: User

2.3

Municipality Personnel

Case Delega-

Solicit Feedback

Case Resolved

User NoƟfica-

Location Awareness

Most modern smart phones come with built-in GPS sensors to get earth coordinates. In our system we use GPS and Google Maps API to provide two features: 1. Location detection: during the reporting process, the reporter may need to provide location of the problem. For that the device GPS must be enabled. 2. On_my_route (directions and routing) tasks: where the system asks the user to define a route by providing a starting point and destination. The system uses Google Directions API to get the route then looks for reported problems within a given distance from the route for the user to look into them by pushing the problems to the user inbox. The user can access the reports displayed on his/her device and act on them while travelling (Fig. 1). For that, we used Google Maps API and the Google Directions API:

22

A. Yahya et al.

Google Maps API: allows users to customize maps with own content and imagery for display on mobile platforms and websites. This API provides four basic map types: roadmap, satellite, hybrid, and terrain [7]. Google Directions API: a service that calculates routes and directions between locations using an HTTP request. This API can calculate directions using one of several modes of transportation, including walking, driving or cycling. It returns directions as a series of waypoints and specifies destinations, origins, and waypoints as strings [8]. The API calculates directions and returns the most efficient routes. The API considers many factors like time, distance, number of turns. Calculating directions is a time and resource intensive process [8]. Once we have the trip route, our system ranks the reported problems by their distance from the travel route and pushes the closest to the user as suggested tasks for possible feedback.

2.4

Data Quality Assessment

User provided data may not always be trusted, accurate or even correct. Thus, one needs to make sure that the quality of such data is properly evaluated before incorporation into the decision-making process [16]. Data Trustworthiness: in any Human Computation System, Data Trustworthiness is one of the most important issues to consider carefully. Since the platform is

Fig. 1 General and On_My_Route Functionality Screen Shots

Human Computation Based Platform for Citizen Services …

23

dependent on human contribution, certain rules must be set to ensure data trustworthiness and to deal with any errors inherent with information provided by human users. Looking for typical approaches to validate user data, we found the following to be of interest [6]: 1. Using majority vote to assess the correctness of data and ensure trustworthiness. 2. Gathering redundant data from users to check the agreement level of reporting users on the same piece of information. For example, collect multiple problem reports and check if some validate others submitted previously. This approach can be costly and time-consuming. 3. Evaluation of user credibility based on quality of information provided earlier then assigning trustworthiness levels of data based on user credibility. Acceptance of generated data will be based on the aggregate (weighted sum) of user-supplied data on the same problem. In our platform for Human Computation Services, we adopted the following approach for assessing the quality of data provided by users: 1. Notify the platform users who are in the same geographical location of the reported problem, and ask them to verify the report. 2. Measure the degree of agreement between respondents on the existence of the problem, whether they are registered or non-registered to assign a metric called “Trust Ratio”. In case of disagreement between users, we use the Cappa Measure (Inter–Judge Agreement). 3. If a user reports a problem and our system declares it trustworthy after proper evaluation, the user’s trustworthiness ranking increases, according to the following formula and his/her future contributions will be weighted accordingly. User Ranking ¼

Number of Accepted Reports  100% Total Number of Reports

4. For municipality services, three major elements increase the quality (trustworthiness) of data if included in the report. These are: i. Digital Attachments: Image or Video demonstrating the problem. ii. Geographical Tagging: of Location where the problem was observed. iii. Time–Stamp: Exact time when the problem was observed. The time stamp and geotags are generated automatically and can be attached to the images, even if data transmission is delayed, say due to lack of internet connection. It is important to stress that data trustworthiness is affected by involved users credibility and that user credibility/rank affects the trustworthiness of the submitted data.

24

2.5

A. Yahya et al.

Assessment of Agreement between Users

It is essential for Human Computation (HC) results to be reproducible. Additionally, we need the results of any Human Computation process to be of high validity or high utility and better than out-of-chance agreement at minimum [14]. Reproducibility can be defined as the extent to which the results of a HC process can be duplicated by different human contributors operating under different conditions using functionally equivalent metrics. Reproducibility is related to consistency of results, meaning that the more the results are consistent, the higher the chance of reproducibility [14]. Measuring Reproducibility: to have confidence in data obtained in HC tasks we must measure reliability, to make sure the evidence obtained is independent of measuring instrument or person. The following three types of reliability are of interest [14]: • Stability: best described as the extent to which the results of a HC process are stable (i.e. unchanging) over time. Stability is focused on results being consistent, such that if someone repeats the same task over time, the results stay the same. • Accuracy: is the extent to which a HC process outputs high validity results. Measuring accuracy requires comparing data obtained by human contributors against valid answers to prior known questions. • Reproducibility: describes the extent to which a process can be repeated many times by human contributors, under different conditions at different locations. Reproducibility mainly has two aspects: inter–rater reliability and inter–method reliability. We focused on inter–rater reliability, which is of most concern for HC. One form of inter–rater reliability is percent agreement but for our HC we opted to use the Cappa Measure (Inter–Judge Agreement) [14].

2.6

Privacy Issues

Many human computation applications are highly dependent on data provided by users, and thus we want to avoid losing users through under-considered or badly documented practices. For highly engaging volunteers in participatory systems, privacy must be supported. Privacy in participatory projects is defined as the right to manage access to personal data contributed voluntarily. Personal data is defined as data that contains identified or identifiable information about the contributor [9]. In our system, we adopt privacy considerations to protect users personal data to encourage users to continue contributing without fear for the privacy of their data and to protect against possible legal litigation and potential lawsuits. Our system must maintain users privacy, and make privacy principles as clear as possible and visible to all. Some design principles were stated to ensure privacy [10]. We include

Human Computation Based Platform for Citizen Services …

25

them in participation terms and give the user the choice to accept or reject participation. These principles are: • Participant primacy: users personal data will be collected with her/his knowledge and explicit agreement. • Data legibility: the users will be informed how their data will be used in the system. • Longitudinal engagement: users will be given the option to change the permissions of sharing their personal data over time if situation changed. Additionally, based on Privacy in Participatory Research [9], some ethical principles were derived for Participatory Research Design. We opted to adopt four of them here: • Ethical Engagement: according to our system guidelines and goals, we will identify the ethical principles and take them into consideration in design, implementation and development processes, and post them beside our system policies. • Informed Participation: detailed explanation of participation as well as legal compliance related issues will be provided to users before accepting any participation. • Evolving Consent: if participation terms change, the system policies will be changed, and users will be informed promptly by repeating the informed consent process. • Evolving Choice: we designed the system to support user’s choice in response to changing contextual situations. For example, a registered user may choose to share personal information in order to receive benefits in one geographic location like a public space but reject sharing the same information in own private home. These principles are offered as guidelines for our system to support stronger relationships between the users and service providers, increase the trust in our system, and encourage user participation by keeping them informed and in control of personal data.

2.7

Reward System

A Reward System is a measure directed at rewarding users to motivate them to participate and use the system. The following clarify the major characteristics of our Reward System: • Typically, the first and most important reward users get is improved city services in that problems they face in their environment are communicated faster to the authorized parties who can work toward a solution. So, instead of a formal

26

A. Yahya et al.

visit to the municipality, users can report the problem whenever they observe it through the platform, without the need for formal, boring procedures. • Whenever a user successfully reports a problem on the platform, a number of points is added to his/her account, and users with highest numbers of points are elevated to premium users with special benefits from municipality [11]. • To encourage participation, we thought of adding elements such as: the first reporter of a problem gets more points when verified, supported reports get better rewarded, quick responses to verification requests get better rewards, inviting new people to enroll and participate results in reward points and so on. • The more points users collect, the higher their ranking becomes. We have three user rankings on the system: Gold, Silver and Bronze, and these are accounted for when assessing trustworthiness of the user and user supplied data.

2.8

Users and Roles

We know that users on the platform are either registered or non–registered (anonymous). On our platform, a series of tasks are always available and geotagged to both registered and anonymous users. However, when it comes to voting on a problem report, registered users vote is given greater weight and higher priority and anonymous user vote value is only 10% of that of a registered user. The platform is flexible in that it accepts multiple roles for the same user: a registered user is allowed anonymous access and non-registered users can upgrade to registered users when desired. The main and most vital role of both types of users is to report problems or collaborate on them. Other roles for users on our platform include: monitoring the status of the report they submitted, including degree of agreement with other users, periodically checking the platform to see if the problem was solved or a feedback is provided. On the other end, designated users manage and keep track of all the reported problems in their domains and their status: these are Authorized Users like a Municipality. Since each reported problem has at least one division responsible for it, we agreed to give direct Authorized Access to a super authorized user in the Municipality who then delegates these rights to the responsible divisions. The workflow is as follows: once a submitted problem is accepted (based on degree of agreement between reporting), it is the super authorized user’s responsibility to forward the problem to the responsible division, and from there the division administrator keeps track of the problem and work towards solving it. Problems not solved for a long time are flagged with red or yellow colors, to alert that follow up is needed. Another point worth mentioning is the fact that our platform is extensible is the sense that any additional tasks and additional roles of users in other fields of human services without the need for major system rewriting.

Human Computation Based Platform for Citizen Services …

27

3 Deployment and Testing We have fully implemented the mobile application, for both iOS and Android devices in addition to the backend components. For the latter we developed a web portal for authorized parties to login and keep track of the problems under their responsibilities, with static login credentials provided by our side, to login and keep track of the problems under their responsibilities. As mentioned earlier, authorized parties (like municipalities) have the ability to change the report status as work progresses. For the implementation, after considering available options we settled on: Ionic for cross platform framework and Firebase for cloud services deployment. The functionalities implemented and available to users are: report a problem, monitor the status of reported problems and check to see if any problems exist on the route a user is taking (on-my-way functionality, Fig. 1). We have released it on a local server. The interface is bi-lingual (Arabic-English) and the language choice is controlled by individual users. During the development, we worked with a local municipality and a group of volunteers. We also asked some potential users to test the main functionalities and the usability of the application, most of their experiments were successful, and we got positive reactions. The users were appreciative of the potential of the platform for improved services to the city residents and provided useful feedback for future modifications.

4 Conclusions and Possible Extensions During the work on the system we faced some challenges. First, is how to encourage citizen users to use the application and our answer was the reward system. Second, is how to trust user data reported on the platform. We agreed to take into account the ranking of the user who had reported and users agreement on the report to decide the degree of trust in the collected data. Third, is how to encourage entities like municipalities to use the application to handle real-life problems reported from real users. Reluctance emanated from concerns regarding many problems being documented and visible to outsiders. Our platform put together the real clients of an organization and the management to help them provide their views and suggestions or talk about the problems facing them and open the way for the organization to view the suggestions and problems of users and work on their solution in a collaborative manner. Fourth, is that the problems encountered by users are now documented, and if not solved, they will be proof that the entity (municipality) failed to cater for its constituents needs, something that may discourage service providers from adopting the system. We tried to emphasize the positive aspects of the system work to allay such fears. The mobile application is still not released on Google Play Store or Apple Store, but we have tested it for the work of main functionalities as well as usability.

28

A. Yahya et al.

Finally, we believe that our system provides the community with an application that utilizes human computation, allowing citizens from community to participate in decision-making process. From another side, the platform makes local authorities (municipalities for example) aware of any problems with public services (garbage collection, street-lights failure, state of streets, etc.) as soon as encountered by citizens which may help save effort and resources and improve response time [15]. Possible improvements include having the approach work on other systems with additional tasks and additional roles. Processing of textual data for better system performance is another improvement aspect. We are also studying adding the ability to “Go Live” when reporting a problem on the platform so that a user can report a problem and engage in real-time and the possible fallout for data quality. “Go Live” feature consumes extra upload bandwidth and thus depends on the network connection type.

References 1. Wikipedia, Human-based computation, https://en.wikipedia.org/wiki/Human-based_ computation. Accessed 2 Aug 2021 2. L. Chen, An overview to human computation. Institute of Information Science, Taiwan (2009) 3. Wikipedia, Universally unique identifier, https://en.wikipedia.org/wiki/Universally_unique_ identifier. Accessed 2 Aug 2021 4. S. Kanoje, S. Girase, D. Mukhopadhyay, User profiling trends, techniques and applications. Int. J. Adv. Found. Res. Comput. 1(1) (2014) 5. SevOne, To Pull or Push IoT Data? That is the Question, https://www.sevone.com/blog/pullor-push-data-iot-data-question, Accessed 11 Mar 2018 6. P. Ipeirotis, P. Paritosh, Managing Crowdsourced Human Computation (Hyderabad University, India, 2011) 7. Google, Maps JavaScript API, https://developers.google.com/maps/documentation/javascript/ tutorial. Accessed 15 May 2018 8. Google, Directions API, https://developers.google.com/maps/documentation/directions/intro. Accessed 15 May 2018 9. A. Bowser, A. Wiggins, Privacy in participatory research: advancing policy to support human computation. Hum. Comput. 2(1), 19–44 (2015) 10. K. Shilton, J. Burke, D. Estrin, R. Govindan, J. Kang, Designing the personal data stream: enabling participatory privacy in mobile personal sensing (Information and Internet Policy, Washington, 2009), pp. 3–4 11. D. Goh, E. Than, C. Lee, An Investigation of Reward Systems in Human Computation Games (Nanyang Technological University, Singapore, 2015) 12. Youknow, ‫ﺇﻧّﻮ ﺑﺘﻌﺮﻑ‬, http://www.youknow.ps/. Accessed 29 Mar 2018 13. City Smart Services, http://citysmartservices.resco.net/. Accessed 13 Mar 2018 14. P. Paritosh, Human Computation Must Be Reproducible (CrowdSearch, 2012), pp. 20–25 15. OECD. Smart Cities and Inclusive Growth (2020), https://www.oecd.org/cfe/cities/OECD_ Policy_Paper_Smart_Cities_and_Inclusive_Growth.pdf 16. C. Neupane, S. Wibowo, S. Grandhi, H. Deng, A trust-based model for the adoption of smart city technologies in Australian regional cities. Sustainability 13, 9316 (2021)

Smart Home Study Within the Scope of Urban Transformation Project: Case of MAtchUP Antalya Project Neşe Özçandır

and Sevim Ateş Can

Abstract The aim of this study is to discuss the concept of “Ecological Smart City” through the “Kepez-Santral Smart Urban Transformation Project”, which includes Kepez and Santral neighborhoods of Antalya’s Kepez District. Known as shantytowns, this area, which contains unhealthy and unstable buildings, was declared a ‘Risky Zone’ within the scope of the “Law on Disaster Relief Areas No. 6306” with the decision of the Council of Ministers dated 24.11.2014 and numbered 7041 and published in the Official Gazette. The practices of the new urban transformation in the concept of “Ecological Smart City”, which was started with the joint decision of more than 3000 landowners living in unhealthy conditions, still continue today. Kepez-Santral Smart Urban Transformation Project, which was prepared under the leadership of Antalya Metropolitan Municipality and received international award and grant support with the European Union MAtchUP project, was started as a pilot project in the fields of smart city, smart transportation and smart housing in Antalya. In addition to the benefits such as demolishing dangerous and risky urban areas and solving property problems, the Project aims to develop a smart area that uses green, livable and innovative technologies. In this study, the technologies used in the housing blocks to be smart within the scope of the MAtchUP project, the anticipated energy consumption and the opportunities it will provide to the users have been evaluated. In this context, a comparison is made between the old and new buildings in terms of residents’ comfort, heating-cooling systems, energy savings, etc. Consequently, in the study, the dimensions of the Ecological Smart City concept are revealed and analyzed through the introduction of the project.




Keywords Urban transformation Ecological Smart City MAtchUP Project Antalya-Kepez





Smart housing


N. Özçandır (&)
S. A. Can Burdur Mehmet Akif Ersoy University, İstiklal Yerleşkesi, 15030 Burdur, Turkey © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_3

29

N. Özçandır and S. A. Can

30

1 Introduction In this study, starting from the concepts of ‘Smart City’ and ‘Smart Home’, the extent of smart home systems included in the concept of ‘Ecological Smart City’, which is an Urban Transformation Project of Antalya Kepez & Santral Neighborhoods, are revealed and analyzed. Literature studies of the concepts of smart city and smart home have been completed and an example is examined based on these definitions. The processes of Kepez&Santral Neighborhoods’ Urban Transformation Project which is one of Turkey’s most important Urban Transformation Projects, were referred. The main purpose of the study is to compare the old houses and new smart home systems in the urban transformation area in terms of energy savings and energy consumption. It is seen that the new houses are more efficient, healthier and have higher energy performance than the old houses. The construction of the area examined in the urban transformation project has been completed, but due to the corona pandemic, the work plan of the smartening system has been delayed, and the smartening system has not yet been integrated. One of the targets of the MAtchUP project, 40% energy saving, is expected to be completed with the introduction of smart home systems. The new residential blocks built in place of the risky and unhealthy buildings before the urban transformation will be more comfortable, energy-efficient and manageable by users. The data to be created by the users will be integrated into the smart city platform. These data will guide the needs and organization of the city in the context of the smart city.

1.1

Smart City

Throughout its history, human beings have tried various methods and developed technologies to meet their needs. “Smartness”, “sustainability” and “smart cities” have become concepts that emerged out of necessity as a result of these experiences. (Zor 2012) The concept of “smart city”, which is the most important point in the development and welfare of societies living in cities, does not have a single definition and has made great contributions to the concepts of urban planning and development at the beginning of the twenty-first century (Greene 2021). According to an estimate by the United Nations, by 2050, 66% of the world’s population will live in urban areas, and this density will pose far-reaching challenges. Issues such as pollution, congestion, waste management and human health are serious issues that need to be adressed and overcomed with sustainable solutions for cities. The European Union (2014) and the United Nations (2016) had an urgent need to develop smart solutions to overcome these problems in climate and energy perspectives and have prepared action plans. In this direction, the smart city

Smart Home Study Within the Scope of Urban Transformation Project …

31

integrates the concepts of “sustainable city” and “digital city” while specifying the requirements for urban innovation and citizenship in new governance, managed by information and communication technology (ICT) (Ateş Can 2014). A smart city has to include forward-looking and sustainable policies. In other words, a smart city must use the features of modernization, technology and digital skills to its advantage in a creative and revolutionary way (Fagadar et al. 2021). The purpose of the smart city is, in a sense, to keep cities livable while continuing to grow and develop, while minimizing the problems associated with increasing population. The ability of a city to be a fully smart city depends on the good execution of many elements. The low emissions of vehicles in traffic, the preference of urban residents for public transportation, efficient and energy-friendly lighting, the maximum performance and efficiency of irrigation and water treatment processes, the supply of energy from renewable sources and many other factors underlie the concept of smart city. In addition to these, of course, it is also important to have advanced information systems and information infrastructure. (Mirghaemi 2019).

1.2

Smart Home

The concept of “Smart house”, which was taken as the subject of the literature in the early 1980s, is defined as a structure that includes automation systems developed to provide building users with advanced command and control over the functions of the building. (Demiris and Hensel 2007; Öztürk and Naimi 2017) Smart home systems can control lighting, temperature, security (window – door control) operations, multimedia and many other functions very easily. (Sine and Koçyiğit 2020) These services, provided by smart home systems, allow the user to instantly monitor and control them from their mobile device or over the web (Peine 2007).

2 Antalya Kepez&Santral Neighbors Urban Transformation The city of Antalya is a city located in the south of Turkey, providing its economic cycle mostly by tourism and agriculture sectors. In this city, which started to receive rapid immigration from the early 1970s, a rapid slum settlement began around the urban periphery. According to the information obtained from the Antalya Governorship Provincial Directorate of Environment and Urbanization, the number of risky buildings in the central districts is shown in Table 1.

N. Özçandır and S. A. Can

32 Table 1 The number of risky buildings

District

Risky buildings

Kepez Konyaaltı Muratpaşa Döşemealtı Aksu

7,660 59 1405 107 20

Especially in Kepez district, which covers the northern parts of the city, the urban structure that became slums in the 90 s caused many physical and sociological problems. Buildings built without being subject to zoning regulations and legal order were at risk of collapse as a result of disasters such as earthquakes. In addition to the disaster risk, the unplanned construction in the region prevented the public services (transportation, health, education facilities, accessibility, etc.) from reaching all citizens. (MAtchUP Homepage 2020). Kepez and Santral Neighborhoods are among the areas declared as “risk zones” for earthquake hazard in Antalya in 2014. Upon this declaration, The Ministry of Environment and Urbanization has commissioned the Antalya Metropolitan Municipality for the re-urbanization of Kepez and Santral Districts. In the project, where it was decided to reconstruct the 1.3 million m2 residential area, there are many residential constructions, as well as public buildings and integrated transportation solutions with other parts of the city. Kepez-Santral District before Urban Transformation is given in Fig. 1. Kepez-Santral District during Urban Transformation is given Fig. 2.

Fig. 1 Kepez-Santral district before urban transformation

Smart Home Study Within the Scope of Urban Transformation Project …

33

Fig. 2 Kepez-Santral district during urban transformation

2.1

Situation Before Urban Transformation

In the region, there were buildings that were built without being subject to legal regulations before the urban transformation. The buildings generally consisted of shanty houses and irregular structures. Although Antalya does not have a very cold climate in winters, the lack of insulation in the buildings caused serious energy losses. In addition to energy efficiency, these structures, which were not durable, were dangerous for human health. In the Fig. 3 a sample Risky Building before Urban Transformation is given.

Fig. 3 Risky building before urban transformation

34

N. Özçandır and S. A. Can

3 MAtchUP Projects and Smart Home Applications The MAtchUP Project is a project that implements Smart City applications in the cities of Antalya—Valencia—Dresden within the scope of the EU Horizon 2020 Smart Cities and Communities (SCC-01) program. The project is carried out with the participation of 8 Countries, 28 Project Partners. One of the main goals of the MAtchUP Project is to use and integrate innovative solutions in the energy, transportation and ICT sectors with a strong monitoring program. Antalya, one of the three pilot cities in the MAtchUP Project, is carrying out a real pilot transformation in the Energy Sector with smart solutions for the integration of smart buildings and homes. These pilots will result in very ambitious energy savings and CO2 emissions reductions while being supported by ICT solutions that encourage the use of open data as the main provider of collecting relevant information. This will be achieved through improvements towards high energy efficiency in buildings, the integration of renewable energy systems and the integration of advanced energy management systems to enable intense building interaction. At the same time, these solutions will be combined with new electromobility solutions (e-vehicles, charging points). One of the most important activities of the MAtchUP project in the field is applications for new high-performance residential construction. It will construct “class B” energy performance buildings, which are well above the Turkish building standards, in 8 residential buildings, which will include 534 flats in Kepez and Santral Districts. In the Turkish Standard TS825, residences have an energy rating of C, but these buildings are designed as high-performance buildings to achieve at least an energy performance rating of B. Energy rating B is defined as having up to 40% better energy performance compared to a standard C-rated building. TS 825 building code is thermal insulation requirements for buildings in the Turkish Standard to energy rating of C. The following Table 2. summarizes the energy demands based on TS825 building code, which has been considered as the baseline value for these new residential buildings (MAtchUP Homepage 2020).

Smart Home Study Within the Scope of Urban Transformation Project …

35

Table 2 Simulated energy demands based on TS825

Block 1 – North-South Orientation

Heating Cooling Lighting Domestic Hot Water Fans&Pumps TOTAL

kWh/m2 year 25.00 41.30 23.70 14.00 4.6 108.60

Block 2 – West -East Orientation kWh/m2 year Heating 21.60 Cooling 59.09 Lighting 23.70 Domestic Hot Water 14.00 Fans&Pumps 5.54 Sub TOTAL 123.93

A better performance will be achieved with several measures including the following interventions: – – – – – – –

Insulation of building envelope Low-u value glazing Energy efficient lighting Gas fired condensing boiler of high efficiency Reference COP of chillers is 3.00 Domestic hot water max 60 °C Only natural ventilation is applied.

The outer walls that are open for weather conditions are designed to have a coefficient of heat permeability of 0.548 W/m2K total (U value). The U value for roof insulations will be 0,428 and 0,730 W/m2K for slabs. This is reached with an insulation material from mineral fibers with a thickness of 50 mm. Double glazed glass with a min U value of 1.8 W/m2K will be used in the building. The lighting of interior will be also based on high-efficient LEDs. The LED lighting systems will not only result in lower installed capacities for lighting but will also lead to energy savings through lower heating effect of lighting in the interior. The energy demand for the interior lighting will be lowered to approximately 4 W/m2 based on an average of 100 lx lighting demand in residential areas.

N. Özçandır and S. A. Can

36 Table 3 Simulated energy demands based on design specifications

Block 1 – North-South Orientation

Heating Cooling Lighting

kWh/m2 year 21.00 33.70 12.30

Block 2 – West -East Orientation

Heating Cooling Lighting

kWh/m2 year 20.60 47.56 12.30

Domestic Hot Water

13.00

Domestic Hot Water

13.00

Fans&Pumps Sub TOTAL

4.10 84.10

Fans&Pumps Sub TOTAL

5.25 98.71

Based on the passive interventions designed for the high performance building the energy demand could be lowered to 84.10 kWh/m2 per year for a building with north–south orientation. The same building would have had an energy demand of 123.93 kWh/m2 based on requirements of TS825 building code. The same building would have had an energy performance of approximately 175 kWh/m2 per year energy demand (G Rating) with no insulation. A significant portion of the building stock in Antalya has no insulation specifically for those buildings constructed before 2008. The following Table 3 summarizes the energy demands based on design specifications new residential buildings (MAtchUP Homepage 2020). Energy efficient lighting systems are used throughout the building and in the landscape areas. The smartening of the houses will be provided by the communication of the objects with each other and the control of the user. With the smartening in residences, users will be able to monitor and control their consumption thanks to smart screens. In the Fig. 4 a sample plan diagram with smart sensors and smart control units in Antalya Kepez&Santral Neighborhoods is given. Figure 5. Data collection schemes on the basis of buildings and residences are shown. Landscape area and exterior facades of buildings can be seen in the Fig. 6.

Smart Home Study Within the Scope of Urban Transformation Project …

Fig. 4 Smart sensors and smart control units plan in Antalya Kepez&Santral

Fig. 5 Data collection schemes on the basis of buildings and residences

37

38

N. Özçandır and S. A. Can

Fig. 6 Landscape area and exterior facades of buildings in Kepez&Santral District

4 Conclusion Smart control systems and high-performance buildings have an important place within the scope of energy solutions for the goals of the Antalya MAtchUP Project. In line with these goals, it is aimed to reduce carbon emissions and environmental pollution as a result of reducing energy consumption and increasing the use of renewable energy in Antalya. Following the pilots, the MAtchUP Project will encourage both citizens and local governments to use renewable energy systems at the urban scale. The urban transformation project is the first of its kind in Antalya and is one of the largest projects in Turkey in terms of scale, and it has brought some challenges with it. Aside from its huge financial and technical requirements, residents in the area had to be relocated during the project timeline. Considering the population of approximately 13,500 people, this operation was managed in close cooperation with the population through participatory processes before and during the project activity. During the covid-19 pandemic that swept the world, disruptions occurred in the project implementation and delays were experienced in the installation of smart home systems. Eligible citizens who have moved to their new homes state that although they do not have smart home systems installed yet, their energy costs have decreased while their comfort increased. It is planned to complete the installation of smart home systems in all 534 residences in the coming months. In the residences built after the urban transformation, the energy consumption is monitored within the scope of the law on the Protection of Personal Data. After the installation of smart home systems, the data flow will continue and will be collected on the urban platform. Thanks to this data, the development and progress of smart city systems will be ensured. In addition to the 40% energy savings provided by the change in the housing class, healthy, innovative and technological opportunities

Smart Home Study Within the Scope of Urban Transformation Project …

39

will be offered to the users. In this study, the effect on the users could not be evaluated because the smart home systems are not integrated in the residential blocks. The project outputs can be updated and presented again in the upcoming conferences.

References A. Peine, The sources of use-information: a review of relevant literature and an exploration into innovation and aging (2007) S. Ateş Can, The concept of smart city and principles of the new approach. J. Archit. 4(2), 12–19 (2014). ISSN 2249-9326 G. Demiris, B. Hensel, Technologies for an aging society: a systematic review of ‘smart home’ applications (2007) Fagadar, C. F., Trip, D.T., Gavrilut, D., Badulescu, D.: Smart cities and the European vision. Ann. Univ. Oradea Econ. Sci. Ser. 30(1), 49–60. https://doi.org/10.47535/1991auoes30(1)004. Accessed 29 Sept 2021 J. M. Greene, Smart city, Salem Press Encyclopedia of Science (2021), https://search.ebscohost. com/login.aspx?direct=true&db=ers&AN=119214377&lang=tr&site=eds-live. Accessed 23 Aug 2021 MAtchUP Homepage, https://www.matchup-project.eu/. Accessed 15 Aug 2021 S.A. Mirghaemi, Akıllı Kentler Üzerine Bir İnceleme: Türkiye Örneği. Beykent Üniv. Fen ve Mühendislik Bilimleri Dergisi 12(2), 37–46 (2019) A. Öztürk, S. Naimi, Akıllı Ev Sistemlerinde Kullanılan Yöntemlerin Farkları, Avantajları ve Dezavantajları. İstanbul Aydın Üniv. Dergisi 36, 115–125 (2017) Ö. Sine, Y. Koçyiğit, İnternet Üzerinden Kontrol Edilen Tam Otomasyonlu Akıllı Ev Sistemleri İçin Örnek Bir Uygulama. Dicle Univ. J. Eng. 11(2), 521–532 (2020) T.C. Cumhurbaşkanlığı Resmi Gazete Homepage, https://www.resmigazete.gov.tr/. Accessed 25 Sept 2021 Turkish Standard EN, TS825, Thermal insulation requirements for buildings, ICS 91.120.10, Ankara, Turkey (2008) United Nations Homepage, https://www.un.org/en. Accessed 25 Sept 2021 A. Zor, Geleneksel Konut Yapılarının Korunmasının Ekolojik Dengeye Sağladığı Katkılar Üzerine Bir İnceleme (Y. Lisans Tezi, İstanbul, 2012)

Spatial Analysis for Smart City Approach: The Case of Beşiktaş-Etiler Neighborhood Anıl Çakir , Enver Murat Karababa , Furkan Talha Erdemir , Berfin Şenik , and Elif Kutay Karaçor

Abstract While the rapid and intense urbanization processes taking place at the global level are transforming the urban landscape at the same speed, they have led to new policy searches in the spatial context at the stage of solving the increasing needs/problems. In this context, one of the most important approaches that come to the fore is the smart city. This approach, which offers technological solutions to the increasing urban problems of metropolitan cities, also offers important spatial opportunities for the sustainability of the urban landscape. In this study, the current status and potentials of smart city applications at the neighborhood scale were determined by drawing a methodical framework (analysis-synthesis-suggestion) based on the smart city concept and containing urban landscape planning goals and objectives in the example of Etiler Neighborhood, which is a part of Beşiktaş district in Istanbul. In the proposal phase, some spatial suggestions were made at the neighborhood scale based on the situations determined in the analysis and synthesis phase. In this direction, smart city targets and strategies that will contribute to future urban landscape planning studies have been determined. Keywords Urban landscape planning


Smart cities
Neighborhood scale
Etiler

1 Introduction “Landscape is a mosaic in which the mix of local ecosystems or land uses is similarly repeated for kilometers” (Forman 1995, p. 13). According to the European Landscape Convention, “landscape is defined as the area that emerges as a result of the action and interaction of natural and/or human factors as perceived by humans” (Council of Europe 2000). Accordingly, as there are natural processes or ecosystems that create the landscape, it also consists of cultural components/processes A. Çakir
E. M. Karababa
F. T. Erdemir
B. Şenik
E. K. Karaçor (&) Department of Landscape Architecture, Faculty of Forestry, Düzce University, Konuralp Campus, 81620 Düzce, Turkey e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_4

41

42

A. Çakir et al.

brought about by agricultural, urban and historical processes. In this systemic structure, it is important to protect, develop and plan all components, interactions and the processes they result from. In this sense, the concept of landscape planning comes to the fore. Landscape planning is a concept that includes all the actions of developing and implementing strategies, policies and plans to create successful environments in both urban and rural environments for the benefit of present and future generations (Landscape Institute 2016). It also provides an approach that envisages alternative land use configurations (Ahern 2005), which is understood as an important factor for a sustainable planning process, and that includes goals and objectives related to the potential of the landscape, current and potential impacts on these potentials (Mander and Uuemaa 2015). While human habitat has changed fundamentally and profoundly since the end of the twentieth century (Steiner 2014), the twenty-first century is recognized as the century of the new urban world (Vasiljević et al. 2018). All of the urban processes are intertwined with the concept of landscape itself. Of course, these developments also include the transformation of the landscape and the natural/rural landscape leaves its place to the concept of urban landscape. Urban landscape is an important visual form that reflects the characteristics of a city (Lin and Chiu 2003). The spatial and temporal boundaries of the urban landscape are also defined by the presence of dense urban structure (Kruhlov 1999). In this respect, urban landscapes should be handled with a landscape planning approach in which the ecological, social and spatial needs of rapid and intense urbanization can be met as required by the process. The multidimensional developments that have taken place in urban areas in the last century and the unstoppable rise of cities reveal new conceptual frameworks in which different solutions are presented to overcome some emerging difficulties. It is critical to internalize the transformation processes of urban landscapes and to carry out effective planning studies in harmony with this process. One of the most important concepts emerging in the context of producing technological solutions to urban problems is the smart city (Caragliu et al. 2011). In its simplest definition, smart cities are places where information, digital and communication technologies are used to make traditional networks and services more flexible, effective and sustainable with the aim of developing activities for the benefit of city residents (European Commission 2019). Smart city applications basically produce solutions under three headings. These are physical environment (ecological sustainability, monitoring system, etc.), society (high-tech industry, knowledge economy, higher education, stakeholder, citizen, and community engagement etc.) and government (policies, transportation, public safety, health and social services, housing, water, energy and electricity, solid waste, etc.) (Gil-Garcia et al. 2015). In other words, the concept of smart city itself encourages policies based on making cities technologically smarter (Anguluri and Narayanan 2017). For example, the European Union supports the concept of smart city in order to reduce greenhouse gas emissions and ensure environmental sustainability in urban areas with the use of innovative technologies (Ahvenniemi et al. 2017). In EU Pilot Smart Cities, different smart city applications come to the fore in accordance with the needs of each city. In Amsterdam, Netherlands, different policies are applied to cities with

Spatial Analysis for Smart City Approach …

43

approaches such as ‘smart society’, Smart Work Centre, Smart Grid. Intelligent traffic data management, Energetic self-sufficiency plan in Barcelona (Spain); An online Solar Maps and CityGuide Digital Mapping System services are provided where residents can decide where the conditions are suitable for installing solar panels in Frankfurt (Germany) (Smart & Cooperation 2014). Thus, a process is defined in which all components and processes that create the urban landscape are blended with technology, social and spatial needs are met more quickly and effectively, and the welfare of the whole society is ensured starting from the individual. At the same time, an understanding that supports the sustainable use of space and holistic spatial solutions required by urban landscape planning is also spatialized. In Turkey, smart city applications are/has been implemented in many cities, especially in metropolitan cities. One of them is Istanbul, the city where metropolitanization first started. In the 2020–2024 Strategic Plan prepared by the Istanbul Metropolitan Municipality (IMM), it is stated that the metropolises focus on smart urbanization applications in an increasingly global competitive environment. In this context, it is emphasized that the results of the IMM’s current project should be developed with a smart city system where the endless data network is made workable with the development of innovation and technology infrastructure and all plans from infrastructure to development are produced technology-centered. In particular, the integration of the rapidly changing urban landscape with smart city components (smart living, smart environment, smart mobility, smart economy, smart governance, smart people) offers an important opportunity in terms of providing a more sustainable urban development in the context of urban landscape planning. According to the data of the Turkish Statistical Institute (TUIK) (2021), Istanbul, which has a population of approximately 15.4 million, has a population of approximately 15.4 million, depending on this population density (2895 people per km2), both in the whole city and in the context of transportation, open green spaces, residential areas covered by the urban landscape. The spatialization of appropriate smart city applications at the neighborhood scale has become a necessity. In this study, the possibilities of smart city applications are emphasized within the framework of urban landscape planning approach in the case of Etiler Neighborhood of Beşiktaş district in Istanbul. In this context, a methodological framework was drawn that includes the stages of “analysis-synthesis-suggestion”. In the analysis phase; The current situation of the neighborhood in the context of the smart city approach was determined by conducting 6 different analysis studies, namely area use analysis, floor height analysis, urban pattern analysis, green space analysis, transportation analysis and image analysis. In the synthesis phase, the potentials of the neighborhood in the context of smart city applications were determined by synthesizing 6 different analysis results. At the last stage, suggestions were made for smart city applications for Etiler Neighborhood, both for the whole neighborhood and for certain parts of the neighborhood. Thus, a framework that will provide a basis for the urban landscape planning studies to be carried out at the neighborhood scale on the basis of the smart city approach has been put forward.

A. Çakir et al.

44

2 Material and Method 2.1

Material

Etiler Neighborhood, which is the study area, is adjacent to neighborhoods such as Bebek, Arnavutköy, Boyacı Village in the Beşiktaş district of Istanbul and has an important commercial and social position. Etiler is the second mass housing settlement established after 1st Levent, which was opened for settlement in 1957. It is one of the most preferred places in Istanbul (Beşiktaş Municipality 2017). The total area of Etiler is 1.3 km2, and its population is 11,040 in 2020 according to TUIK (2021). Beşiktaş district constitutes the majority of the monthly average household income in the province of Istanbul, and Etiler Neighborhood provides a large part of this majority. As a result, Etiler Neighborhood is a socio-economically developed district of Istanbul (see Fig. 1). 1/1000 scaled Implementation Zoning Plan of 2010 covering the Etiler region of Beşiktaş district was used as the main base for all analyzes related to the area. AutoCAD 2021, ESRI ArcMap 10.7.1, Google Maps, Adobe Illustrator 2020, Adobe Photoshop 2020 were also used for analysis and visualizations.

2.2

Method

The study consists of analysis-synthesis and suggestion stages (see Fig. 2). Based on the 1/1000 scaled Implementation Zoning Plan of 2010, which is already layered

Fig. 1 Study area

Spatial Analysis for Smart City Approach …

45

Fig. 2 Flowchart of the method

in the acquisition of data during the analysis phase; filtering building units in Urban Pattern Analysis, marking the locations and routes of metro and bus stops in the area in Transportation I-II Analyzes, updating the floor heights expressed in the zoning plan in Floor Height Analysis and transferring them to the map with Google Maps data, zoning plan and Google Maps in Active-Passive Green space Analysis actively and passively decomposing the regions specified in. The data in the above-mentioned analyzes have been reached, together with the mapping. At the synthesis stage, the common denominators of these data were overlapped on a single sheet and the problems in the field were determined based on the 2020–2023 National Smart Cities Strategy and Action Plan, by examining both the area and the street. At the proposal stage, these problems were evaluated within the framework of the 2020–2023 National Smart Cities Strategy and Action Plan, and solutions were produced so that the public could feel the concept of Smart City with maximum efficiency.

3 Results 3.1

Land Use Analysis

In the study area, the settlement with single and multi-family structures in general, and the mixed use areas developed depending on the commercial use in the same residence, in addition to these family structures, were determined by the area use

A. Çakir et al.

46 Fig. 3 Land use analysis

analysis. A significant part of the study area consists of residential areas (11.17%). In terms of land use, residential areas are followed by mixed-use (4.48%), commercial (3.22%), parking (2.94%), green spaces (2.69%) and institutional (1.25%). As a result of the determination, it has been seen that mixed-use and residential areas are concentrated on Cengiz Topel Avenue, which is generally expressed as the main artery. At the same time, on Nispetiye Avenue, where the carrying capacity of the area is the highest, there is a dense commercial establishment and the presence of small-scale green spaces along the main arteries (see Fig. 3).

3.2

Floor Height Analysis

As a result of the floor height analysis carried out on the Etiler Neighborhood, it was determined that there is an average floor height (5–6 floors, 15.5–18.6 m) on Nispetiye and Cengiz Topel Avenues. According to the 10th article of the Bosphorus Law No. 2960 and dated 18.11.1983, in line with the statement “On the condition that the Base Area Floor Number is 15% and 5 floors (H = 15.50 m. altitude) do not exceed again in the affected zone”, it has been determined that the Etiler Neighborhood is in the affected area and the residences exceeding 5 floors are located in the area. In this context, it has been determined that the rate of 4 and 5-floor dwellings is higher than other dwellings, and the rate of 8, 9, and 10-floor dwellings is less than other dwellings. Finally, it was observed that the ratio of the buildings on 5 floors to the buildings in the whole area is 5% (see Fig. 4).

Spatial Analysis for Smart City Approach …

47

Fig. 4 Floor height analysis

3.3

Urban Pattern Analysis

Once the solid (buildings) and void (open spaces) rate of the urban pattern on the study area is examined, it is seen that the building density is spread over the whole area. At the same time, it has been determined that the density of buildings on Cengiz Topel and Nispetiye Avenues decreases as one goes to the northeast of the study area. As a matter of fact, when the numerical data obtained were examined, it was determined that the ratio of the solid rate to the entire study area was 19.71% (see Fig. 5).

A. Çakir et al.

48

Fig. 5 Urban pattern analysis

3.4

Green Space Analysis

Active green spaces define the areas where recreational activities are carried out, while passive green spaces refer to the open green spaces (medians, cemeteries, etc.) outside of these. According to the green space analysis on the study area, it is seen that active green spaces are more than passive green spaces (see Fig. 6). It is stated that the amount of active green space that should be provided per person in urban environments included in the Annex-2 Table of the Spatial Plans

Fig.6 Green space analysis

Spatial Analysis for Smart City Approach …

49

Construction Regulation (SPCR) dated 14.06.2014 and numbered 29,030 is 10 m2. According to TUIK (2021) data, when the population of Etiler Neighborhood in 2020 (11.040) is taken into account, it has been determined that the active green space per person is 2.5 m2. This shows that the study area cannot meet the open green space standard specified in SPCR and has the amount of open green space much below this standard.

3.5

Transportation Analysis

According to the 12th article of SPCR, “In the zoning plans; children’s playground, playground, open district sports field, family health center, kindergarten, kindergarten and primary school functions are approximately 500 m” walking distance standard. In this context, the 500-m access distance foreseen for the common areas was also taken as a basis in the transportation analysis of the study. According to the related analysis, it has been determined that the main arteries and lower arteries are not of sufficient size, and the areas for pedestrian use or cycle use are not in sufficient order and are in a piecemeal state. Likewise, the presence of metro, bus stations, and taxi stands in the area stands out from the results of the relevant determinations, but the fact that the bus routes are only on the main arteries, the metro station is in a far location on the area and the alternative transportation options are insufficient to show that the transportation systems for the region are not healthy. Likewise, the fact that pedestrian crossings, overpasses, and traffic lights are only on main arteries is one of the biggest factors supporting this problem (see Fig. 7).

Fig. 7 Transportation analysis

A. Çakir et al.

50

Fig. 8 Image analysis

3.6

Image Analysis

It is often forgotten that living in a city is actually living in an image (Lynch 1960). In line with the urban image analysis method, nodes formed by the combination of main and sub-axes, which create an edge effect on the area, landmarks that should be seen visually and architecturally, and districts with different characteristics were determined. As a result, different districts are formed on Cengiz Topel and Nispetiye Avenues, the lower arteries and its surroundings, the low number of landmarks in such a special district and the high edge effect in the area strikes (see Fig. 8).

4 Conclusion The rapid development of information technology can lead to urban development and different structural elements that reflect the concept of “smart city”, which increases the importance of not only land planning and layout, but also related landscape technology (Yang et al. 2021). In this context, sustainable urban development and smart cities are defined together with the characteristics of cities such as being more responsive to the needs of citizens, providing conditions that promote high quality of life and increasing competitiveness in an increasingly globalized environment (Angelidou et al. 2018). At the same time, it shows that the

Spatial Analysis for Smart City Approach …

51

concept of smart city has data-centered and multidisciplinary features due to its conceptual content (Yin et al. 2015). Therefore, the reason for different definitions and studies on smart cities is that the concept is multidimensional. In this study, the concept of smart city was discussed within the scope of urban landscape planning and evaluated within the scope of analysis-synthesis-suggestion phases. In the analysis and synthesis phase, it has been determined what Etiler Neighborhood needs in order to become a smart city at the city scale and what kind of requirements it already provides, and a proposal list has been prepared at the last stage. The focal point of the city can be divided into multiple points by creating smart green points that are homogeneously distributed on the scale of Etiler Neighborhood. These smart green spots can include landscape elements, recreational areas, solar systems, Wi-Fi points and electrical device charging services. Thus, people can feel the smart city life in every part of the neighborhood and can be directly involved in the smart city approach. By aiming to increase the capacity of the existing transportation infrastructure and ensure the safety and efficiency of highways, transportation to designated areas can be made easier and smarter. Real-time and dynamic intersection management systems can be applied to the main arteries in the neighborhood. By increasing the presence of bicycle and pedestrian paths in the neighborhood, it can reduce the vehicle traffic on the main arteries, and smart bus and cycle stand systems can be placed on these main arteries. Within these systems, it is possible to contribute to the integrity of the green space in the area by creating green space with solar systems, Wi-Fi services, and plants. Thanks to the traffic management centers, all existing systems, subsystems, and sensors in the traffic network (dynamic intersection control systems, smart attendance control systems, variable message signs, signalized intersections, CO2/temperature/Bluetooth sensors, etc.) can be remotely controlled only on main streets. Public transportation and pedestrian transportation, which are the main ones, can be distributed in proportion to the neighborhood scale, and new transportation solutions (bus routes, rail system lines, etc.) can be brought accordingly. It is aimed to ensure that the green banquettes, which come with smart bus stations, are insufficient to use green dots, to a sufficient level. Considering the rainfall in the Etiler Neighborhood, the hard ground density determined in the urban pattern analysis, rainwater can be recovered by replacing these hard floors with permeable smart ground systems. At the same time, with the integration of smart systems, it can calculate the amount of water falling to the ground and recovered, and use it to increase soft soils (green shoulders, soil, etc.) and to ensure sustainability. In a possible urban transformation process, according to the 10th article of the Bosphorus Law dated 18.11.1983 and numbered 2960, “In the affected area, the Base Area Floor Number is 15% and 5 floors (H = 15.50 m. altitude) is not exceeded. The border effect in the neighborhood can be eliminated by taking into account the phrase housing can be built”. Thus, the entire neighborhood can feel the effect of the landscape and the integrity of the use of the area in the neighborhood can be achieved by spreading the use of green space homogeneously throughout the neighborhood.

52

A. Çakir et al.

Although the amount of green space per person in the neighborhood is low, we can see partially green roof systems. By applying the green roof systems to suitable areas throughout the neighborhood, the amount of green space per person can be increased and studies on smart city systems can be made. In this context, solar systems for each green roof, water recycling and recovery systems that can provide recovery of used and falling rain water can be applied at the points where green roofs are located.

References J. Ahern, Theories, methods & strategies for sustainable landscape planning, in From Landscape Research to Landscape Planning: Aspects of Integration, Education and Application, vol. 12, ed. by B. Tress, G. Tres, G. Fry, P. Opdam (Springer, Heidelberg, 2005), pp. 119–131 H. Ahvenniemi, A. Huovila, I. Pinto-Seppä, M. Airaksinen, What are the differences between sustainable and smart cities? Cities 60, 234–245 (2017) M. Angelidou, A. Psaltoglou, N. Komninos, C. Kakderi, P. Tsarchopoulos, A. Panori, Enhancing sustainable urban development through smart city applications. J. Sci. Technol. Policy Manag. 9(2), 146–169 (2017) R. Anguluri, P. Narayanan, Role of green space in urban planning: outlook towards smart cities. Urban For. Urban Green. 25, 58–65 (2017) Beşiktaş Municipality, Etiler Mahallesi (Etiler Neighborhood) (2017), https://besiktas.bel.tr/sayfa/ 1659/Etiler-Mahallesi European Commission (2019), https://ec.europa.eu/info/eu-regional-and-urban-development/ topics/cities-and-urban-development/city-initiatives/smart-cities_en A. Caragliu, C. Del Bo, P. Nijkamp, Smart cities in Europe. J. Urban Technol. 18(2), 65–82 (2011) Council of Europe, The European Landscape Convention. Strasbourg (2000) R.T.T. Forman, Land mosaics. The Ecology of Landscapes and Regions (Cambridge University Press, Cambridge, 1995) J.R. Gil-Garcia, T.A. Pardo, T. Nam, What makes a city smart? Identifying core components and proposing an integrative and comprehensive conceptualization. Inf. Polity 20(1), 61–87 (2015) I. Kruhlov, The structure of the urban landscape. Univ. Ostraviensis Acta Fac. Rerum Naturalium GeographiaGeologia 181, 7 (1999) Landscape Institute, Landscape planning introduction (2016), https://www.landscapeinstitute.org/ technical-resource/landscape-planning-introduction/ H.T. Lin, M.L. Chiu, From urban landscape to information landscape: digital Tainan as an example. Autom. Constr. 12(5), 473–480 (2003) K. Lynch, The İmage of The City, vol. 11 (MIT Press, Cambridge, 1960) Ü. Mander, E. Uuemaa, Landscape planning. Encycl. Ecol. 532–544 (2015). https://doi.org/10. 1016/b978-0-12-409548-9.09478-1 EU-China Smart and Green City Cooperation, Comparative study of smart cities in Europe and China. Current Chinese Economic Report Series (Springer, 2014) F. Steiner, Urban Landscape Perspectives. Land 3(1), 342–350 (2014) TUIK (Turkish Statistical Institute) (2021). Nüfus ve Demografi (Population and Demographics) 2020, https://data.tuik.gov.tr/Kategori/GetKategori?p=Nufus-ve-Demografi-109 N. Vasiljević, B. Radić, S. Gavrilović, B. Šljukić, M. Medarević, R. Ristić, The concept of green infrastructure and urban landscape planning: a challenge for urban forestry planning in Belgrade, Serbia. iFor.-Biogeosci. For. 11(4), 491 (2018) W. Yang, X. Xi, L. Guo, Z. Chen, Y. Ma, Guangzhou digital city landscape planning based on spatial information from the perspective of smart city. Math. Probl. Eng. 2021, 11 p. (2021). Article ID 5572652 C. Yin, Z. Xiong, H. Chen, J. Wang, D. Cooper, B. David, A literature survey on smart cities. Sci. China Inf. Sci. 58(10), 1–18 (2015)

Utilization of the Visiting Jogja Mobile Application as a Provider of Information Regarding Limitations of Tourism Activities During the COVID-19 Pandemic in the Special Region of Yogyakarta Candra Triastutiningsih and Rini Rachmawati

Abstract COVID-19 pandemic has caused reduction of tourist arrival, and thus it has resulted in crisis of tourism sector in Special Region of Yogyakarta. The government has implemented health protocols in accordance to Decree of the Minister of Health No. HK.01.07/MENKES/382/2020 as an effort to revive the tourism sector while minimizing the spread of COVID-19. The information regarding the regulation can be conveyed by utilizing ICT, one of them is through Visiting Jogja Mobile Application. This study aimed to identify the provision of information on the Visiting Jogja Mobile Application regarding restrictions on tourism activities and identify the utilization of the application among people who carry out tourism activities during the COVID-19 pandemic. This study used qualitative research methods. Primary data were collected through in-depth interviews and online questionnaires. Secondary data were obtained from literature studies and browsing applications with smartphones. All of the data were analyzed by qualitative and quantitative descriptive analysis. The results show that the application provides information and helps implement 4 of the 8 health protocols through its features. Utilization of applications by tourists is still low both from system management and survey results. Therefore, branding is still needed about the application to the public. Keywords Information and communication technology application Tourism COVID-19








Visiting Jogja mobile

C. Triastutiningsih (&)
R. Rachmawati Department of Development Geography, Faculty of Geography, Universitas Gadjah Mada, Yogyakarta, Indonesia e-mail: [email protected] R. Rachmawati e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_5

53

54

C. Triastutiningsih and R. Rachmawati

1 Introduction Coronavirus Disease or the so called COVID-19 is a disease that emerged in December 2019 in China caused by the SARS-CoV-2 corona virus and has quickly become a pandemic on a global scale [1]. The transmission of COVID-19 can be through various modes, such as droplets, air and formites. The rapid spread of the virus has caused it to reach various regions in Indonesia, one of which is the Special Region of Yogyakarta (DIY) starting in March 2020. The COVID-19 pandemic in Special Region of Yogyakarta has resulted in a crisis in the tourism sector [2]. It occurs as people avoid crowds, including tourist destinations, in order not to get COVID-19. This condition has caused a decrease in the number of tourist visits. From March to April the number of visits decreased by 93% from 180,872 visits to only 5,270 visits. The tourism sector is an economical driving force in DIY. In fact, the tourism sector contribution to the economy in DIY is 55.37% [3]. The decrease in tourist arrivals causes a decrease in regional income. The decline in revenue from the tourism sector in 2021 reached 52.9% [4]. Some efforts have been conducted by the government and the public to prevent the spread of the SARS-CoV-2 corona virus. One of which is the establishment of a health protocol policy in accordance to Decree of the Minister of Health No. HK.01.07/MENKES/382/2020 concerning Health Protocols for Communities in Public Places and Facilities in the Context of Preventing COVID-19 in June 2020. The implementation of health protocols has been carried out so that the economy continues to grow through the tourism industry but still minimizes people exposed to COVID-19 [2]. The implementation of health protocols in tourism activities can occur if the provision of information related to tourism and restrictions on tourism activities is conveyed to the public. As a tourist area, DIY requires information technology that can support the tourism sector [5]. One of the implementations in ICT is its application. Mobile applications are the ones running on mobile devices that are easy to move, use and access anywhere and anytime [6]. The application can provide information more quickly and accurately through its features [7]. The Visiting Jogja Mobile Application is one of ICT implementations in the tourism sector in DIY. As one of the public service applications related to tourism, this application is expected to play a role in conveying public information related to the application of health protocols to tourism activities in DIY. The purpose of this study is therefore to analyze the provision of information on the Visiting Jogja Mobile Application regarding restrictions on tourism activities during the COVID-19 pandemic and analyze the use of the Visiting Jogja Mobile Application among people who carry out tourism activities during the COVID-19 pandemic. A similar study was also conducted by Rini Rachmawati et al. [5] with the title Smart Province Application “Jogja Istimewa”: Provision of Integrated Information and Its Utilization. This study aims to identify the provision of integrated information in the Smart Province Application “Jogja Istimewa” and analyze the optimal

Utilization of the Visiting Jogja Mobile Application …

55

use of the Smart Province Application “Jogja Istimewa”. This is different from this research which focuses on providing information and its uses by the community during the COVID-19 pandemic in tourism activities as one way to realize a smart city.

2 Literature Review 2.1

The Use of Information and Communication Technology in Tourism

ICT refers to all communication technologies that enabling users to manipulate information in a digital form such as the internet, software, and other media applications [8]. Internet is a form of ICT utilization rapidly growing as it is able to support rapid communication between one party and another without recognizing the boundaries of space and time [9]. One of the internet usage is in public services since it can assist to convey information. The application of ICT in the tourism sector is known as e-tourism. It is referred to as a way to build commercial relationships using the internet with the aim of offering tourism-related products. These products include flights, hotel reservations, car rental, and many other [10]. The ICT development and progress is one of the factors driving the advancement of the tourism industry [11]. The progress of tourism sector will have impact on the local area. The tourism sector has major contribution to an area since the presence of tourist sites can increase the number of visits and thus resulting in a source of income for the region [12]. The great ICT contribution in tourism has remain constrained by several things. One of the obstacles in the use of ICT is the limited capacity of human resources as managers of tourist sites, for example the expertise of managers in displaying tourism packages that are creatively and uniquely offered via websites so that they can attract potential tourists [11]. The high utilization of ICT is in line with the rapid changes in the industry but the dynamics of the use of ICT is influenced by several factors, one of which is human resources [13].

2.2

Tourism During the COVID-19 Pandemic

The government has implemented restrictions on mobility and recommendations not to travel and gather in large numbers, causing tourist visits to drastically drop and resulting in a crisis in the tourism sector [14]. By observing the conditions, the government implements health protocols so that the tourism industry can continue to operate but still minimize the spread of the virus that causes COVID-19. The health protocol policy in accordance to Decree of the Minister of Health

56

C. Triastutiningsih and R. Rachmawati

No. HK.01.07/MENKES/382/2020 concerning Health Protocols for Communities in Public Places and Facilities in the Context of Preventing COVID-19. Submission of information regarding health protocols in tourism activities is important to be conveyed to tourists. This can be supported by the use of ICT, one of which is in the form of applications. To prevent the spread of the virus during the COVID-19 pandemic, several regions have implemented the smart city concept to prepare ICT-based applications to overcome the pandemic, making it easier for people to carry out their daily activities [15].

2.3

Visiting Jogja Mobile Application

Smart city concept is an urban development concept that is closely related to Information and Communication Technology. It is the application of Information and Communication Technology in order to actualize good public services in order to create a comfortable, safe and strong city in economic competitiveness [16]. Public services in the smart city concept can be supported by the application. The application can provide information more quickly and accurately through its features [7]. Several cities in Indonesia have also developed the smart city concept using ICT-based applications, one of which is DIY. The existence of this application can help people in period of pandemic [15]. One of the ICT-based applications in DIY is the Mobile Visiting Jogja Application. Visiting Jogja Mobile Application is a DIY tourism information portal managed by DIY Tourism Office. It consists of several features facilitating the process of delivering tourism information in DIY. Some of the features available in this application are information regarding tourist destinations in the Special Region of Yogyakarta, the latest news and information on tourism in the Special Region of Yogyakarta, online ticket reservations, recommendations for tourist destinations, the nearest hotel and culinary, and directions to tourist destinations.

3 Methodology This study used qualitative research methods. Primary data were obtained from in-depth interviews with 4 informants by a sampling technique in the form of snowball sampling. Primary data was also obtained from the distribution of online questionnaires to 100 respondents whose sample was determined by using a sampling technique in the form of purposive sampling. Secondary data were obtained from literature studies from Decree of the Minister of Health No. HK.01.07/MENKES/382/2020, system management data from PT. Integra Innovation Indonesia and DIY Tourism Office, and also browsing the Visiting Jogja Mobile Application with a smartphone. Secondary data regarding the

Utilization of the Visiting Jogja Mobile Application …

57

distribution of the use of the Visiting Jogja Mobile Application, especially the ticket reservation feature, will be processed using ArcGIS first to obtain the spatial distribution in DIY. Data were analyzed based on triangulation technique. All of the data were analyzed by qualitative descriptive analysis and quantitative descriptive analysis. The analysis of the provision of information is based on the Decree of the Minister of Health No. HK.01.07/MENKES/382/2020 which contains 8 health protocol. Analysis of application utilization is carried out based on several things, namely: (1) application downloads; (2) knowledge of applications; (3) intensity of use and distribution of utilization locations. The system flow chart can be seen in Fig. 1.

4 Result and Discussion Result and Discussion 4.1

Provision of Information Related to Restrictions on Tourism Activities During the COVID-19 Pandemic

The analysis of the provision of information is based on the Decree of the Minister of Health No. HK.01.07/MENKES/382/2020 which contains 8 health protocols, namely: (1) limiting the number of visitors who enter; (2) setting back operating hours; (3) adjusting the distance when queuing by giving a marker on the floor at least 1 m; (4) optimizing open space for places sales/transactions in order to prevent crowds; (5) limiting the capacity of elevator passengers by labeling on the elevator floor; (6) setting a minimum distance of 1 m in elevators and stairs; (7) regulating the flow of visitors in tourist attraction areas; (8) using barriers/partitions in table or counter. The provision of information related to restrictions on tourism activities in this application is divided into several features, as follows:

Primary Data • Questionnaire • Indepthinterviews Data Collection

• Presented in graph, table • Create interview transcript Qualitative and quantitative descriptive analysis

Secondary Data • Literature review • System management data • Browse the application

Fig. 1 System flow charts

• Presented in graph, diagram, and table • Create map with ArcGIS

Utilization of Mobile Visiting Jogja Application

58

C. Triastutiningsih and R. Rachmawati

Information Features in Terms of Tourist Destinations in the Special Region of Yogyakarta. The information feature regarding tourist destinations in DIY contains general information, such as object descriptions, operating hours, user fees and facilities provided. This feature has a role to convey information related to health protocols, particularly social distancing, such as restrictions on tourism activities, facilities provided in the context of preventing COVID-19 and also setting back operating hours as it has been conducted by the Jogja Exotarium tourist destination manager. The Latest News and Information on Tourism in the Special Region of Yogyakarta Feature. The latest news and information feature of the Special Region of Yogyakarta provides information on the latest regulations related to the COVID-19 Pandemic, agency activities, seminars/webinars/talkshows regarding tourism, competitions and news of destinations, culinary, hotels to accommodation. Through this feature, prospective tourists can find out various rules that apply in DIY related to the COVID-19 Pandemic, such as information on the implementation of closing destinations to prevent the spread of COVID-19 and the importance one is provide the book guide called “Buku Pranata Anyar Plesiran Jogja” which the implementation of the the Decree of the Minister of Health No. HK.01.07/ MENKES/382/2020 in DIY. Online Ticket Reservation Feature. This feature playes a role in limiting the number of visitors’ arrival, adjusting the distance when queuing by placing a marker on the floor at least 1 m and optimizing open space for sales/transactions to prevent crowds. It provides information on the total capacity of tourist destinations and the number of visitors in the last 2 h at that destination. The combination of the two information is useful for monitoring the number of visits so as not to exceed the maximum capacity of only 50% during the COVID-19 pandemic for both managers and destination visitors. By implementing the ticket reservation feature, tourists do not necessarily queue to buy tickets, however they can buy it online and scan the QR Code as entering tourist destinations so that the application of a queue distance of 1 m can be applied in open spaces. The feature can minimize contact between tourists and managers and prevent crowds at once. The online ticket reservation is the most feature that provide and implement the health protocol which is 3 from 8 health protocol. Recommendations for Nearby Tourist Destinations, Hotels and Culinary Delights Feature. The recommendation feature for nearby tourist destinations, hotels and culinary delights can facilitate the potential tourists to look around the nearby destinations regarding the distance and travel time. It helps potential tourists in choosing tourist destinations according to new trends during the COVID-19 pandemic to choose open tourist destinations in a relatively close distance [17]. So, this feature can facilitate the implementation of tourism trends during the covid-19 pandemic even does not provide information regarding the 8 heatlh protocols.

Utilization of the Visiting Jogja Mobile Application …

59

Directions to Tourist Destinations Feature. This feature does not provide information regarding the 8 health protocols but provides information needed during the COVID-19 pandemic. The directions feature to tourist destinations is useful for tourists from outside the city to choose the best route to tourist destinations. By providing directions feature to tourist destinations, it will help tourists from outside area to find out the route so that they can estimate the post and prepare files to be checked when passing through it as a travel requirement from the government.

4.2

The Utilization of Visiting Jogja Mobile Application

Analysis of Visiting Jogja Mobile Application is carried out on the research by Rachmawati et al. [5] that uses some variables, namely application download performance, average of user utilization, respondent knowledge, and respondent utilization. Application Download. Total downloads from initial release to March 31, 2021 were 18,197. Based on Fig. 2, it can be seen that the number of downloads of the Visiting Jogja Mobile Application tended to decrease every month from the initial release until March 2021. The decline in the number of application downloads might be due to the COVID-19 pandemic in which it resulted in non optimal branding or advertising. Advertising is an effective way to increase the number of downloads of an application and popularity of an application is recognized when it has a high download rate [18]. The high number of downloads at the beginning of the release might be due to intense promotion and branding. In contrast to the current condition, the focuses is on the application maintenance process, and thus decreases the promotion and branding.

Number of downlod

The Application Knowledge. Based on the primary data collection outcome via online questionnaire can be seen that out of 100 respondents, there were only 34 people knowing about the Visiting Jogja Mobile Application while 66 other did not know about it. It’s means that the knowledge of Visiting Jogja Mobile Applications is still low because the comparison does not reach 50%. The origin area distribution of respondent knowing the Visiting Jogja Mobile Application was 20 people from DIY and 14 people from outside DIY. The majority of respondents were aware of

6000 4000 2000 0 June

July

Aug

Sept

Oct

Nov

Dec

Jan

Fig. 2 The number of downloads for the visiting Jogja mobile application

Feb

Mar

60

C. Triastutiningsih and R. Rachmawati

the Visiting Jogja Mobile Application from DIY because the socialization carried out was indeed more in DIY than outside DIY. Based on questionnaire it can be seen that tourist visits were dominated more by non DIY tourists than insider itself. The number of tourists from outside of DIY was not accompanied by the branding delivered. The lack of branding and promotion to the community resulted in not many people knowing about the Visiting Jogja Mobile Application, especially from outside visitors of DIY. Socialization using social media can be done to increase public awareness and knowledge of the application so that the application can be more familiar and effective [15]. Application Utilization Intensity. Based on the questionnaire show that out of 100 respondents, 15% of them used the application and 85% did not use the Visiting Jogja Mobile Application. Total of 19 respondents were aware of the Visiting Jogja Mobile Application but did not use it due to several reasons, such as feeling that they did not need the application and problems on their smartphone. The respondents found out that they did not need it as there were many alternative choices of tourist destinations. Other respondents did not use it since there was other option such as Google to find references and directions to tourist destination. The application is not used for several reasons, such as not needing the application and having other applications that are more helpful [19]. The mostly used feature by respondents was the information regarding tourist destinations in the Special Region of Yogyakarta feature, while the one least used was the directions feature to tourist destinations and ticket reservations feature. Ticket reservations feature utilization was only 0.4% for all visits to tourist sites in Special Region of Yogyakarta as and 2–3% for visits to Breksi Cliffs. The Utilization Locations Distribution. The ticket reservation feature utilizations between July 2020 and March 2020 were 2660 ticket reservations. It was carried out at various tourist destinations in Special Region of Yogyakarta. Information about the number of ticket reservation feature utilizations are divided into 3 classifications, namely low, medium and high. This classification made by ArcGIS application with natural breaks classification. The number of 33 to 206 is in the low classification, the number of 207 to 783 is in the medium classification, and the number of 784 to 1158 is in high classification. Bantul Regency is the regency with the highest number of visits, which is 1,158 visits by using ticket reservations. Sleman Regency and Gunungkidul Regency are in the middle classification with 783 and 480 visits, respectively. Meanwhile, Kulon Progo Regency and Yogyakarta City are in the low classification, which is only 206 and 33 tourist visits by utilizing the ticket reservation feature. This condition is influenced by the number of tourist destinations in Bantul Regency that provide ticket reservation features, as many as 29 tourist destinations while the City of Yogyakarta only has 5 tourist destinations that provide ticket reservation features. The implementation of ICT-based public services can vary in each region. This difference is influenced by human resources and capital owned by each region [20].

Utilization of the Visiting Jogja Mobile Application …

61

Table 1 Table of the number of application users and tourist visits Month

Total users of the application

Number of tourist visits

June 2020 July 2020 August 2020 September 2020

51 4876 9773 12,289

4991 20,352 40,731 43,446

5 Implementation of Visiting Jogja Mobile Application From 15 respondents who used the Visiting Jogja Mobile Application claimed to have benefited from the implementation of the application. These benefits come from the information obtained more quickly. The existence of an application during a pandemic is very important because it is very useful for the community to reduce movement, be fast, effective, and efficient in accessing public services [15]. Respondents were able to obtain information from application, for example knowing travel arrangements during a pandemic, knowing information on every tourist location in DIY, and purchasing tickets made easy. The existence of these benefits makes it easier for tourists to visit tourist destination in DIY during the COVID-19 pandemic. This is in line with the data on the number of tourist visits in DIY which also increased along with the total users of the applications at the same time. This shows that the application is able to attract tourists to visit DIY (Table 1).

6 Conclusion As an implementation of ICT, Visiting Jogja Mobile Application can provide information accordance to Decree of the Minister of Health No. HK.01.07/ MENKES/382/2020, namely limiting the number of visitors’ arrival, rearranging operating hours, adjusting the distance when queuing by giving a marker on the floor at least 1 m and optimizing open space for sales/transactions to prevent crowds. Beside that, this application can provide information that needed during COVID-19 pandemic. The provision of this information has remained not optimal. Therefore, it is necessary to increase the awareness of tourist destination managers to take advantage of the features available in the application. The Visiting Jogja Mobile Application utilization based on the system and the survey results using questionnaire was still low. It is shown by the number of downloads which continues to decline, the use of the ticket reservation feature was only 0.4% and the public knowing and using it was only 44% or 15 respondents. This low knowledge might be due to the lack of promotion and branding to the community. More intensive branding and promotion are therefore needed to the community both inside and outside Special Region of Yogyakarta to increase the utilization of the Visiting Jogja Mobile Application.

62

C. Triastutiningsih and R. Rachmawati

Acknowledgements This research was a part of Rekognisi Tugas Akhir (RTA) 2021 from Universitas Gadjah Mada, Indonesia led by Dr. Rini Rachmawati, S.Si., M.T. Therefore, the authors thank Universitas Gadjah Mada who has supported all of this research activities and their funding.

References 1. B. Marroquín, V. Vine, R. Morgan, Mental health during the COVID-19 pandemic: effects of stay-at-home policies, social distancing behavior, and social resources. Psychiatry Res. 293, 1–9 (2020). https://doi.org/10.1016/j.psychres.2020.113419 2. M. Prayudi, The Impact of COVID-19 on tourism in the special region of Yogyakarta. J. Archipelago 3, 14–20 (2020) 3. B. Listiyo, Wujudkan Jogja Tourism Goes to Digital Ecosystem, Melalui Sinergi Pengembangan Pariwisata DIY, https://www.sonora.id/read/422358156/wujudkan-jogjatourism-goes-to-digital-ecosystem-melalui-sinergi-pengembangan-pariwisata-diy?page=all. Accessed 24 Oct 2020 4. DIY Tourism Office, Tourism Statistics 2020. DIY Tourism Office, Yogyakarta (2021) 5. R. Rachmawati, E. Ramadhan, A. Rohmah, Smart province application “Jogja Istimewa”: provision of integrated information and its utilization. Indon. Geogr. Mag. 32, 14–23 (2018). https://doi.org/10.22146/mgi.31662 6. R. Islam, R. Islam, T.A. Mazumder, Mobile Application and Its Global Impact. Int. J. Eng. Technol. IJET-IJENS 10, 104–111 (2010) 7. R. Rachmawati, P. Rachmadani, V.N. Anifa, F. Lutfiana, Various ICT-based applications and their uses to support smart city implementation in the regency of Blora, in The 1st Geosciences and Environmental Sciences Symposium (ICST 2020). E3S Web of Conferences, Yogyakarta (2020). https://doi.org/10.1051/e3sconf/202020007004 8. Information and Communication Technologies (ICT), http://aims.fao.org/information-andcommunication-technologies-ict 9. S. Enjelina, E. Insannudin, HALOAPP interactive web-based application design for android and iOS. Web Based Appl. J. 10(2) (2016). https://doi.org/10.13140/RG.2.2.15774.95044 10. I. Condratov, E-tourism: concept and evolution. Ecoforum 2, 58–61 (2013) 11. Suleman, Utilization of ICT in Tourism Promotion in Indonesia. J. Comput. Inf. Acad. Bina Sarana Inform. Yogyakarta 1(1) (2013) 12. R. Rachmawati, E. Ramadhan, A. Rohmah, Smart province application “Jogja Istimewa” in support of tourism activities, in Planoearth National Seminar Proceedings: Implementasi New Urban Agenda Through Tourism Development Based on Local Culture and Community Empowerment (Universitas Muhammadiyah Mataram, Mataram, 2017), pp. 500–509 13. U. Choirunnisa, R. Rachmawati, ICT usages in Yogyakarta’s tourist villages development. J. Indon. Tour. Dev. Stud. 8(2), 112–118 (2020). https://doi.org/10.21776/ub.jitode.2020.008. 02.08 14. A. Hasibuan, O.K. Sulaiman, Smart city, smart city concept as an alternative for solving regency/city urban problems, in big cities of North Sumatra Province. Tech. Main Bull. 14, 127–135 (2019) 15. R. Rachmawati, A.D. Sari, H.A.R. Sukawan, I.M.A. Widhyastana, R.A. Ghiffari, The use of ICT-based applications to support the implementation of smart cities during the COVID-19 pandemic in Indonesia. Infrastructures 6(9), 119 (2021). https://doi.org/10.3390/ infrastructures6090119 16. M.G. Gunagama, Y.R. Naurah, A.E. Prabono, Post-pandemic tourism: important lessons and prospects for development. J. Urban Residential Archit. 5(2), 56–68 (2020). https://doi.org/ 10.33096/losari.v5i2.76

Utilization of the Visiting Jogja Mobile Application …

63

17. R.I. Kemenparekraf, New normal travel trends when the pandemic still hits, https:// pedulicovid19.kemenparekraf.go.id/tren-wisata-new-normal-saat-pandemi-masih-melanda/. Accessed 3 June 2021 18. T. Wahyono, H.L.H.S. Warnas, B.S. Wijaya, A. Fahri, Sasmoko, T. Matsuo, Building a popular mobile application by utilizing user feedback, in 2017 International Conference on Innovative and Creative Information Technology (ICITech) (2017), pp. 1–6. https://doi.org/ 10.1109/INNOCIT.2017.8319136 19. C. Koljonen, Marketing Plan for a Mobile Application (Oulu University of Applied Sciences, Finlandia, 2016) 20. R. Rachmawati, Pengembangan Perkotaan dalam Era Teknologi Informasi dan Komunikasi (UGM Press, Yogyakarta, 2014)

Intelligent Competitiveness of Logistics Companies Based on Benchmarking Approach Mohamed Achraf Laissaoui, Ouail El imrani , and Aziz Babounia

Abstract The practice of benchmarking logistics performance is taken seriously by contemporary organizations because of all the advantages it presents knowing that they have a very big impact on international trade. For example, the benchmarking process helps minimize the gap between an organization’s vision and mission. At a fundamental level, each organization has goals to achieve. This requires that the organization present certain levels of performance in order to achieve these organizational objectives. However, it often happens that the organization presents a level of performance which is not up to this standardized level of performance. Benchmarking helps the organization to understand the level of performance it gives in relation to the performance standard it has set to achieve these organizational objectives. Thus, if the achievement of the organizational objective is the vision, the mission consists in taking measures to achieve this objective by making performance equivalent to the performance standard. Benchmarking therefore makes it possible to minimize the gap between vision and mission. Benchmarking business performance also helps reduce the cost of service in the long run. Indeed, performance benchmarking contributes to the process of quantitative measurement of the organization’s performance. Unnecessary service overheads can be eliminated. In addition, as benchmarking contributes to improving the quality of service, it also allows an organization to continuously improve the quality of its performance, by developing the trust of customers in the organization as well as in chain stakeholders value of the organization. This paper should focus on the basics of benchmarking in the logistics industry, particularly in Morocco, by presenting a brief overview of the concept of Benchmarking in the field of logistics. It is followed by an examination of the advantages of benchmarking in logistics for an organization and an elaboration of the logistics sector in Morocco which will include a discussion on its main players, competition and key statistics. M. A. Laissaoui
A. Babounia Ibn Tofail University, Kenitra, Morocco O. El imrani (&) Abdelmalek Essaadi University, Tetouan, Morocco e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_6

65

66

Keywords Smart competitiveness Benchmarking

M. A. Laissaoui et al.


Performance
Optimization
Trade


1 Introduction In any economy, companies have a great interest in taking a close interest in this tool (Benchmark) which should be an integral part of the management of organizations, because due to globalization, customers have increased, but they also offers, for large as well as for small and medium enterprises. It is therefore necessary to adapt it to the contexts of the company, taking into account the internal constraints that are specific to it (cultural, political, and social). It is also important to know that SMEs, due to their small size, do not really have the opportunity to benefit from internal benchmarking. It is therefore inevitable that they compare themselves with the outside to find new ideas, improve their practices and thus increase their market share. This research aims to understand the impact of Benchmarking on the logistics sector and how it contributes to improving the competitiveness of an organization. Here are the objectives which have been developed in correspondence with the aim of the research: – Understand the current state of the practice of benchmarking in the logistics sector in Morocco. – Identify the main actors of the logistics sector in Morocco and understand if they will be ready to adopt it. – Understand the organizational changes that will be necessary to make them suitable for benchmarking practices in the logistics sector. The expectations hoped for by this research are very promising for all logistics activities, given that this sector of logistics providers constitutes a motor for the development of other activities such as: port logistics and terminal development, storage infrastructure, road infrastructure, etc.

2 Need/Importance of Benchmarking in the Competitiveness of Companies The benchmarking process has become an essential tool for modern organizations in terms of strengthening their market position and their competitiveness. Basically, benchmarking has become the backbone of modern organizations because of the set of benefits it generates (Krishnamoorthy and D’Lima 2014). Below is a list of some examples of the overwhelming importance of benchmarking in logistics.

Intelligent Competitiveness of Logistics Companies …

67

The notion of Benchmarking is better known through the expression “best practices”. Benchmarking (from the English benchmark means benchmark, reference or standard), this translation reduces the method to almost nothing, the word is translated into French by standard, or calibration, thus making it possible to measure oneself against others. It is also said parangonnage in Franco-Canadian, illustrates this practice of methodical adjustment to one’s fellow man. Whatever name you choose, it is about spotting a stallion or a paragon, that is, a model with which to compare, in order to close the performance gap. Benchmarking touches on management and management practices, that is, the work that staff do to run the business. When benchmarking is mainly aimed at establishing performance indicators, it is referred to as “benchmarking” and “markup”. Longbottom (2000) presents three main categories of benchmarking which are internal benchmarking, “competitive benchmarking” and “generic benchmarking” which consists in comparing the practices of the company evaluated with those used by world-class companies, which these are called best practices. Regardless of the form of benchmarking used, it is a strategic tool that allows the company to identify possible sources of improvement to increase its performance, its degree of competitiveness (Haughton et al. 1999) and improve the learning of employees who see what is being done elsewhere, thus leading them to better understand why they sometimes have to change their ways of doing things (Elmuti and Kathawala 1997). In addition, benchmarking is seen as a way of identifying new objectives to be achieved by allowing the company to compare its own business practices with those used by the best (Voss et al. 1997). The definition of benchmarking has been the subject of the interest of a large number of economists, managers and professionals in the field. Each of them has proposed a definition which can either be complementary or provide a new aspect to be taken into account. Among the many definitions of Benchmarking we can distinguish: - Operational definition According to Robert C. CAMP, the initiator of the approach who considers that “Benchmarking is the search for the most efficient methods for a given activity, making it possible to ensure superiority”. The literature supports the use of Benchmarking as a performance vector. Its use responds to the need for improved profitability, efficiency and process quality implied by the rapidly changing environment (Haugton et al. 1999). However, we notice that most of the studies are based on illustrations from business cases and there are currently very few empirical verifications of its real effect on the performance defined as well in an operational way, at the level of the different dimensions of the organization, or more globally by financial indicators (El Imrani et al. 2021). According to David Kearns, former general manager of Xerox Corporation. “Benchmarking is a continuous process of evaluating products, services and methods against those of the most serious competitors or partners or organizations recognized as leaders or leaders”.

68

M. A. Laissaoui et al.

3 Develop a Comptetitive Advantage Increasing competitiveness is often seen as an important factor for all contemporary organizations. In the current scenario, meeting and exceeding quality standards is essential. Thus, benchmarking responds appropriately to this organizational need (Knipe 2002). According to (Bagchi 1997), the comparative analysis of logistics is particularly important to improve the competitiveness of the organization. Through the performance benchmarking process, organizations are able to identify “best-in-class” performance standards against which they can compare their own organizational performance and then the strategies to achieve that standardized level of performance across the board (Pekuri et al. 2011). their product delivery process. Ultimately, this helps to strengthen their competitive advantage over time (Bagchi 1997). One of the characteristics of the goods transport and logistics sector in Morocco is that it is dominated by multiple actors in the form of transport operators, maritime operators, international logistics agents, customers, suppliers (Khaddam et al. 2020). Infrastructure, support sectors. Among the other actors who play an important role in the smooth running of operations in the logistics sector in Morocco are technology centers, the association of exporters, the association of shipping agents and logistics development agencies (Akoudad and Jawab 2018). Based on the number of employees and annual turnover, the logistics sector is dominated by companies with foreign capital such as UPS, Rhenus, IPSEN and DB SCHENKER. The subsidiaries of companies with foreign capital such as DachserMaroc, GEODIS Wilson Maroc and Bolloré Africa Logistics Maroc also dominate in this area. All these companies have their headquarters in Europe.

4 Integration of Technological Model The comparative study between the models of the integration of technology and the out of stock using the traceability presents an inventory of the traceability where taking example between the United States and the European community who have chosen to secure their drug circuit (for example) by introducing the concept of serialization, which facilitated full traceability of the latter. To locate current trends in the research, the next section presents a results and analysis of mathematical models of supply chain design in order to highlight important problems surrounding supply chain design related to comparative studies in logistics (Kassou et al. 2021). Benchmarking within companies is also linked to the process of improving an organization’s long-term business prospect (Jagtap et al. 2021)s. The goal of benchmarking is to improve the company’s own processes by drawing inspiration from formulas that work in the sector or for a particular area, this is possible because continuous performance evaluation and improvement helps the organization to satisfy its customers (Bourekkadil et al. 2021). This in turn increases customer loyalty and strengthens relationships with them. Ultimately, the organization’s business prospects improve.

Intelligent Competitiveness of Logistics Companies …

69

5 Competition of Logistics Companies in Morocco: Main Players, Competition and Key Statistics The Moroccan economy shows steady annual growth with an average growth rate of 4% over the past ten years. The country’s logistics sector has grown equally steadily over the years (El Imrani 2021). The main reason for this accelerated growth of Morocco’s logistics sector is mainly due to significant investments in the physical structure of the country (El Imrani 2021). This includes the creation of 1,800 km of new expressways, the expansion of the rail network and the development of logistics and industrial zones throughout the country (Rensma and Hamoumi 2018). However, the major growth of Morocco’s logistics sector took place between 2010 and 2015, in particular due to the massive growth rate of public development projects. In 2010, Morocco only had a few dozen hectares of modern logistics areas. In 2015, this area increased by 550 hectares around areas such as Casablanca and Tangier as well as other regions that have hosted integrated industrial platforms (PII). This IIP project was implemented by Moroccan companies such as MedZ, SNTF and ONCL in order to develop logistics sites in the country (Akoudad and Jawab 2018). Morocco’s logistics sector is an important pillar of the country and contributes significantly to the country’s GDP. In 2011, Morocco’s logistics sector accounted for US $ 9.07 billion. This represents 5.8% of total added value and 10.3% of tertiary activities (Akoudad and Jawab 2018). In 2015, the sector’s contribution to Morocco’s economy was around 6% (Aparicio, 2015). In 2018, the contribution of Morocco’s logistics sector to the country’s economy was 4% of GDP. The sector provides an average rate of 4.7% of jobs to young Moroccans each year. One of the characteristics of Morocco’s freight transport and logistics sector is that it is dominated by multiple players in the form of transport operators, shipping operators, international logistics agents, customers, suppliers. infrastructure, support sectors. Other actors that play an important role in the smooth running of operations in Morocco’s logistics sector include technology centers, the association of exporters, the association of shipping agents and logistics development agencies (Akoudad and Jawab 2018, Ennima et al. 2021). Based on the number of employees and annual turnover, the logistics industry is dominated by foreign-invested companies like UPS, Rhenus, IPSEN and DB SCHENKER. Subsidiaries of foreign-invested companies such as DachserMaroc, GEODIS Wilson Maroc and Bolloré Africa Logistics Maroc also dominate in this area. All these companies have their headquarters in Europe. Morocco’s national logistics companies, excluding the public body SNTL, are comparatively smaller than those of these foreign companies. Nevertheless, the national and foreign actors who dominate this logistics sector in Morocco have goods transport and warehousing activities (Rensma and Hamoumi 2018). In order to further strengthen the logistics sector as well as that of the country’s economic infrastructure, various sectoral strategies have been implemented for 2020. These include the Road Plan 2035, the National Port Strategy 2030, the Logistics Strategy

70

M. A. Laissaoui et al.

National 2030 and the 2035 Airports Development Master Plan. The 2035 Road Plan provides for the construction of 1,600 km of interurban motorways and rural roads over 30,000 km. As part of this plan, the rail network for passengers and goods across the country will also be extended. At the same time, the first high-speed TGV line with a length of 1,500 km would also be built. In another plan, the National Port Strategy 2030, five new ports will be built, four new ports will be enlarged and five ports will be integrated into the city. As part of the master plan for the development of the airport in 2035, the position of Casablanca airport will be strengthened and transformed into a regional center. In addition, the capacity of all airports in Morocco will be increased so that they can accommodate from 24 million passengers to 75 million passengers by 2035. Again, the national logistics strategy 2030 will have a great influence in the future of international trade, efforts will be made to reducing logistics costs and CO2 emissions, on the one hand, and creating 36,000 jobs and contributing even more to national GDP, on the other hand (Rensma and Hamoumi 2018).

6 Conclusion By focusing on the examination of benchmarking practices implemented by the logistics industry in Morocco, this paper analyzes the impact of benchmark practices on the competitiveness of companies. Using the analysis of the existing situation, we were able to quantify the close relationship between the improvement and the good use of benchmark practices within companies as well as organizational performance which will directly impact the level of competitiveness of the companies.

References S. Nagarajan, HR challenges faced by road transportation service providers in Tamilnadu, in International Conference on Contemporary Challenges in Management, Technology & Social Sciences, at Atmgimt, Lucknow (2014), pp. 3–4 S. Bourekkadil, K. Slimani, O. El Imrani, A. Babounia, S. Khoulji, A. Aboulhassane, Intelligent technology solutions at the service of energy consumption management, in E3S Web of Conferences, vol. 234 (2021), p. 00108. https://doi.org/10.1051/e3sconf/2021234 00108 S. Ennima et al., Innovation of the photovoltaic system, general and maximum power point tracking device for smart cities. J. Theor. Appl. Inf. Technol. 99(11), 2712–2721 (2021) A. Pekuri, H. Haapasalo, M. Herrala, Productivity and performance management - Managerial practices in the construction industry. Prod. Perform. Manag. – Manager. Pract. Constr. Ind. 1 (1) (2011) O. El Imrani, Study to reduce the costs of international trade operations through container traffic in a smart port, in Innovations in Smart Cities Applications Volume 4. SCA 2020, ed. by A.M. Ben, K.İ. Rakıp, D. Santos, O. Sergeyeva, A.A. Boudhir. Lecture Notes in Networks and Systems, vol. 183 (Springer, Cham, 2021), pp. 477–488. https://doi.org/10.1007/978-3-03066840-2_36

Intelligent Competitiveness of Logistics Companies …

71

O. El Imrani, M. Ben Messaoud Layti, S. Bourekkadi, A. Boulaksili, I. Kabbassi, Optimization of the international trade activities in the period of COVID-19 by proposing an algorithm, in Artificial Intelligence for COVID-19, ed. by D. Oliva, S.A. Hassan, A. Mohamed. Studies in Systems, Decision and Control, vol. 358 (Springer, Cham, 2021), pp. 187–193. https://doi.org/ 10.1007/978-3-030-69744-0_11 A.A. Khaddam, H.J. Irtaimeh, B.S. Bader, The effect of supply chain management on competitive advantage: The mediating role of information technology. Uncertain Supply Chain Manag. 8, 547–562 (2020) M. Kassou et al., Digital transformation in flow planning: the case of container terminals at a smart port. J. Theor. Appl. Inf. Technol. 99(9), 1966–1976 (2021) S.T. Jagtap et al., A framework for secure healthcare system using blockchain and smart contracts, in Proceedings of the Second International Conference on Electronics and Sustainable Communication Systems (ICESC-2021), IEEE Xplore Part Number: CFP21V66-ART; ISBN: 978-1-6654-2867-5 (2021) Krishnamoorthy, C. D’Lima, Benchmarking as a measure of competitiveness. Int. J. Process Manag. Benchmarking 4(3), 342–359 (2014) D. Knipe, M. Lee, The quality of teaching and learning via videoconferencing. Br. J. Educ. Technol. 33(3), 301–311 (2002) D. Longbottom, Benchmarking in the UK: an empirical study of practitioners and academics. Benchmarking Int. J. 7(2), 98–117 (2000) O. El Imrani, Prediction of international trade evolution through the support vector machine: The case of a port complex, in ACM International Conference Proceeding Series, 2021, 3456587 (2021) D. Elmuti, Y. Kathawala, An overview of benchmarking process: a tool for continuous improvement and competitive advantage. Benchmarking Qual. Manag. Technol. 4(4), 229–243 (1997) C. Voss, K. Blackmon, P. Hanson, B. Oak, The competitiveness of Europeanmanufacturing – a four country study. Bus. Strategy Rev. 6(1), 1–2 (12) (1997) M. Haughton, W. Grenoble, E. Thomchick, R. Young, The role of benchmarking in the performance of the import process. Int J. Phys. Distrib. Logist. Manag. 29(9), 511–568 (1999) P.K. Bagchi, A proper mentality for benchmarking. Logist. Inf. Manag. 10(1), 28–39 (1997) K. Akoudad, F. Jawab, Difficultés De Transport International De Marchandises: Proposition D’un Cadre D’étude: Cas Du Maroc, Eur. Sci. J. 14(14) (2018) M. Rensma, S. Hamoumi, Business opportunities Dutch companies in Transport & Logistics sector Morocco, rapport commandé par le ministère néerlandais des affaires étrangères (2018)

An Integrated Human-AI Framework Towards Organizational Agility and Sustainable Performance Mohamed Amine Marhraoui, Mohammed Abdou Janati Idrissi, and Abdellah El Manouar

Abstract Companies are facing important challenges related to markets’ internationalization, regulatory restrictions and fierce competition especially during the COVID19 crisis. They should embrace change and be agile in order to prevent risks and seize opportunities quickly and efficiently. In this article, we examine how artificial intelligence (AI) can help companies to enhance their organizational agility. Based on two systematic literature reviews on the subject, we identified the weaknesses on relying only on a digital enabler or human resource (HR) practices. Thus, we propose a Framework integrating artificial intelligence and human practices in order to help companies in their efforts towards agility. This latter allows companies to adapt to new regulations in the society, to customers’ expectations and to environmental changes. Ultimately, agile companies can ensure a sustainable performance.







Keywords Artificial intelligence (AI) Organizational agility Sensing Seizing Decision-making Human practices Sustainable performance













1 Introduction The pace of environmental change is increasing exponentially at the economic, social, environmental levels especially during the COVID19 context. Moreover, the world is generating data massively through internet and all the connected sensors. This 4th digital revolution presents challenges for companies in order to cope with change. Companies should then be agile on the organizational level. AI, which M. A. Marhraoui (&)
M. A. Janati Idrissi
A. El Manouar ENSIAS Engineering School, Mohammed V University in Rabat, Rabat, Morocco e-mail: [email protected] M. A. Janati Idrissi e-mail: [email protected] A. El Manouar e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_7

73

74

M. A. Marhraoui et al.

encompasses technologies and models allowing machines to think and behave like humans, can help companies to be more agile. In this work, we propose based on a socio-technical approach a Framework combining HR practices and AI technologies allowing the firm to sense its environment, to make adequate decisions and to act effectively. Indeed, our socio-technical approach is based on social and technical components as interactive and additive to produce outcomes [1]. Our technical component is related to the broad range of artificial intelligence technologies which interact with the social component namely human resources practices in order to increase the desired outcome which is the improvement of the firm’s agility level. This latter will help the company to boost its sustainable performance [2, 3]. The present paper is structured as follow. The second section is related to a review of the literature presenting artificial intelligence and organizational agility concepts and to two systematic literature reviews presenting the impact of artificial intelligence and human practices on agility. After describing the research gaps and the shortcomings of considering only one enabler, the research proposes a framework combining the artificial intelligence capabilities with human resources practices allowing employees to be more flexible and to design/adopt adequate AI technologies for sensing and seizing risks/opportunities.

2 Review of Literature 2.1

Artificial intelligence

The artificial intelligence subject has started early in the 50s. Indeed, (Turing, 1950) defined intelligence as the ability to achieve human cognitive tasks [4]. A machine is considered intelligent if a human interrogator cannot distinguish whether there is a human or a computer in the other end [4]. In addition, (Shannon, 1950) has proposed an intelligent machine able to play chess [5]. 1959 has marked the beginning of machine learning thanks to its pioneer Arthur Samuel who conceived a checkers program [6]. Since 1980, almost all big corporates possessed an AI department relying on decision networks and expert systems for optimal decision making [7, 8]. Progress in AI research witnessed a period a decline called “AI winter” from 1974 to 1980 due to several reports criticizing the developments in the AI field [9]. This latter experienced a second slowdown from 1987 to 1993 due to a lack of perception and funding from governments and public agencies [10]. AI research has regained its importance after the Robot Polly has navigated using vision in 1993 [11], and Deep Blue system has defeated the world chess champion [12]. Since 2011, some emerging practices in AI have emerged, especially those related to machine learning, deep leaning, Natural language processing (NLP), Robotic process Automation (RPA) and automated/assisted/augmented/autonomous intelligence [13].

An Integrated Human-AI Framework ...

75

Fig. 1 The Hype cycle for AI [10]

Figure 1 below summarizes the visibility of AI subject which complies with the Gartner Hype cycles [10]. It describes the maturity and adoption of AI and its applications, and how AI is potentially relevant to solving real business problems. Through this historical review, Artificial intelligence is considered as a scientific approach aiming to construct artificial models by identifying phenomenon and operating in the environment by acquiring the adequate competence [14]. In addition, artificial intelligence can be defined as the ability of a computer to do things normally associated with humans or perform tasks at which people are better [15]. Nevertheless, AI can also be associated with thinking or acting based on an ideal rationality through intelligent agents which are expected to maximize goal given the available knowledge of the environment [16]. Based on whether AI thinks or acts like a human or rationally, there are four possible goals of AI (Table 1 below) [8]. The areas of artificial intelligence research include knowledge presentations and logic, expert systems, natural language, robotics, computer vision, neural network and machine learning. Their applications can be grouped into three categories: cognitive science applications, robotics applications and Natural interface applications[17] (Fig. 2 below).

2.2

Organizational Agility

Organizational agility is studied in the literature from a dynamic capability perspective which allows the firm to address uncertainty and deal with environmental

76

M. A. Marhraoui et al.

Table 1 Four possible goals of AI [8] Reasoning-based Behavior-based

Human-based

Ideal rationality

Systems that think like humans Systems that act like humans

Systems that think rationally Systems that act rationally

Fig. 2 Main applications of artificial intelligence

changes [18]. It is the firm capability to deal with unexpected changes in the business environment and to develop innovative responses which exploit changes as an opportunity of growth [19, 20]. There are three main components of organizational agility [21]. The first one is the sensing capability, which is the ability of the firm to scan its environment and to detect opportunities. Thus, the company is able to anticipate an eventual danger through exploring and incorporating new knowledge. The second is the ability to take adequate decisions regarding the environment context based on relevant data according to a variety of sources. The third capability is the ability of the firm to seize opportunities and to respond efficiently and adequately by reconfiguring its resources, modifying business processes and redesigning the organizational structure to choose the best the best solution among the possible alternatives [22, 23].

2.3

Impact of AI on Organizational Agility: Systematic Literature Review

• 1st SLR Method: We’ve conducted a systematic literature review in order to explore the relationship between AI adoption, organizational agility and firm’s performance. We’ve

An Integrated Human-AI Framework ...

77

chosen three inputs for our study including the resources indexed in the two main databases (Scopus and Web of Science) and the ones derived from the IEEExplore digital library. Our query on all fields since the last twenty years is the following: (“Artificial intelligence” AND “performance” AND “agility”) OR (“AI” AND “performance” AND “agility”). For each of the query’s results of the three inputs, we assessed the titles and abstracts in order to exclude the papers related to other disciplines. The rest of the papers were fully analyzed and the papers related to the subject of our study were kept. Finally, we’ve merged the four results and removed the duplicates as explained in Fig. 3 below. • 1st SLR results and discussion: Our input consists of 166 papers resulting from executing our query on the four sources described above. In step 2, the papers excluded from our four inputs were mainly related to other disciplines namely medicine, neuroscience, sport, astronomy or physics. The 76 remaining papers were fully assessed in order to include only the ones related to our research question. Thus, 24 relevant papers were selected in step 3. The final output was obtained in step 4 after removing duplicate papers among the four streams of selected papers in step 3. Table 2 below describes the 16 papers resulting from our 1st systematic literature review. The selected papers emphasize the role of artificial intelligence in delivering better customer experience and enhancing productivity. However, most of the papers studies this impact only from one lens namely supply chain agility [25, 28, 33, 35, 36, 38], agile operations/manufacturing [29, 37] or purchasing [24]

Fig. 3 The steps of our 1st systematic literature review

78

M. A. Marhraoui et al.

Table 2 SLR results Paper

Cite N°

Research approach

Type

Main contributions

(Allal-Chérif et al., 2021) [24]

–

Qualitative

Case study

(Brintrup et al., 2020) [25]

10

Qualitative

Case study

(Roumani et al., 2020) [26]

3

Qualitative

Comparative study

(Williams et al., 2020) [27]

1

Conceptual

–

(Pattanaik et al., 2020) [28]

1

Qualitative

Case study

(Gomes et al., 2020) [29]

1

Qualitative

Case study

(Chen et al., 2020) [30]

1

Quantitative

Survey

(Gupta et al., 2020) [31]

8

Conceptual

–

(Souza et al., 2019) [32]

7

Qualitative

Case study

(Panichayakorn and Jermsittiparsert, 2019) [33] (Bogdan and Borza, 2019) [34]

4

Quantitative

Survey

How AI can help the company to improve the performance of purchasing How machine learning can help the company to predict supply chain disruptions Compares the classification performances of predicting US software firm’s failure Explores how the integration of emergent information systems can enhance enterprise agility Proposes a two objectives algorithm in order to optimize the combination of suppliers Evaluates the impacts of implementing Artificial intelligence in dam operations Study of how business analytics use on an individual level can have a positive impact on organizational agility Proposes a Framework organizing how cutting-edge technologies can help the company to create value and enhance customer experience Studies how user experience can be enhanced by using AI in mouse tracking Studies the impact of AI awareness on supply chain agility and performance

–

Literature review

Systematic

(Fu and Chien, 2019) [35]

13

Qualitative

Case study

Extracts the most relevant papers linking Big data analytics and firm’s performance Proposes and tests a framework relying on AI for predicting demands and enhancing the supply chain agility (continued)

An Integrated Human-AI Framework ...

79

Table 2 (continued) Paper

Cite N°

Research approach

Type

Main contributions

(Lee et al., 2011) [36]

96

Conceptual

–

(Guo and Zhang, 2010) [37]

59

Conceptual

–

(Tse et al., 2009) [38]

9

Conceptual

–

(Guo and Zhang, 2010) [37]

92

Conceptual

–

Studies how AI and RFID can enhance the responsiveness of the supply chain Describes the impact of multi-agent technology on the performance of manufacturing through optimized scheduling Proposes an integrated solution based mainly on multi-agent and ANN in order to achieve an agility of logistics flows Proposes an intelligent system based on autonomous agents which help the company to optimize the scheduling of new product development

Moreover, only one paper has studied the impact of the effective use of Business analytics on organizational agility [30]. • Identified research gaps: However, there some insufficiencies on relying only on a digital enabler. Indeed, although artificial intelligence allows things have certain properties, digital machines may not possess a spirit or creativity as humans do. Also, from an epistemological point of view, artificial intelligence can only respond to some specific problems by relying on facts that can be represented to a digital computer and drawing conclusions. We can pretend that artificial intelligence may not replace the human mindset which is mandatory for an agile transformation of the firm. In addition, AI systems can manifest unforeseeable actions and behaviors in real life after leaving the care of their designers. Humans should then be implicated in regulating and controlling the artificial intelligence engine deployed in order to help the company to be more agile.

80

2.4

M. A. Marhraoui et al.

Impact of HR Practises on Organizational Agility: Systematic Literature Review

• 2nd SLR Method: We’ve followed the same method for the second SLR by using the three main databases as inputs and the same inclusion/exclusion criteria. For this SLR, our query on all fields since the last twenty years is the following: (“Human Resources AND “performance” AND “agility”) OR (“HR” AND “performance” AND “agility”). The 4 steps for papers selection are described in Fig. 4 below. • 2nd SLR results and discussion: Table 3 below describes the 16 papers resulting from our 2nd systematic literature review. The results of our 2nd systematic literature review confirm that companies should invest in developing human resources practices in order to embrace agility. As per [48], the human resources practices include three dimensions: attitudes, behaviors and skills of employees. On one hand, strengthening the mindset of employees is likely to stimulate innovation throughout the organization [54]. On the other hand, employees’ behavior is related to their resistance to change [55]. This latter may be motivated by a lack of vision on the future of the company, a low motivation among employees and a lack of creativity when looking for an appropriate change strategy [56]. In addition, the pro-activity of employees is another desired behavior. It is linked to initiative taking [57], autonomy and commitment to work [58]. As the new economy is hugely based on knowledge,

Fig. 4 The steps of our 2nd systematic literature review

An Integrated Human-AI Framework ...

81

Table 3 2nd SLR result Paper

Cite N°

Research approach

Type

Main contributions

(Heilmann et al., 2020) [39] (Hamidianpour et al., 2016) [40]

21

Quantitative

Survey

5

Quantitative

Survey

(Harsh and Festin, 2019) [41] (Doz, 2020) [42]

25

Qualitative

Interviews

HRM practices have a positive impact on profitability Electronic human resource management (E-HRM) affect positively organizational agility Dynamic talent management increases organizational agility

32

Conceptual

–

(Dyer and Shafer, 1998) [43] (Al-Qaralleh and Atan, 2021) [44]

239

Conceptual

–

−

Quantitative

Survey

(Marhraoui and EL Manouar, 2017) [45] (Wu and Li, 2008) [46]

4

Conceptual

–

4

Quantitative

Survey

(Saha et al., 2017) [47]

46

Conceptual

–

(Marhraoui and El Manouar, 2020) [48]

2

Conceptual

–

(Ananthram and Nankervis, 2013) [49] (e Cunha et al., 2020) [50] (Teimouri et al., 2017) [51]

35

Qualitative

Interviews

30

Conceptual

–

12

Quantitative

Survey

–

Conceptual

–

–

Qualitative

Case study

(Asfahani, 2021) [52] (Karman, 2019) [53]

Identifies HR practices’ enablers for strategic agility Identifies personal competencies and behaviors for an agile organization Knowledge-based HRM practices, business analytics and organizational agility have positive impact on innovative performance Identify groups of organizational agility enablers including HR practices IT and managerial capabilities have a positive impact on agility and performance Proposes a conceptual framework linking HR management, performance and organizational agility Proposes a Framework examining the complementary enabling role of IT and Human Resources practices on organizational agility Alignment between business and HRM strategies allows more agility Proposes six HRM domains of action which enables greater agility Finds strong relationship between human resource management actions and organizational agility Identifies HR practices in the literature which increases organizational agility HR flexibility has a positive impact on sustainable competitiveness

82

M. A. Marhraoui et al.

human resources should be qualified and able to disseminate and exploit new knowledge [59]. Moreover, in today’s turbulent environment, employees should work on self-development through continuous learning, self-determination and motivation [51, 60]. This lifelong learning is facilitated by employees’ risk-taking to get out of the comfort zone, active listening and open-mindedness about new ideas. Finally, the proactive/innovative mentality and the appropriate behaviors of the employees require an effective and adaptable strategy of human resources management [49]. Indeed, human resources are at the heart of the development of the company, which, in order to be agile, must adopt adequate methods of managing human resources. This includes training and skills development, performance evaluation and compensation systems, talent recruitment, job descriptions and benefits. • Identified research gaps: However, there are insufficiencies for relying only on HR enabler. Indeed, some human agile attributes cannot be accomplished efficiently if they are not supported by adequate tools and processes. For example, the competency of taking initiative is fully adopted if the leader of the employee is able to scan efficiently the firm environment [61]. The willingness and openness of employees can be enhanced through the use of IT resources [62, 63]. In addition, the adoption of new AI tools may not be sufficient if managers and employees are not conscious of the real value of these IT resources in helping their companies to be more agile [64].

3 Proposition of an İntegrated Human-AI Framework for an Agile Firm As seen earlier, there are shortcomings when relying only on the technological enablers (especially artificial intelligence) or on human resource practices. In order to overcome these shortcomings, we’ve proposed a new Framework allowing firms to enhance their sensing, decision making and responding capabilities by enabling the adequate AI technologies and the flexible HR practices allowing the firm to adapt to society’s changes and to be more performant. Hereunder is our proposed Framework (Fig. 5 below). Image recognition can help companies to discover new ways for delivering seamless customer experience (Amazon Go concept for example) [65]. Furthermore, natural language processing applications like chatbots allow companies to sense its environment by identifying new leads. Otherwise, the sensing capability is enhanced by HR practices including customer centricity, creativity and learning ability of employees in order to find innovative ideas which suits customers’ needs and market trends.

An Integrated Human-AI Framework ...

83

Fig. 5 Our proposed framework

Furthermore, intelligent agents can help for decision making like Anti-money laundering application based on multiple agents for data collection, monitoring and behavior diagnosing [66]. Older AI like expert systems may have either a replacement role on a tactical level or a support role on a strategic level [67]. Besides relying on AI-based decision support systems, managers should take into consideration their intuition and have a flexible mindset/critical thinking in order to take adequate decisions. In order to respond to changes in external environment, companies should deploy both AI-based applications and HR practices. Indeed, using robotics allows automating repetitive tasks and enable workers to focus on higher added value ones. Also, artificial neural networks (ANN) enhance 3D printing in terms of design and process [68]. Thus, companies are able to design and deliver products with changing features in a rapid manner. Furthermore, companies can act more efficiently and rapidly by continuously training their employees to the use of disruptive technologies and reorganizing in small teams. This enhances collaboration, autonomy and cross-functionality and thus, allows the company to optimize the time to market. These companies are more agile and ultimately ensure a sustainable performance. Indeed, Different studies in the have linked organizational agility and firm’s performance. Indeed, [69] has proposed a Framework linking the strategic alignment of information systems to the performance of the company through organizational agility. This latter enables the company with a set of options in response to changes in the environment and therefore future gains in terms of profitability and cost reduction. The empirical study carried out over 241 companies has shown that alignment has a positive effect on agility and this latter has a positive effect on the performance of the company, especially in a turbulent environment. In addition, [70] has examined the relationship between IT capabilities/resources and business performance with emphasis on the intermediary role of organizational agility. An

84

M. A. Marhraoui et al.

empirical study, carried out on 109 companies, concludes that entrepreneurial and adaptive agility affects positively the performance of the company more particularly in a dynamic environment. This strengthens the positive relationship between organizational agility and sustainable performance in our proposed Framework.

4 Conclusion and Perspectives In this article, we have proposed a new Framework helping companies to take advantage either from artificial intelligence technologies or human resource practices enablers in order to be more agile and ultimately ensuring a sustainable performance. This is an original approach as the literature emphasizes on the impact of one of the two enablers. Future work will focus on a case study in order to get an in-depth knowledge of the practical impact of artificial intelligence and human resource practices on the agility of the enterprise. The objective is to track the firm’s agility after the adoption of AI technologies through adequate use cases.

References 1. S. Sarker, S. Chatterjee, X. Xiao, A. Elbanna, The sociotechnical axis of cohesion for the IS discipline: its historical legacy and its continued relevance. MIS Q. 43(3), 695–720 (2019) 2. M.A. Marhraoui, A. El Manouar, IT innovation and firm’s sustainable performance: the mediating role of organizational agility, in ACM International Conference on Proceeding Series (2017), pp. 150–156 3. M.A. Marhraoui, A. El Manouar, IT Innovation and firm’s sustainable performance: the intermediary role of organizational agility—an empirical study. Int. J. Inf. Eng. Electron. Bus. 10(3), 1–7 (2018) 4. A.M. Turing, Computing machinery and intelligence. Mind 49, 433–460 (1950) 5. C.E. Shannon, philosophical magazine series 7 XXII. Programming a computer for playing chess. Philos. Mag. Ser. 7314(41), 256–275 (1950) 6. A.L. Samuel, Eight-move opening utilizing generalization learning. (See Appendix B, game G-43.1 some studies in machine learning using the game of checkers. IBM J. 3, 210–229 (1959) 7. D. Srinivasan, S. Member, M.C. Choy, R.L. Cheu, Neural networks for real-time traffic signal control. IEEE Trans. Intell. Transp. Syst. 7(3), 261–272 (2006) 8. Y. Dai, J. Hu, D. Zhao, I. Member, F. Zhu, Neural network based online traffic signal controller design with reinforcement training (2011), pp. 1045–1050 9. J. Mccarthy, Review of artificial intelligence: a general survey. Artif. Intell. 5, 317–322 (1974) 10. T. Menzies, 21st-century Al: proud, not smug. IEEE Intell. Syst. 18(3), 18–24 (2003) 11. I. Horswill, Polly: a visisn-based artificial agent, in AAAI 1993 Proceedings (1993), pp. 825– 829 12. F.H. Hsu, IBM’s deep blue chess grandmaster chips. IEEE Micro 19(2), 70–81 (1999) 13. PwC, The essential eight technologies board byte: artificial intelligence. PwC Glob., no. October, pp. 1–8 (2017)

An Integrated Human-AI Framework ...

85

14. L. Steels, V.U. Brussel, The articial life roots of articial intelligence. Artif. Life 1(1_2), 75– 110 (1993) 15. E. Rich, Artificial intelligence and the humanities. Comput. Hum. 19(2), 117–122 (1985) 16. M. Wooldridge, An automata theoretic approach to multiagent planning, pp. 1–15 (2003) 17. J. Borana, Proceeding of international conference on emerging technologies in engineering, biomedical, management and science (ETEBMS-2016). Appl. Artif. Intell. Assoc. Technol. (March), 5–6 (2016) 18. D.J. Teece, M. Peteratd, S. Leih, Dynamic capabilities and organizational agility. Calif. Manage. Rev. 58(4), 13–35 (2016) 19. Z. Zhang, H. Sharifi, A methodology for achieving agility in manufacturing organisations. Int. J. Oper. Prod. Manag. 20(4), 496–513 (2000) 20. S. Goldman, R. Nagel, K. Preiss, Agile competitors and virtual organizations. VANNOST (1995) 21. N.A. Wageeh, Organizational agility: the key to organizational success. Int. J. Bus. Manag. 11 (5), 296 (2016) 22. R. Dove, Agile enterprise cornerstones: Knowledge, values, and response ability. IFIP Adv. Inf. Commun. Technol. 180, 313–330 (2005) 23. E. Overby, A. Bharadwaj, V. Sambamurthy, Enterprise agility and the enabling role of information technology. Eur. J. Inf. Syst. 15(2), 120–131 (2006) 24. O. Allal-Chérif, V. Simón-Moya, A.C.C. Ballester, Intelligent purchasing: how artificial intelligence can redefine the purchasing function. J. Bus. Res. 124(October 2020), 69–76 (2021) 25. A. Brintrup et al., Supply chain data analytics for predicting supplier disruptions: a case study in complex asset manufacturing. Int. J. Prod. Res. 58(11), 3330–3341 (2020) 26. Y.F. Roumani, J.K. Nwankpa, M. Tanniru, Predicting firm failure in the software industry. Artif. Intell. Rev. 53(6), 4161–4182 (2020) 27. O. Williams, F. Olajide, T. Al-Hadhrami, A. Lotfi, “Exploring process of information systems and information technology for enterprise agility, in Emerging Trends in Intelligent Computing and Informatics. IRICT 2019, ed. by F. Saeed, F. Mohammed, N. Gazem. Advances in Intelligent Systems and Computing, vol. 1073 (Springer, Cham, 2019), pp. 1042–1051. https://doi.org/10.1007/978-3-030-33582-3_98 28. L.N. Pattanaik, P. Agarwal, S. Ranjan, U. Narayan, Bi-objective optimization of a reconfigurable supply chain using a self-organizing migration algorithm, in Advances in Computational Intelligence, ed. by S. Sahana, V. Bhattacharjee. Advances in Intelligent Systems and Computing, vol. 988 (Springer, Singapore, 2020), pp. 39–53. https://doi.org/10. 1007/978-981-13-8222-2_4 29. M.G. Gomes et al., Economic, environmental and social gains of the implementation of artificial intelligence at dam operations toward industry 4.0 principles. Sustain. 12(9), 3604 (2020) 30. X. Chen, K. Siau, Business analytics/business intelligence and IT infrastructure. J. Organ. End User Comput. 32(4), 138–161 (2020) 31. S. Gupta, A. Leszkiewicz, V. Kumar, T. Bijmolt, D. Potapov, Digital analytics: modeling for insights and new methods. J. Interact. Mark. 51, 26–43 (2020) 32. K.E.S. Souza, M.C.R. Seruffo, H.D. De Mello, D.D.S. Souza, M.M.B.R. Vellasco, User experience evaluation using mouse tracking and artificial intelligence. IEEE Access 7, 96506– 96515 (2019) 33. T. Panichayakorn, K. Jermsittiparsert, Mobilizing organizational performance through robotic and artificial intelligence awareness in mediating role of supply chain agility. Int. J. Supply Chain Manag. 8(5), 757–768 (2019) 34. M. Bogdan, A. Borza, Big data analytics as a strategic capability: a systematic review, in Proceedings of the International Management Conference (2019), pp. 575–583 35. W. Fu, C.F. Chien, UNISON data-driven intermittent demand forecast framework to empower supply chain resilience and an empirical study in electronics distribution. Comput. Ind. Eng. 135(July), 940–949 (2019)

86

M. A. Marhraoui et al.

36. C.K.M. Lee, W. Ho, G.T.S. Ho, H.C.W. Lau, Design and development of logistics workflow systems for demand management with RFID. Expert Syst. Appl. 38(5), 5428–5437 (2011) 37. Q. Guo, M. Zhang, An agent-oriented approach to resolve scheduling optimization in intelligent manufacturing. Robot. Comput. Integr. Manuf. 26(1), 39–45 (2010) 38. Y.K. Tse, T.M. Chan, R.H. Lie, Solving complex logistics problems with multi-artificial intelligent system. Int. J. Eng. Bus. Manag. 1(1), 37–48 (2009) 39. P. Heilmann, R. Forsten-Astikainen, S. Kultalahti, Agile HRM practices of SMEs. J. Small Bus. Manag. 58(6), 1291–1306 (2020) 40. F. Hamidianpour, M. Esmaeilpour, H. Firoozi, Assessing the impact of electronic human resource management on creation of organizational agility: a study in the Bushehr Banks, Iran. Asian Soc. Sci. 12(7), 105–118 (2016) 41. K. Harsch, M. Festing, Dynamic talent management capabilities and organizational agility—a qualitative exploration. Hum. Resour. Manage. 59(1), 43–61 (2020) 42. Y. Doz, Fostering strategic agility: How individual executives and human resource practices contribute. Hum. Resour. Manag. Rev. 30(1), 100693 (2020) 43. L. Dyer, R. A. Shafer, Working paper series from human resource strategy to organizational effectiveness: lessons from research on organizational agility, pp. 1–35 (1998) 44. R. Enad Al-Qaralleh, T. Atan, Impact of knowledge-based HRM, business analytics and agility on innovative performance: linear and FsQCA findings from the hotel industry. Kybernetes 51(1), 423–441 (2021) 45. M.A. Marhraoui, A. El Manouar, IT-enabled organizational agility–proposition of a new framework. J. Theor. Appl. Inf. Technol. 95(20), 5431–5442 (2017) 46. W. Jun, L. Ruixue, Modeling the relationship between IT capability, uncertainty and organizational performance: an empirical study, in 2008 IEEE Symposium on Advanced Management of Information for Globalized Enterprises AMIGE 2008 - Proceedings (2009), pp. 191–195 47. N. Saha, A. Gregar, P. Sáha, Organizational agility and HRM strategy: do they really enhance firms’ competitiveness? Int. J. Organ. Leadersh. 6(3), 323–334 (2017) 48. M.A. Marhraoui, A. El Manouar, Organizational agility and the complementary enabling role of IT and human resources: proposition of a new framework, in ICT for an Inclusive World, ed. by Y. Baghdadi, A. Harfouche, M. Musso. Lecture Notes in Information Systems and Organisation, vol. 35 (Springer, Cham, 2020), pp. 55–65. https://doi.org/10.1007/978-3-03034269-2_4 49. S. Ananthram, A. Nankervis, Strategic agility and the role of HR as a strategic business partner: an Indian perspective. Asia Pac. J. Hum. Resour. 51(4), 454–470 (2013) 50. M.P.E. Cunha, E. Gomes, K. Mellahi, A.S. Miner, A. Rego, Strategic agility through improvisational capabilities: implications for a paradox-sensitive HRM. Hum. Resour. Manag. Rev. 30(1), 100695 (2020) 51. H. Teimouri, K. Jenab, H.R. Moazeni, B. Bakhtiari, Studying effectiveness of human resource management actions and organizational agility: resource management actions and organizational agility. Inf. Resour. Manag. J. 30(2), 61–77 (2017) 52. A. Asfahani, The impact of modern strategic human resources management models on promoting organizational agility. Acad. Strateg. Manag. J. 20(2), 1–1 (2021) 53. A. Karman, The role of human resource flexibility and agility in achieving sustainable competitiveness. Int. J. Sustain. Econ. 11(4), 324–346 (2019) 54. J.S. Boudrias, P. Gaudreau, A. Savoie, A.J.S. Morin, Employee empowerment: from managerial practices to employees’ behavioral empowerment. Leadersh. Organ. Dev. J. 30(7), 625–638 (2009) 55. J. Shin, M.S. Taylor, M.G. Seo Resources for change: the relationships of organizational inducements and P...: EBSCOhost. Acad. Manag. J. 55(3), 727–748 (2012) 56. M.P. Del Val, C.M. Fuentes, Resistance to change: a literature review and empirical study. Manag. Decis. 41(2), 148–155 (2003)

An Integrated Human-AI Framework ...

87

57. M. Frese, D. Fay, T. Hilburger, K. Leng, A. Tag, The concept of personal initiative: operationalization, reliability and validity in two German samples. J. Occup. Organ. Psychol. 70(2), 139–161 (1997) 58. D.N. Den Hartog, F.D. Belschak, Personal initiative, commitment and affect at work. J. Occup. Organ. Psychol. 80(4), 601–622 (2007) 59. Banque mondiale, Building Knowledge Economies: Advanced Strategies for Development. WBI Development Studies. Washington, DC: World Bank. © World Bank (2007) 60. M. London, J.W. Smither, Empowered self-development and continuous learning. Hum. Resour. Manage. 38(1), 3–15 (1999) 61. P.M. Wright, S.A. Snell, Toward a unifying framework for exploring fit and flexibility in strategic human resource management. Acad. Manag. Rev. 23(4), 756–772 (1998) 62. R. Broderick, J.W. Boudreau, Human resource management, information technology, and the competitive edge. Acad. Manag. Perspect. 6(2), 7–17 (1992) 63. T. Davenport, Saving IT’s soul: human-centered information management. Harv. Bus. Rev. 72(2), 119–131 (1994) 64. T.C. Powell, A. Dent-Micallef, Information technology as competitive advantage: the role of human, business, and technology resources. Strateg. Manag. J. 18(5), 375–405 (1997) 65. E.Y. Li, Amazon go concept. J. Bus. Manag. 24(1), 79–92 (2018) 66. S. Gao, D. Xu, Conceptual modeling and development of an intelligent agent-assisted decision support system for anti-money laundering. Expert Syst. Appl. 36(2 PART 1), 1493– 1504 (2009) 67. Y. Duan, J.S. Edwards, Y.K. Dwivedi, Artificial intelligence for decision making in the era of big data—evolution, challenges and research agenda. Int. J. Inf. Manage. 48(February), 63–71 (2019) 68. G.D. Goh, S.L. Sing, W.Y. Yeong, A review on machine learning in 3D printing: applications, potential, and challenges. Artif. Intell. Rev. 54(1), 63–94 (2021) 69. P.P. Tallon, A. Pinsonneault, Competing perspectives on the link between strategic information technology alignment and organizational agility: Insights from a mediation model. MIS Q. Manag. Inf. Syst. 35(2), 463–486 (2011) 70. A. Chakravarty, R. Grewal, V. Sambamurthy, Information technology competencies, organizational agility, and firm performance: enabling and facilitating roles. Inf. Syst. Res. 24(4), 976–997 (2013)

The Place of Stock Photography as a Digital Commerce in Turkey İsa Avcı , Murat Koca , and Büşra Uysal

Abstract With the development of technologies, digital commerce is increasing in many sectors today. The concept of digital commerce has gained more importance in recent years due to the pandemic. Stock photography, which is a branch of digital commerce and passive income model, is a sector that has no development in Turkey and benefits from foreign-sourced services. This study aims to find a domestic solution to this issue and make a study that will facilitate the work of digital content producers. In order to carry out these studies, it is necessary to follow and implement the websites and the innovations that come with web 2.0. In addition, the software used forms the basis of every detail that serves its purpose. The fact that the software used in stock photography is flexible and has a structure that meets the needs increases the work’s applicability. In the globalizing world, giving importance to such issues has become as important as health and education. The state of development in countries’ economies is such an essential issue that it can direct these issues worldwide. Working on this issue will contribute significantly to developing countries and digital commerce in terms of stock photography. Especially in this study, evaluations will be made regarding the place and application areas of stock photography as digital commerce in Turkey. Keywords Digital content production


Digital commerce
Stock photography

İ. Avcı (&)
B. Uysal Engineering Faculty, Computer Engineering, Karabuk University, 78050 Karabuk, Turkey e-mail: [email protected] M. Koca Provincial Directorate, Republic of Turkey Ministry of Industry and Technology, Hakkari, Turkey © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_8

89

90

İ. Avcı et al.

1 Introduction Stock photography, one of the subjects that should be dealt with in web technologies, is not an unknown and valuable concept in our country. Working and developing on this subject will make a domestic contribution on behalf of our country. Internet use is widespread in our country, where the youth population density is high. The work that comes with the internet has gained importance today. In new sites developed with web technologies, users can open their sessions, share their posts, and establish their friendship networks. It should be a project developed in the light of these technologies. Although the administrator’s authority is comprehensive, member transactions should be an environment where the member can handle his work comfortably. Stock photography is becoming one of the most popular areas of digital content production. When we look at this sector, it is seen that many people, from the site designer to the photographer, are employed in this sector. The site to be designed for the project must meet the customer’s needs and must not provide easy and reliable service in trading transactions. Digital commerce is a concept that we have encountered frequently in recent years and that is developing day by day. It is necessary to develop and work on stock photography in this field. Although it is not a subject we frequently encounter today, it is promising for the future. Considering these and similar areas contributes to the catching up of developing technologies. This study aims to conduct a study in the field of the economy depending on the developing and changing technologies in the digital age. It is seen that there are innovations in many areas in the virtualized world. Being aware that these areas are developing every day and their structure is growing, opportunities in these issues should be seized. Although web design may seem like an old method for computer science, it is a concept that keeps up with today’s technologies and improves itself day by day. The webspace is constantly changing and following new technologies. In this study, studies on stock photography, methods, Turkey’s place will be given information on the subject.

2 Literature Reviewer In the globalizing world, it is seen that the concept of virtual has entered human life more and more every day. In addition to areas such as digital commerce, the concept of digital commerce, which is the transfer of commerce to the internet, has recently become a field of interest. Digital commerce is a fast, time-saving, and financially efficient method. The most significant support in the development of this subject stems from the increase in studies on the web environment. Every development in the web environment has supported the development of the digital commerce environment. It is a subject that is very open to development and change. In the article titled “Trade for Entrepreneurs: Digital commerce”, published by

The Place of Stock Photography as a Digital Commerce in Turkey

91

Mehmet Marangoz, it was mentioned that the concept of digital commerce is of great importance for the globalizing world and that the potential workforce in this field is high [1]. M. Marangoz et al., in their study titled “Examination of Web and Social Network Sites of Digital commerce Businesses by Content Analysis Method”, mentions that businesses that want to achieve commercial success should meet the needs in terms of web and social networking. It is seen that businesses that pay attention to web and social network issues show high turnover and a high amount of development in a short time. The networks established as a result of the experimental observations made various suggestions. These recommendations are used for the business to obtain higher profits with less cost. It handles the operations of this concept with an infrastructure originating from data science. Internet technologies, which have gained a different dimension in their field, have succeeded in bringing a new dimension to commercial developments in technological innovations. The results of digital commerce projects have opened a great door to content analysis [2]. M. Demirdövmez et al. addressed this issue in the article titled “Development of Digital commerce Sector in Turkey by Years”. Depending on the young population in Turkey, the use of the internet to rise to high levels and to adapt to the digital commerce environment is quite fast. It is seen that some developing and developing countries use digital commerce as the primary source of income [3]. Peter Trasewich et al., “Issues in Mobile Digital commerce”, written the convenience of digital commerce in mobile devices used as a wireless communication tool to humanity and the difficulties it brings along with this ease are emphasized with a different perspective [4]. J. B. Schafer et al. This issue was mentioned in the article titled “Recommender System in Digital commerce,” written by Many of the websites used today have developed systems that recommend their customers to find the product to sell. These systems that provide advice provide the most valuable product available [5]. J. Glückler is a long evaluation on this subject in the article titled “The Evolution of a Strategic Alliance Network: Exploring the Case of Stock Photography.“ It is thought that connecting the phenomenon of geographically recognizing nature and exploring the environment with an economic link will significantly contribute and start to stock photography [6]. P. Frosh has examined this issue in his work “Inside the Image Factory: Stock Photography and Cultural Production.“ Many areas such as advertising, marketing processes, website designs are supplied with stock photography. Since this work area is provided within specific countries and borders, it causes significant financial losses to those in need [7]. Z. Kalazic et al., in the article titled “The Stock Photography is a Part of Cultural and Creative Industries of the Digital Age,” explains that, just like in other media, photography has lost its intangibility by being transferred to the virtual environment. It is emphasized that it is a subject that needs to adapt to the virtual environment. Stock photography is seen as a necessity of the modern age that can be used repeatedly in the digital environment. Following the developments in this field has become a necessity rather than a choice. It has become a compelling reason [8]. P. Frosh, “Digital Technology and Stock

İ. Avcı et al.

92

Photography”, emphasized that digital photography is an important field and that these studies are used in many areas such as site designs and advertising. It has been determined that the resources used in using ready-made templates for websites and photoshop designs come from this sector [9]. Z. Karabacak, “An Evaluation on Stock Photography,” first introduced the websites and then shared extended information about the member login and posting admin processes on these sites to be able to sell. The acceptance and sale of the uploaded photos should be determined by the site [10]. W. Feighery, in his study titled “Tourism, Stock Photography, and Surveillance: a Foucauldian Interpretation”, emphasized the importance of stock photography in terms of culture and tourism and mentioned that these areas should be carried out in a digital environment [11]. Z. Maamar, “Commerce, Digital commerce, and M-Commerce: What Comes Next?” In his article named digital commerce and m-commerce, the development and change of the concept known as mobile commerce and how it took shape over time are explained. In addition, evaluations were made about what kind of future awaits us in the next step of these fields. Depending on the developing IT technologies, the course of the concept of commerce takes shape. Future studies are carried out in this area [12]. E. Runge, “The Travels of Photography within The Global Image Market. In his study titled “How Monopolisation, Interconnectedness, and Differentiation Shape The Economics of Photography”, it was written that digital image systems were reshaped according to analog image systems and monopolization in this area continued through the web. The website and agency used in the development of such an industry have great importance. It is known that the site undertakes the most excellent load for marketing and selling the products. If the site has a fluid and exciting structure, the customer will want to spend more time on these sites [13].

3 Material and Method 3.1

PHP Programming Language

PHP programming language was used in the stock photography project. PHP programming language is a preferred and widely used language in web studies. It needs a compiler and also has a structure similar to the C and PERL programming languages. So it is easy to understand for a programmer. Today, it is shaped according to object-oriented programming languages. Recently, it contains many plugins, functions, and libraries [14]. The innovations that came with Web 2.0 contributed to many projects, as seen in Fig. 1. It is available in many plugins for doing business using databases in PHP language. The most common usage is MySQL. MySQL database was also used in the stock photography project.

The Place of Stock Photography as a Digital Commerce in Turkey

93

Fig. 1 Web development [15]

Fig. 2 The route in the Laravel framework [17]

3.2

Laravel Framework

Laravel is recognized as the best programming tool for developing PHP-supported programming languages [16]. These results were obtained with the use of methodology and the prominence of the ability to observe. Laravel Function, a framework used in stock photography applications, was preferred because it has a comprehensive and understandable structure. Laravelin life cycle is shown in Fig. 2. Codes written using the PHP programming language come to the route determination parts called a route. From here, it is responsible for

94

İ. Avcı et al.

performing whatever operations are to be performed with the controller. The model part is responsible for the design of the database. The view part represents the parts created by the HTML and CSS codes displayed in the browser.

4 The Place of Stock Photography at Digital Commerce in Turkey Electronic commerce (digital commerce) is a concept that eliminates time and space limits with low cost. It provides transportation to more than one person with sound and image. The concept of digital commerce continues to develop day by day, and it is not possible to make a clear definition of digital commerce. Many different issues that digital commerce includes have appeared as in Fig. 3. These topics are electronic communication, business process, service, and online situations. Digital commerce should not be evaluated in a way that can be ordered, just like a catalog. It can think of it as a more comprehensive management network.

Fig. 3 Architecture for a standard digital commerce system [18]

The Place of Stock Photography as a Digital Commerce in Turkey

4.1

95

Role of Social Media and Web Sites

The internet protocol known as TCP/IP is developing day by day. This protocol is based on a high-performance telephone line. In this context, the internet is of great importance for humanity. It has a great philosophy, especially in the millennium age. After 2001, the use of free internet, known as Web 2.0, increased its prevalence and enabled us to have a greater say on the web [2]. The websites used in the early days of the internet were utterly admin-managed sites. Users entering these sites had to take advantage of only admin posts. However, with the innovations that came with Web 2.0, the dot-com concept was renewed as a name, and then the websites became member-oriented instead of user-admin. In this way, users can become members, share posts, and communicate with the private message system. In addition, members have transferred multimedia files such as pictures and videos to the internet, depending on their wishes. In a world where many innovations are transferred to the virtual environment, economic power affects day by day. Therefore, commercial expectations are directed towards these areas. The job of meeting these needs in these areas is primarily up to the software developers. A good software developer can use this field as a significant gain for himself and his environment with his flexible ideas. It is one of the methods of providing job opportunities to many software developers. On the other hand, a social network is a structure that connects people on such a website. The word network has many definitions and structures in computer science. It is a concept that is open to further research as a feature. However, the concept that is viewed as a social network has a slightly different structure from the basis of the word network. Individuals provide access to information through the internet as a business, friend, travel destination, and one of the most important. When we look at it from this point of view, we see that the concept of social network has great importance on the internet, on computer science, and commercial aspects. Today, we are witnessing the development and importance of social networking channels for our future. It is also of great importance in terms of software. A well-designed project provides excellent financial and social benefits.

4.2

The Development of Digital Commerce in Informatics

The concept of digital commerce is directly proportional to the use of the internet. The widespread use of the internet has provided the most significant contribution to the development of the sector. Many factors support the development of digital commerce. However, in our project, it is handled in terms of the IT sector. It has contributed to the development of the youth population in our country in this area (Fig. 4). It differs according to the structure of the countries, their political and economic conditions. In addition, the legal structure of the countries is also essential. Another

96

İ. Avcı et al.

Fig. 4 Computer use, internet access, and web page ownership in entrepreneurs [19]

critical issue that needs to be addressed is strengthening the internet infrastructure, which can only be done with investments. The speed and quality of the internet show the power of a project done on the internet [3]. Internet infrastructure performance provides support to increase service quality and increase financial gains. Quality service will lead to gains within the country and will provide an expectation for commercial relations with different countries in the future. In order to realize all these situations, internet service should be provided at the highest level, and new technologies should be followed. The ecosystems that feed the concept of digital commerce are shown in Fig. 5. When we look at this figure, the ecosystems required for a country are grouped. The increase in internet technologies supports the increase in current job opportunities. The sale of marketed products is offered to the user in easier ways in this environment. A user can shop more accurately for a product by comparing different vendors over the internet. The developments available on the internet are helpful to entrepreneurs in generating new business ideas. In addition, entrepreneurs can adapt these business ideas to their lives more quickly. The most crucial building block between entrepreneur, producer, and consumer is web channels. The development of web channels contributes to production, consumption, the country’s economy, and political power. Continuing its development, IT (Information Technologies) provides access to vast internet areas. It is not correct to deal with digital commerce from a single aspect. Multiple options are available. The most important for us is direct digital commerce. It is a system that includes digital commerce multimedia areas directly, away from physicality, and handles trading transactions via virtual. Geographical location is not essential in the concept of direct digital commerce. It

The Place of Stock Photography as a Digital Commerce in Turkey

97

Fig. 5 Digital commerce ecosystem [20]

goes far beyond geography. It is expected that electronic commerce will lead to significant changes in business life in the coming days. It offers a clear and straightforward use. The designs of websites are slightly different from the use of the personal blog or social media. First of all, by knowing how many pieces of the product there are in real life, a field should be opened accordingly in this information database. The Buy tab should be right next to the product. The color and theme of the page are also important. Ideal colors should be used for the product being sold. The most exciting products about the product should be available in the slider section. There should be multiple products at the bottom of the site’s title and add new products in the middle. Information such as contact references should be available at the bottom and top. Apart from the sales part of the site, there should be many items such as surveys, comparisons, advertisements. There are more boxes with advertisements on sites that share personal information. This is to direct the sales process when logged into the site for a different purpose and to meet the site revenues through advertising. A critical issue is that it should not close the existing site when moved to these designed items. It should open on a new page. The site administrator is as essential as the site itself. The person in charge should check the site at regular intervals. The site design should be a competent person in software. Otherwise, healthy site management is not provided. The security of the site is as important as its design. Card information entered by

İ. Avcı et al.

98

individuals with their information should be protected with high-security measures. With the development of digital commerce, the cases of theft have also increased in the virtual environment.

4.3

The Need for Stock Photography

The use of photography has become a necessity in many visual content areas such as websites, magazines, and brochures. The first stock photography agency in the world was established in 1920. It has continued to evolve since its inception. It has turned into an industrial area that is still needed today. It is also called “Image Bank” in Turkish. These organizations act as a bridge between the photographer and the buyer with a particular contract. Stock photos are sold under license without copyright. The customer who makes the purchase purchases the product together with the copyrights. It is obligatory to use it in areas that work with visual content. Because by using a royalty-free product, individuals have taken over the rights of the photo without permission. For this reason, stock photography was born out of a need. It continued its development by feeding on this need [10]. Over time, manual photography has left its place in digital photography, which is why stock photography has gained prominence. Stock photography, which meets today’s needs, has become a sector that employs many workers. This aspect has attracted attention recently. In the first period, the number of photographs was low, and the demand rate was high. For this reason, the fees for the photos have come up in high amounts. In order to normalize this situation, new alternatives were offered, and new websites and environments continued their activities. The increase in the segment dealing with stock photography has led to low prices in product costs. Developments in this area have pushed people to monopolize and facilitate the buyer’s business, not a single centralized administration. Another critical aspect of the developments in this field is the increase in the use of the internet. The development of stock photography is directly proportional to the increase in internet and website usage with today’s innovations. There are some points to be considered in the field of stock photography. One of the essential points is that if a model is used, a form must be filled in that person’s consent is obtained. If he is not an adult, parental consent must be obtained. If the person appears in the photograph without permission, he has the right to file a lawsuit. No trademark or license should appear. Historical artifacts, museums, and buildings are also within this scope. When it is desired to shoot and be published on the website, it should be done by obtaining legal permissions. In addition, the quality of the photos taken is of great importance. Quality photos should be allowed to be used by site administrators, photos with insufficient quality levels should not be allowed.

The Place of Stock Photography as a Digital Commerce in Turkey

4.4

99

Stock Photo Sites

The most common stock photography sites are; Sites such as www.shutterstock. com, www.istockphoto.com are coming. Similar sites have been established in Turkey as well. However, the quality levels are insufficient. If you need to sell on the stock photography site, you must first become a member. Afterward, the terms of membership must be read and accepted. Each site has its criteria. These; conditions such as good image quality and a confidentiality agreement. It is crucial for users who buy an excellent image quality. Considering customer satisfaction, such conditions must be fulfilled. The stock photo uploaded by the member is expected to be purchased on demand. Customers can buy the photo they need from the categories section. Stock photography applications are not limited to this, but they also hand in multimedia environments such as videos and icons. Stock videography is used to purchase any video [18, 21]. The habit of staying at home and working from home, which continues its effect during the pandemic period, is also reflected in the stock photography culture. It is known that such content has increased in similar areas such as home education and the use of masks, and the demand for purchase has increased (Fig. 6). The IT sector aims not to provide quality photographs but to make evaluations on graphics and design in the virtual environment. This type of interest is preferred for adaptation to the virtual environment. It is known that every photograph transferred to the virtual environment is processed as a database and used in data science. It is a sector that also includes amateur shooting. Site managers evaluate these sectors from their perspectives. Therefore, site administrators should be experienced people about the site. Today, digital cameras, cameras, smartphones,

Fig. 6 Primary data about the operations of the Shutterstock agency in 2013 [21]

100

İ. Avcı et al.

Fig. 7 A rights holder in the stock photography market [22]

and drone technologies that come with digital technologies have made a significant contribution to the increase in competition in stock media. Many people, including amateur users, can shoot at professional scales. The issue to be considered here is the development of this scaling and its effect on stock photography. These sites, which proceed following specific rules, offer new experiences to users. In this way, people started to be interested in content production and to work in this field (Fig. 7). All of the photographs we see in images such as posters belong to one source. It is crucial to be entitled to these resources. This right owner and purchase transactions are made through a site on the internet. Users have two options; the first method is to acquire their photos with their efforts. The second is to buy these photos from a photo agency [23]. In Turkey, the use of foreign-sourced sites is every day, and this situation does not provide a financial return to our country. The development of such concepts and the increase in domestic applications in our country make outstanding contributions to us in informatics. Digital content production with applications such as stock photography and video and its derivatives is gaining importance day by day. In our country, such areas should be taken seriously.

5 Conclusion and Discussion There is not much information about stock photography in Turkey. In terms of digital commerce, there are newer beginnings in this field compared to developed countries. In this regard, the concepts should be well known in all aspects and well adapted to the IT sector. Developments in these areas are followed in some major countries. In general, these issues and similar digital content gains should be developed by our country. This article aims to adapt and develop a concept that is

The Place of Stock Photography as a Digital Commerce in Turkey

101

widely used abroad and is not well known in our country. The steps taken in this area contribute to the development of similar projects. Considering the increase in the use of social media and the increase in digital content production, the development of stock multimedia makes significant contributions to human life. In the IT sector, many areas such as health and education are mentioned. However, economic and commercial areas should never be neglected. The economy, which is an element that directs informatics and software, must always be kept alive in this sector. In the project, the concept of digital commerce in the informatics sector was introduced and explained. Such areas should be developed in Turkey. The most important rule of developing these areas is to reach the young generation with such projects. Our country is in a new awakening to such concepts. It cannot be said that such studies have developed, but the young and dynamic population in the country will contribute to the development of these areas in the future. Such studies should not be neglected and should always be taken into consideration and supported.

References 1. M. Marangoz, Girişimciler İçin Sınırsız Ticare:E-Ticaret. Çanakkale Türkiye 6(1), 182–197 (2011) 2. M. Marangoz, B. Yeşildağ, I. Arıkan Saltık, E-Ticaret İşletmelerinin web ve Sosyal Ağ Sitelerinin İçerik Analizi, Yöntemiyle İncelenmesi. İnternet Uygulamarı ve Yönetim Dergisi, Türkiye, Muğla 3(2), 53–78 (2012) 3. M. Demirdöğmez, N. Gültekin, H.Y. Taş, Türkiye’de E-Ticaret Sektörünün Yıllara Göre Gelişimi. Uluslararası Toplum Araştırmaları Dergisi, Şanlıurfa, Türkiye 8, 1–16 (2018) 4. P. Tarasewich, R. Nickerson, M. Warkentin, Issues in mobile digital commerce. Commun. Ais-CAIS 8 (2002). https://doi.org/10.17705/1CAIS.00803 5. J.B. Schafer, J. Konstan, J. Riedl, Recommender systems in digital commerce, in Proceedings of the 1st ACM conference on Electronic commerce (EC 1999) (Association for Computing Machinery, New York, 1999). https://doi.org/10.1145/336992.337035 6. J. Glückler, The Evolution of a Strategic Alliance Network: Exploring the Case of Stock Photography (Edward Elgar, 2010) 7. P. Frosh, Inside the image factory: stock photography and cultural production. Media Cult. Soc. 23(5), 625–646 (2001) 8. Z. Kalazic, J. Horvat, J. Mijoc, The stock photography as a part of cultural and creative industries of the digital age. Interdisc. Manag. Res. Josip Juraj Strossmayer Univ. Osijek Fac. Econ. Croatia 11, 189–203 (2015) 9. P. Frosh, Digital technology, and stock photography: And God created photoshop, in Image Ethics in the Digital Age, ed. by L. Gross, J.S. Katz, J. Ruby (2003), pp. 183–216 10. Z. Karabacak, Stok Fotoğrafçılık Üzerine Bir Değerlendirme. Sanat ve Dil Araştırmaları Enstitüsü Ankara Türkiye 8, 167–177 (2011) 11. W. Feighery, Tourism, stock photography, and surveillance: a Foucauldian interpretation. J. Tour. Cult. Chang. 7(3), 161–178 (2009) 12. Z. Maamar, Commerce, digital commerce, and m-commerce: what comes next? Commun. ACM 46(12), 251–257 (2003) 13. E. Runge, The travels of photographs within the global image market. How monopolisation, interconnectedness, and differentiation shape the economics of photography. Photographies 13(3), 365–384 (2020)

102

İ. Avcı et al.

14. H. Doğan, Php’ye Giriş. Akademik Bilişim, Adana, Türkiye (2003), pp. 1–5 15. WordPress, https://thepaisano.wordpress.com/2008/03/08/web-20-vs-web-30/. Accessed 15 July 2021 16. E.S. Soegoto, Implementing laravel framework website as brand image in higher-education institution, in IOP Conferences Series: Materials Science and Engineering, Bandung City, Indonesia, vol. 407 (2018), pp. 1–16 17. Phpwomen, https://phpwomen.gitbook.io/project/untitled-2. Accessed 11 July 2021 18. M. Agarwal, G. Kar, R. Mahindru, A. Neogi, A. Sailer, Performance problem prediction in transaction-based e-business systems. IEEE Trans. Netw. Serv. Manag. 5, 1–10 (2018) 19. Anthonybmasters, https://anthonybmasters.medium.com/how-many-people-use-the-internet5a361ff7b436. Accessed 18 July 2021 20. Smarttrustuit, https://www.smarttrustuit.com/whitepaperspost/navigating-the-digital commer ce-ecosystem-for-growth-opportunities/. Accessed 08 July 2021 21. X. Chen, Z. Ji, Y. Fan, Y. Zhan, Restful API architecture based on laravel framework. J. Phys. Conf. Ser. 910(1), 012016 (2017). https://doi.org/10.1088/1742-6596/910/1/012016 22. Z. Klazic, J. Horvat, J. Mizoc, The stock photography as a part of cultural and creative industries of the digital age. Interdisc. Manag. Res. XI, 189–203 (2015) 23. G. Johannes, The evolution of a strategic alliance network: exploring the case of stock photography. Handb. Evol. Econ. Geogr. 14, 1–16 (2010)

Fuzzy Classification of the Flow of Events for Decision-Making in Smart Systems Anatolii Kargin

and Tetyana Petrenko

Abstract Decision making engine based on the classification of the data stream in real time is the main challenge of the Smart City Application (SCA). Fuzzy logic system (FLS) can be used as such engine if the data stream dimension will be significantly reduced. The two-stage Computing with Words (CWW) approach overcomes the dimensionality problem by using abstraction engine at the first stage, which maps the meaning of stream data from sensors to the meaning of a few words. In this article the CWW approach is extended by a Short-Term Memory (STM) model that stores a time sequence of events in the form of fuzzy characteristics of words of a high level of abstraction. The STM footprint model is supported by the footprint blur algorithm, in which calculations are performed with blurring of information about events with time expiration. The classification is proposed to be performed by a fuzzy assessment of the proximity of the STM footprint and the prototype of the class of the flow of events. An example of the use of fuzzy classification of the flow of events model is given. Keywords Fuzzy logic systems Short-term memory footprint


Stream classification
Computing with words


1 Introduction A significant place is occupied by Smart City Applications (SCA), in which requires a classification of the events stream in real time. Logistics and city traffic management are as examples. In the first case, the delivery of goods in a non-deterministic environment is considered, when at any stage of the actions plan implementation, unforeseen events may occur. To make a decision, the SCA must A. Kargin (&)
T. Petrenko Ukrainian State University of Railway Transport, Feuerbach Sq., 7, Kharkiv 61050, Ukraine e-mail: [email protected] T. Petrenko e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_9

103

104

A. Kargin and T. Petrenko

know the history in the form of a sequence of events and actions [1, 2]. A similar problem arises when managing the logistics of an automated warehouse in the unexpected failures. The subsequent logistics of their delivery depends on the order in which the containers were previously served [3, 4]. Controlling a smart traffic light in normal and extreme situations requires knowledge of previous events [5, 6]. Without data about what the signals were before and what dynamics (flow of events) filling the queue of cars and pedestrians, it is impossible to make a rational decision about switching. Smart home security systems require an analysis of the flow of events to make a decision on possible actions in unauthorized situations [7, 8]. Such a flow of events can be generated by the movements of an unidentified object in the room, the composition and sequence of actions, atypical for authorized events, and others. Another example of an SCA that requires the classification of the flow of events is the prevention of emergency situations [2, 9]. Monitoring the equipment operating modes and comparing the sequence of events associated with changes in parameter values with prototypes of events preceding accidents is required to make decisions about turning off the equipment and turning on the backup one. This is not a complete list of tasks that are solved when creating smart systems of various types, for which the classification of the events flow is required. The flow of events that are processed by the previously listed applications are non-stationary [10], which is characterized by variations in the time intervals between events, the fuzzy and ambiguous description of the events themselves, and possible variations in their sequence. The ability to continuously collect and process data streams creates additional problems in the construction of classification algorithm. They are one-pass constraint, concept drift and massive domain constraint. A large arsenal of Artificial Intelligence’s (AI) methods, models and algorithms are used to overcome the listed challenges [11]. All of them are aimed at solving the problem of establishing the relationship between a set of feature variables and a target variable of interest. The most related works are devoted to the development and encapsulation in SCA of an algorithm suitable for a domain in which classification results are used for autonomous decision making. They discuss integrating the flow of events with a decision engine based on a specific development framework. Such solutions exist based on technology of pattern design of the IoT. In this patterning technologies, decision-making is developed on the basis of the Rules Engine (RE) pattern, the concept of which is borrowed from AI, namely, rule-based systems [12]. If the specificity of the domain is such that the decision-making model is a simple mapping of the classes of the flow of events into actions, then the RE does not require a special decision engine. In such SCA, the RE directly implements the procedure for classifying the data flow in real time and its basis is the model and algorithm on which the classification is based. So frameworks are known, the RE of which processes the data stream using different models and methods: Decision Trees, Rule-Based, Nearest Neighbor, SVM methods and Probabilistic Models for Classification and Neural Network Classifiers. Comparison of the capabilities of some REs based on different models can be seen in the work [13]. If the specificity of the SCA domain is such that decision-making requires multi-step logical

Fuzzy Classification of the Flow of Events …

105

inference using not only the results of classifying the flow of events, but also other facts and data, then RE must use a special decision engine. Smart Rules Engine (SRE), in which Fuzzy Logic System (FLS) is used as a decision engine, is proposed in [14]. SRE implements the two-stage CWW approach [15] and overcomes the dimensionality problem by using abstraction engine at the first stage, which maps the meaning of stream of data from sensors to the meaning of a few words. The use of two-stage processing of the data flow from the sensors fundamentally distinguishes the SRE from the previously mentioned REs. SRE allows you to move from the classification of the stream of primary raw data from sensors to the classification of the flow of events, represented by the meaning of the words of a high level of abstraction and generalization of data from sensors. The issues of preliminary processing of data from sensors in SRE by the method of abstraction and generalization, as well as the representation of fuzzy decision-making rules in FLS are disclosed in [16, 17]. The purpose of this work is to propose a sequence of events fuzzy classification method which is based on matching the STM Footprint (STMF) with prototype of events flow and to encapsulate it in a two-stage SRE technology. Section 2 begins with foundations of events flow classification problem and continues with a detailed outline of the new algorithm of certainty calculation that STMF belong to the Events Flow Class (EFC). Section 3 will present experiments and their discussion.

2 STMF-Based Events Flow Classification Model 2.1

Problem Formulation

The SCA makes decisions based on the data stream from the L sensors. At the intersection of possible changes in data values from sensors, a set of possible events W = {w1, …, wi, …, wM} is fuzzy defined. An arbitrary set of data changes according to the generalization algorithm of the SRE abstraction engine is mapped into a set of certainties [16, 17] CFw ¼ fcf ðw1 Þ; . . .; cf ðwi Þ; . . .; cf ðwM Þg

ð1Þ

where cf(wi) characterizes the confidence that the event wi has occurred. The numerical characteristic of confidence is a Certainty Factor (CF). The CF is calculated based on the fuzzy characteristics of the event which is represented by a special fuzzy L-R number X : fxjmX ðxÞ; 8x 2 ½1; þ 1g with Gaussian L-R membership function [16]

106

A. Kargin and T. Petrenko

mLX ¼ expððx  aÞ2 =2s2L Þ; 8x 2 ½1; a

ð2Þ

mRX ¼ expððx  aÞ2 =2s2R Þ; 8x 2 ½a; þ 1

the parameters of which are the certainty (−1  a  +1) and the time intervals s = sL + sR (0  s < ∞), where sL and sR are the time intervals since the last data acquisition from the sensor and the data change, respectively. Based on the fuzzy characteristic (2), the CF as a fuzzy numerical characteristic − 1  cf  +1 of the event’s meaning is calculated [17] P cf ¼ a
hðsÞ; where hðsÞ ¼ 1  j


8x2½1;a

mLX ðxÞ þ

P 8x2½a; þ 1

Cardð½1; þ 1Þ

mRX ðxÞ ð3Þ

Let t0, t−1, …, t−(q+1), t−q, … time instants of changes in data values from sensors that generate an events flow. At time t0, the finite sequence of q + 1 events that occurred at this time t0 and earlier at times t−1, …, t−(q+1), t−q is current fragment of events flow. Let N different classes of event flows be allocated for the domain under consideration. To make decisions, the SRE must know with what confidence the current fragment of events flow can be attributed to each of the N classes. The event flow class is specified by the description of its typical representative, that is, the Prl prototype. In the general case, the prototype, as an ordered sequence of events, will be represented as a set of pairs Prl ¼



 ðwi ; t0 Þ; ðwj ; t1 Þ; ðwp ; t2 Þ; . . .; ðws ; tq Þ :

ð4Þ

Prototypes representing different classes can have different lengths (the value of q in (4)). The specific form of representation of the class prototype (4) depends on the model of knowledge representation about the SCA domain. An example of the presentation of specific prototypes will be considered below. The problem of events flow classifying is posed as follows. For moment t0 of localization of an “fresh” event it is necessary to calculate the CFs that the current fragment of events flow belongs to each of the N classes CFPr ¼ fcfPr1 ; cfPr2 ; . . .; cfPrl ; . . .; cfPrN g:

ð5Þ

The values of the CFs cfPrl in (5) are then used as input numerical variables FLS, with the help of which decisions are made in the SRE [17]. There are two possible approaches to calculating the set of CFs (5), differing in the presentation of the current fragment of events. In the first approach, STM is used, the mechanism of which supports storing and processing a fuzzy STMF of the current fragment of events [18]. To calculate the CF cfPrl, the degree of proximity of the STMF to this prototype Prl is found. In the second approach, the calculation cfPrl is carried out directly on the elements of all prototypes, and the current

Fuzzy Classification of the Flow of Events …

107

fragment of events, as such, is not stored and STM is not used for this. The first approach of STMF-based is discussed below. The second approach is highlighted in [16].

2.2

Short-Term Memory Footprint Model

A dynamic discrete flat vector field is taken as a model of STMF (Fig. 1). The xaxis indicates the element of the events set W. The y-axis indicates the r ordinal number of the event in the current fragment of events, and the z-axis indicates the event CF cf(wi). A special discrete vector field is considered. A vector cf(wi, r) is associated with each point of the two-dimensional space (wi, r). A vector can only have two directions along the z-axis: positive and negative (Fig. 1). The numerical model of the vector field is the matrix 0

cf ðw1 ; 0Þ B cf ðw2 ; 0Þ cf ¼ B @ ... cf ðwM ; 0Þ

cf ðw1 ; 1Þ cf ðw2 ; 1Þ ... cf ðwM ; 1Þ

1 . . . cf ðw1 ; rÞ . . . cf ðw1 ; qÞ . . . cf ðw2 ; rÞ . . . cf ðw2 ; qÞ C C: A ... . . . cf ðwM ; rÞ . . . cf ðwM ; qÞ

ð6Þ

Since in the introduced vector field model only a limited set of vectors is used (directed only along the z-axis in the positive or negative direction), then in (6), instead of vectors, scalars are indicated in the form of signed numbers cf(wi, r). In the classification algorithm, the vector field as a footprint of STM is represented by two components: the matrix a = |a(w, r)| and a row vector s = |s(r) = t−r − t−(r+1)|,

Fig. 1 A graphic illustration of a discrete flat vector field

108

A. Kargin and T. Petrenko

Fig. 2 An example of SM routes

which are consistent with matrix (6). The first contains the a and the second s parameters, on the basis of which the CF in (6) are calculated according to (2), (3). A STMF is considered, in which at the event occurrence moment the characteristics values of all previously occurred events are changed. Firstly, each event is shifted “deep into memory” (the parameter r of the event is replaced by r + 1) and, secondly, the vector modulus is reduced by the “aging” or “forgetting” coefficient exp(−j), where 0  j  1 is the blurring rate. This operation simulates the results of studies of memory, obtained in cognitive psychology about the gradual “blur” of event footprint in the memory. Thus, the content of an STM at an arbitrary time represents a time-ordered sequence of events in the form of STMF. All subsequent explanations will be considered on the example of a transport autonomous mobile robots or co-bot, namely Smart Machine (SM), that shifting of goods between automated warehouse zone [4]. SM decides at position F to continue moving either to position H or to position G (Fig. 2). The decision depends on the prehistory: where (from position A or B) and by what route (via position D or E) SM arrived at position F. To decide, the SM must store a sequence of events which reflects movement, receipt of goods, their storage, picking, dispatch and other technological operations at different positions. Let, for the considered example, the sing model of events are symbols denoting the positions of the routes (Fig. 1) W = {A, B, C, D, E, F, G, H}. For example, the meaning of the event B can be formulated as follows: “SM is on the position B, the container is loaded and preparation for shipment along the route is completed now”. Abstraction engine as a component of SRE calculates the meaning of each of these events cf(A), …, cf(H) based on the data stream received from the sensors [16]. If the data are received from all sensors (complete data) and fully correspond to the meaning of the event, then the cf value will be close to +1. If the data from the sensors are complete, but it do not correspond at all to the meaning of the event, then cf will be close to −1. And, if the data are incomplete or noisy then cf will be close to 0.

Fuzzy Classification of the Flow of Events …

2.3

109

Event Flow Classification Algorithm

The place of the Event Flow Classification Algorithm in the general scheme of processing data from sensors when decision making based on FLS is shown in Fig. 3. At each clock cycle of receiving data from the sensors, AE maps data into words of different levels of abstraction. Analysis of the CF values changes of these words gives whether event has occurred at this time or not. If the event is localized, it is stored into STMF and footprint is modified. This is done by the Footprint Blur Algorithm described in detail in [18]. In the next phase of the algorithm, the confidence that the STMF belongs to each class is calculated sequentially on the base of classes prototypes. This phase is indicated by the CFPr calculation block in Fig. 3. At the last phase of the algorithm, these values are transferred to the FLS as input numerical variables for decision making. The CFPr calculation block of the algorithm is discussed below. The CF that the STMF belongs to each of the l = 1, …, N classes are calculated by comparing the STMF with the Prl prototype. The prototype, like a vector field, is represented by two components, namely the matrix Prla = |aPrl(w,r)| and the row vector Prls = |sPrl(r) = q|. In Prla for the events of interest w there are aPrl(w,r) = +1 or aPrl(w,r) = −1, depending on what is required by the problem logic: an event has occurred or not. For events that are not considered at the rth place of the sequence, aPrl(w, r) = 0 is indicated.

Fig. 3 The flow chart of data processing in an SCA using FLS for decision making

110

A. Kargin and T. Petrenko

• Calculation of the first (vertical) projections of the matrix a = |a(w, r)|. For this, a modified operation of matrix convolution was used [19] a1 ¼ ða1 ðrÞ ¼ a0 ðwq ; rÞ
aPrl ðwq ; rÞ; r ¼ 0; 1; . . .; ql Þ; where a0 ðwq ; rÞ ¼ MAX ðjaðwi ; rÞj
jaPrl ðwi ; rÞjÞ

ð7Þ

i¼1;M

• Calculation of CFs cf1 = (cf1(r), r = 0, −1 …, −ql) for each position of the sequence of events of the prototype Prl by (2), (3) based on the values of a1(r) from (7) and s1(r) = |s(r) − sPrl(r)|, where s(r) is taken from the renew footprint; • Calculation of the CF cfPrl that the STMF is of class l. The value is found as the weighted sum of the found CFs cf1 for individual positions of the events sequence cfPrl ¼

X 1 cf 1 ðrÞ: ql r¼1;2;...;q

ð8Þ

l

3 Experimental Study of the Model For the purpose of clarity, the article discusses the results of modeling a simplified example of performing operations in an automated warehouse, which was considered earlier. SM moves through positions B, C, E, F and its service at positions B, E was simulated (see Fig. 2). Numerical simulation results are given for three cases, when SM arrived at positions C, E and F. The STMF, used to classify the events flow when the SM consecutive serves three positions C, E, F are shown in Table 1 in the format for each input event wi. The initial input STM data are cf(wi, 0). In the first example, the values cf(wi, 0) = 0.9 are used for all events wi and at all positions. The influence of the initial data fuzziness is considered in the second and third examples. The following values of STM parameters were used in the simulation: q = 4, j = 0.1. Two classes are considered, represented by prototypes (4) Pr1a = |a(F,0) = +1, a(E, −1) = +1, a(C, −2) = −1, a(B, −3) = +1|, Pr1s = |s(0) = 5, s(−1) = 10, s(−2) = 8|, Pr2a = |a(F,0) = +1, a(D, − 1) = +1, a(C, −2) = −1, a(A, −3) = +1|, Pr2s = |s(0) = 10, s(−1) = 15, s(−2) = 8|. The first prototype describes the situation when SM arrived at position F and was previously loaded at positions B and E and was not loaded at position C. The second prototype describes the subsequence of events when SM was loaded at positions A and D and was not loaded at position C. In Table 1, the Position C column indicates the data that was generated at the time of SM arrival from position B to position C. The first digit in each line in bold is the value of cf(wi, 0), which were calculated based on data from the sensors. At this

Fuzzy Classification of the Flow of Events …

111

Table 1 Modeling Results of the SM movement along the route B, C, E, F wi

Position C

Position E

Position F

A B C D E F G H Pri 1

0; (0, 0, 0, 0) 0; (0.8, 0, 0, 0) −0.9; (0, 0, 0, 0) 0; (0, 0, 0, 0) 0; (0, 0, 0, 0) 0; (0, 0, 0, 0) 0; (0, 0, 0, 0) 0; (0, 0, 0, 0) Step 1 cf1= (0, 0, 0, 0) cfPr1 = 0 cf1= (0, 0, 0, 0) cfPr2 = 0

0; (0, 0, 0, 0) 0; (0, 0.7, 0, 0) 0; (−0.8, 0, 0, 0) 0; (0, 0, 0, 0) 0.9; (0, 0, 0, 0) 0; (0, 0, 0, 0) 0; (0, 0, 0, 0) 0; (0, 0, 0, 0) Step 2 cf1= (0, 0, 0, 0) cfPr1 = 0 cf1= (0, 0, 0, 0) cfPr2 = 0

0; (0, 0, 0, 0) 0; (0, 0, 0.6, 0) 0; (0, −0.7, 0, 0) 0; (0, 0, 0, 0) 0; (0.8, 0, 0, 0) 0.9; (0, 0, 0, 0) 0; (0, 0, 0, 0) 0; (0, 0, 0, 0) Step 3 cf1=(0.8, 0.7, 0.6, 0) cfPr1 = 0.7 cf1= (0, 0.7, 0, 0) cfPr2 = 0.23

2

point, STMF contains data (four numbers in parentheses in each row of Table 1) that previously there was one event associated with loading SM at position B cf(B, −1) = 0.8. At the moment of occurrence of event E, the STMF is represented by the Position E column in Table 1. It reflects that the previous event was C since cf(C, −1) = 0.8, before this event B occurred since cf(B, −2) = 0.7. From this fragment you can see how, as new events appear, the CF of events that happened earlier is blurred (compare cf(B, −1) = 0.8 in the Position C column with cf(B, −2) = 0.7 in the Position E column). In the lower darkened part of Table 1 shows the results of calculating the first projections (7) using the characteristics of the two prototypes Pr1 and Pr2 introduced above. The first projections are calculated as the STM receives information about new events. Table 1, each position contains two first projection vectors cf1 for the first and second prototypes. The results of calculations of the CF of the footprint belonging to the first and second classes of the events flow are also given there. It can be seen that there is not enough data on the current fragment of events at the moment of completion of the SM service at position C for classification, therefore, cfPr1 = cfPr2 = 0 at position C. Similarly, at position E. And when the F event appeared, the STMF was already classified with high confidence cfPr1 = 0.7, cfPr2 = 0.23. These results allow us to decide that the current fragment of events is of the first class, which is true. In the second example, the study of the information incompleteness influence on the results of classification at the same position F as before was carried out. At the same time, a situation was simulated when data from sensors do not allow to unambiguously localize the position at which the SM is. Two cases were considered. In the first case, the incompleteness and fuzziness of the data from the sensors did not allow unambiguously identifying the location of the SM, when it was actually at position E. The SM perception system with certainty cf(E) confirmed that this is position E and with certainty cf(D) that this is position D. In the second case, a similar situation was simulated for positions A and B. The calculation

112

A. Kargin and T. Petrenko

results, like those considered earlier for the first example shows. For example, for cf (E) = 0.3 compared with a high confidence in the false identification of a non-occurred event cf(D) = 0.7, the correct prototype is matched with the current fragment of events cfPr1 = 0.5, cfPr1 = 0.43. Since cfPr1 > cfPr1 this allows to make the correct classification. Another important parameter affects the recognition results is the time interval between events. It, as shown earlier, is set in the prototype Pr1s of the class. In the event flow classification algorithm, it affects the CF cf1 when calculating by formulas (2), (3). In (2), sR is specified as the normalized deviation of the actual interval s(r) from the STMF and the prototype sPrl(r). The degree of influence of the prototype mismatch is controlled by the blurring rate j in (3). At j = 0, the discrepancy between the time intervals is not taken into account at all. At small values of the blurring rate j, significant differences lead to a slight decrease in the CF. Thus, using the blurring rate allows to customize the model for a specific application. The task can be posed when it is not required to take into account the time intervals between events and to recognize simply the sequence of events (j = 0.0). Or, the classification problem may be posed, provided that the time intervals must exactly match the prototype (j = 1.0).

4 Conclusion FLS can be used as powerful decision engine in SCA including when decisions are based on the results of classifying data streams. For this, the SCA should be developed based on the CWW approach in which the FLS is integrated with two stages pre-preprocessing. It is mapping the stream data from sensors to the events in the form of fuzzy characteristics of words of a high level of abstraction and matching the events footprint of STM with class’s prototypes. Inaccuracies obtained at the stage of mapping a stream data into events, inevitably associated with the one-pass constraint, concept drift and massive domain constraint, are largely compensated for by a fuzzy classification algorithm based on a STMF. In the future, it is planned to investigate the dependence of the certainty factor of the decision making by FLS on the STMF parameters, in particular, the memory depth r, the blurring rate, and inaccuracies obtained at the stage of mapping a stream data into events.

References 1. M. Garcia Alvarez, J. Morales, M.-J. Kraak, Integration and exploitation of sensor data in smart cities through event-driven applications. Sensors 19, 1372 (2019) 2. S. Pandey, A. Srivastava, B. Amidan, Real time event detection, classification and localization using synchrophasor data. IEEE Trans. Power Syst. 35(6), 4421–4431 (2020)

Fuzzy Classification of the Flow of Events …

113

3. K. Azadeh, R. Koster, D. Roy, Robotized warehouse systems: developments and research opportunities. SSRN Electron. J. 61 (2017) http://dx.doi.org/10.2139/ssrn.2977779. Accessed 14 Aug 2021 4. White paper. Robots in the warehouse. IAM Robotics, 15 p. (2020), https://info.iamrobotics. com/robots-in-the-warehouse. Accessed 14 Aug 2021 5. C. Calafate, D. Soler, J. Cano, P. Manzonil, Traffic management as a service: the traffic flow pattern classification problem. Math. Probl. Eng. 2015, 14 (2015). http://dx.doi.org/10.1155/ 2015/716598. Accessed 14 Aug 2021 6. Y. Sokha, J. Karpjoo, L. Jonghyun, J. Woojin, A complex event processing system approach to real time road traffic event detection. J. Converg. Inf. Technol. (JCIT) 8(15), 141–148 (2013) 7. M. Yamauchi, Y. Ohsita, M. Murata, K. Ueda, Y. Kato, Anomaly detection in smart home operation from user behaviors and home conditions. IEEE Trans. Consum. Electron. 66(2), 183–192 (2020) 8. S. Birnbach, S. Eberz, I. Martinovic, Peeves: physical event verification in smart homes, in Proceedings of the 2019 ACM SIGSAC Conference on Computer and Communications (2019), pp. 1455–1467 9. S. Kandanaarachchi, R. Hyndman, K. Smith-Miles, Early classification of spatio-temporal events using time-varying models. PLoS One 15(8), e0236331 (2020). https://doi.org/10. 1371/journal.pone.0236331 (Accessed 14 Aug 2021) 10. J. Stefanowski, D. Brzezinski, Stream classification, in Encyclopedia of Machine Learning and Data Mining. ed. by C. Sammut, G. Webb (Springer, Boston, 2016). https://doi.org/10. 1007/978-1-4899-7502-7_908-1 11. C. Aggarwal (ed.), Data Classification Algorithms and Applications, 1st edn. (Chapman and Hall/CRC, New York, 2014) 12. S. Russell, P. Norvig, Artificial Intelligence. A Modern Approach, 3rd edn. (Pearson Education, Upper Saddle River, 2010) 13. White paper. Guide To Rules Engines for the Internet of Things, https://static.waylay.io/ white-papers/011_A-Comparative-Guide-to-Rules-Engines.pdf. Accessed 14 Aug 2021 14. A. Kargin, T. Petrenko, Abstraction as a way of uncertainty representation in smart rules engine, in Proceedings of XIth International Scientific and Practical Conference on Electronics and Information Technologies (ELIT) (IEEE, Lviv, 2019), pp. 136–141 15. L. Zadeh, Computing with Words Principal Concepts and Ideas. Studies in Fuzziness and Soft Computing, vol. 277. (Springer, Heidelberg, 2012). https://doi.org/10.1007/978-3-64227473-2 16. A. Kargin, T. Petrenko, Spatio-temporal data interpretation based on perceptional model, in Advances in Spatio-Temporal Segmentation of Visual Data, ed. by V. Mashtalir, I. Ruban, V. Levashenko. Studies in Computational Intelligence, vol. 876 (Springer, Cham, 2020), pp. 101–159. https://doi.org/10.1007/978-3-030-35480-0_3 17. A. Kargin, T. Petrenko, Multi-level computing with words model to autonomous systems control, in Proceedings of 9th International Conference on Information Control Systems and Technologies (ICST-2020), vol. 2711, ed. by A. Pakštas, T. Hovorushchenko, V. Vychuzhanin, H. Yin, N. Rudnichenko (CEUR Workshop Proceedings Odessa, Ukraine, 2020), pp. 16–30 18. A. Kargin, T. Petrenko, Planning and control method based on fuzzy logic for intelligent machine, in: Proceedings of the 5th International Conference on Computational Linguistics and Intelligent Systems (COLINS 2021), vol. 2870, ed. by N. Sharonova (CEUR Workshop Proceedings, Lviv, 2021), pp. 1716–1730 19. A. Kaufmann, Introduction to the Theory of Fuzzy Subsets, 1st edn. (Academic Press, Cambridge, 1975)

A Decision Tree-Based Model for Tender Evaluation Samuel Kumbu Mandale and Bernard Shibwabo Kasamani

Abstract Subjective tender evaluation and contract award in public procurement is prevalent in various contexts. This has contributed to low quality of goods, services and projects. Successful implementation of building projects is heavily impacted by taking the right decision during tendering processes. Manning tender procedures can be complex and uncertain, involving coordination of numerous tasks and persons with different priorities and objectives. Bias and inconsistent decisions are inevitable if the decision-making process is wholly dependent on intuition, subjective judgement, or emotions. In making transparent decision and beneficial competition tendering, there is need for a flexible tool that could facilitate fair decision making. The purpose of this research was to present a model of an IT solution integrating the concepts of supervised machine learning techniques in the context of tender evaluation in public procurement. Independent variables used as inputs included “Experience”, “Equipment capacity”, “Professionalism”, and “Number of Personnel”. A set criteria was used to determine the values of the variables based on the documents submitted by applicants. The model combines the values of these attributes and determines the category of the entity as either “PASS” or “FAIL”. J48 decision tree classifier was used for this classification problem. This algorithm was preferred due to its relatively simple model among other benefits stated herein. The dataset was divided into test data and training data for the model. The performance appraisal of the model was based on the accuracy of the classification, the precision, recall ratio, ROC curve and the F-Measure. The model was proven to be impressively accurate with an accuracy of 91.1765% while the precision obtained was 0.857. The recall ratio was 1 and an F-measure of 0.923.




Keywords Decision tree Tender evaluation Supervised learning Tendering process





Procurement


S. K. Mandale
B. S. Kasamani (&) Strathmore University, Nairobi, Kenya e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_10

115

116

S. K. Mandale and B. S. Kasamani

1 Introduction Procurement covers the entire process of acquiring and utilizing goods or services. It begins when a department identifies a need and decides on its requirements. On the other hand, Evaluation is defined as the assessment of the strengths and weaknesses of programs, policies, personnel, products and organisations to enhance productiveness. Usually, we should make inclusive evaluation of many contractors in the process of bid evaluation and choose the best one. If unhelpful choice is made, it will affect the schedule of the project and economic efficiency and can even result in the failure of the project and cause great loss. It is therefore of crucial importance to choose a scientific and objective evaluation method [1, 2]. The main objective of a procurement system is to provide value for money by ensuring that funds are utilised in a transparent, efficient and fair manner. In the context of this study, “Public Procurement” means procurement by procuring entities using public funds (Public Procurement and Disposal Act 33, 2015). It is outlined that public money shall be used in a prudent and responsible way. Tendering is an effective contracting method to achieve favourable outcomes for both public and private entities [3]. Tender evaluation is a critical activity in a capital construction project and is normally the accepted means of obtaining a fair price and good value for undertaking construction works. Public entities in many countries usually organise tenders where firms bid for projects supported and financed by the government directly or indirectly. In such cases, there are usually guidelines, manuals, and regulations to enhance the procurement and tendering methods. Various agencies are mandated to ensure that all public entities adhere to the regulations. However, despite all these measures in place, selection of the right suppliers remains a big challenge to most public entities. The technical evaluation stage of a tendering process is a very crucial one. This work is commonly done by an ad hoc team of experts/committee as per the new regulations. Most public entities in Kenya apply paper-based process in the technical evaluation stage. One such public entity is Technical University of Mombasa (TUM) where the case study is based. Commonly, decision makers tend to make decisions founded on a mix of their intuition, subjective judgment which is rooted on past experience and emotions [4]. Such criteria lack consistency and objectivity in the tender evaluation process, negatively affecting the outcome. Subjecting a set criterion for execution by an inherently impartial system addresses the problem of intuition and subjective judgment. There is therefore a need for application of Information Communication Technology (ICT) to standardise bid evaluation. The technical evaluation of construction contractors using a model based on machine learning techniques is the focus of this research study. The objective of supplier selection is to identify suppliers with the greatest potential for meeting a firm’s needs consistently and at an affordable cost. Therefore, contractor selection is a crucial decision that needs to be taken by the client and his representative, in order to ensure that projects are completed within

A Decision Tree-Based Model for Tender Evaluation

117

cost, time and quality standard. When wrong decisions are taken, they can lead to delays, and abandonment of projects [5–7]. Many public entities apply paper-based system in evaluation of bidders and the human factor has compromised the credibility of the process. Poor quality of work or unfinished projects by incompetent contractors due to unreasonable evaluation has been the result. This steers to the research problem addressed by this study, which is the fact that, there is wastage of resources in contractor selection occasioned by unfair technical evaluation process in tendering. A reliable and fast decision tool is needed to assist the decision makers. In this study, a decision tree-based model that will be used to classify bidders into Pass and Fail is proposed. The use of a training data set, which is a set of records, for which we know all feature-attributes (independent variables) and classifying attribute (dependent variable) are fed to a decision tree-based classifier to create a model. It is this model that is used to classify unseen data as Pass and Fail in construction projects. Therefore, the proposed solution is expected to be relatively faster, accurate and fair compared to the current approach. A new challenge.

2 Literature Review The principle established to analyze and evaluate tenders is based on the use of criteria called criteria for awarding contracts [8]. They must be designed so as to be non-discriminatory and linked to the object of the contract. The award criteria generally used for the analysis and evaluation of tenders is based on technical value, quality, profitability, performance, performance in regard to environmental protection [9]. The Simple Method of Analysis and Evaluation of Tenders is the simplest. It gives a note for each tender according to each criterion. Then, for each tender, a weighted sum of these notes obtained according to all criteria is made. This sum represents the total result obtained by the tender. The same is done for the other tenders and total results obtained are compared. The best tender is the one that will have the highest total result [10]. The Comparison Method of Analysis and Evaluation of Tenders makes a correlation of the tenders as per every aspect and gives the best 100% of points and makes a determination of the extent of selection based on different tenders. It ensures the standards of proportionality and fairness in the treatment of tenders [10]. The problem with this method is its relative aspect. Indeed, the tender to which 100% of the points are assigned is better than the others; this does not mean that it is intrinsically a good tender. The prioritization of criteria method of Analysis and Evaluation of Tenders is based on the establishment of hierarchical order among the criteria: it comes to draw up a list in which the criteria are generally in descending order of importance. The use of methods based on the prioritization of criteria must be justified by demonstrating

118

S. K. Mandale and B. S. Kasamani

that it is inappropriate to use the weighting method considering the specificities of contract. The Prioritization method is simpler than weighting method [9]. Weighting methods are preferred to methods of prioritization because they globally evaluate tenders according to all the criteria in order to decide between them. Every company knows precisely the assessment made of its tender according to each criterion [9]. Concerning weighting methods, some are disadvantaged by the non-intrinsic evaluation of tenders and the rest is disadvantaged by the non-relative evaluation of tenders to the others [10, 11]. A good weighting method is the one that will be able to evaluate offers intrinsically and relatively from the other offers.

3 Methodology Rapid Application Development approach was adopted for this study. There are three broad phases to RAD that engage both users and analysis in assessment, design, and implementation. This study is experimental in the sense that the development of the model of interest was done in WEKA environment. WEKA is a set of machine learning software tools and algorithms that is used to build models through experiments with datasets. Evaluation of the performance of the built model was also done within the WEKA environment. Requirements Planning: During requirements planning for the prototype development under rapid application development approach, the objectives and information requirements were identified. This study applied observation, informal and informal interviews as techniques of collecting data or information needed for planning. Prototype Analysis: There are three approaches in information system development section; data-oriented, process-oriented, and object-oriented approaches. The object-oriented method, unlike its two predecessors that lay emphasis either on data or process, combines processes and data into single entities called objects. Object-oriented Analysis (OOA) is the concept used in this research. Related techniques including use-case modelling and class modelling were applied. Prototype Design: Design class diagram was used for general conceptual forming of the software. Entity Relationship Diagram (ERD) was also used, which is a graphic that demonstrates the relationships between objects, places, people concepts or events within a system. It facilitated the outlining of business procedures and to develop relationships between entities and their attributes in a relational database. Prototype Implementation: In this study, Java programming language was used in the NetBeans environment to build the user interface. MySQL was used as the relational database management system for data storage supported on Apache Server. The classifier model was developed in WEKA environment and saved in an application configuration file format for use in the prototype.

A Decision Tree-Based Model for Tender Evaluation

119

Prototype Testing: Usability testing was used to test the functional and non-functional requirements of the system. Usability testing entails testing; validation of communicating components on each screen e.g., text inputs and buttons, validation of navigation flow, Ease of navigation, responsiveness and user friendliness. Prototype demonstrations were done and users were given questionnaire to provide feedback. Research Site: This study focused on Technical University of Mombasa as an example of a public entity. This University is located in the coastal town of Mombasa, Kenya. It was elevated to a fully-fledged Public University in 2013. The University is fast expanding its physical infrastructure to meet the growing demand of higher education in the country. Many construction projects have been undertaken and more are lined up. Most completed projects fell short of the expected quality and standard. Perhaps improving the criteria of evaluating contractors could remedy this trend. Target Population: This study had two types of population as a target to be used in the inquiry. For the development of the model in the WEKA environment, a population of instances was needed. Technical University of Mombasa could only provide 100 instances. Getting more instances would have required extra time as the records were in different physical files. For the purpose of usability test of the built prototype the targeted population was 80, covering all types of stakeholders. Sampling: The sample was taken from the database of all stakeholders. The study used specifically simple random sampling. Lottery method was employed. This sample was of the respondents used in the usability testing. The study picked 70 respondents whom in the opinion of the researchers were representative enough. Data Collection Instruments: Various data collection techniques can be used which include using available information, Observing, interviewing (face-to-face), administering written questionnaires, and focus group discussion. This study applied literature reviews, questionnaires, observations and interviews. In order to build the model, a criteria for evaluating contractors is first defined by experts. Values for the vectors to be used are then calculated/determined. Input examples from database are used for training. The input examples are defined by the set criteria. This data is fed to the J48 algorithm after pre-processing for training and validation. Document`s dataset is fed to the new model for prediction of category class. This model is a binary classification which gives a two-state output of PASS or FAIL. Output of PASS means a bidder moves to the final phase of financial evaluation and the FAIL output means a bidder did not qualify for the next phase. Figure 1 shows general process of the proposed approach for tender evaluation.

120

S. K. Mandale and B. S. Kasamani

Fig. 1 Conceptual design of the proposed model

Fig. 2 Proposed prototype architecture

4 Proposed Solution 4.1

System Architecture

The prototype architecture operates as shown in Fig. 2. During evaluation of tenders the procurement officers through a graphical user interface requests data from the database. The MySQL database has been employed on Apache server. The database responds after which the officer passes the data to the model for evaluation. The model returns results upon which the officer will download or save to the database for the records.

A Decision Tree-Based Model for Tender Evaluation

121

Fig. 3 Prototype context diagram

The Context Level Diagram in presented in Fig. 3. The prototype is available for four categories of users namely: System administrator, Vice-Chancellor, Bidders, and Procurement Officer. The main process is to evaluate bidders’ credentials by classifying them into pass and fail based on set criteria. The System administrator registers and views authorised persons. This creates credentials for authorised persons to log in. The authorised parties can login and view reports of the evaluation process and they include Vice Chancellor and bidders. The procurement Officer does the actual evaluation on the system. They can also view and download reports of the evaluation process.

4.2

Experimental Setup

The experiments were conducted on an Intel Centrino 1.6 GHZ Processor with MySQL 5.0.45, Java, NetBeans, Apache Server and Python. In order to develop the classification model based on J48 algorithm, an experiment was set up in WEKA environment. Real data from database was used in the experiment. The process of building the model took several stages in sequence so as to achieve the best results.

4.3

Dataset Description

In this study, technical evaluation dataset from TUM was used to conduct the experiments. The dataset contained 7 independent variables whose values are either

122

S. K. Mandale and B. S. Kasamani

Table 1 Dataset description Variable

Description

Possible Values

Experience CRB report

Similar successful jobs done Financial integrity

Professionalism Equipment capacity Number of workers Location

Registered with professional bodies In terms of technology and number of equipment Number of personnel able to do the job

>5 yrs,   < ðXiÞ ¼ Yi and wXi ¼ 0 Por QSimTV R Vcategory ; Vprogram ¼ i¼a a w > i¼1 i > Pi¼c else : Pi¼a a

i¼1

wi þ c

i¼1 þ a

wi

ð4Þ Table 2 represents a comparison between QSimTV and QSimTV_R applied on the recommended profile of a child and then an teenager watching television alone. the recommended profile is: Recommended-Profile (child) = {(comedy, 0.23), (animation, 0.3), (superhero, 0.23), (Fantasy, 0.23), (horror, 0)}. TV Recommendation for Multi-users. In what follows, we will present an approach to recommend to a group of people who watch TV at the same time the most suitable programs for their profiles. In the case where children are part of the user group, our QSimTV_R approach favors them in the selection of the contents to be recommended. The solution consists in generating a common profile based on the concept of profiles emerging. From the recommended TV profile for each user, a common profile Ecommon is calculated by applying a formula to have the most relevant common elements of each of them: Ecommon ¼

[i¼n i¼1

felement TV profilei g;

Where: n: number of persons watching TV. The method is as follows: – f a child is present, then filter the Ecommon set on the basis of forbidden-keywords (defined in Sect. 5.1) in order to eliminate words not suitable for children.

Table 2 QSimTV vs QSimTV_R for a child recommendation Movie

Rating

QSimTV

QSimTV_R

Age  13

13 < age  17

Age  13

13 < age  17 ✓

Ratatouille

G

✓

✓

✓

Night at the museum

PG

✓

✓

✓

✓

Thunder Force

PG-13

✓

✓





Suicide Squad: Hell to Pay

R

✓

✓





SAW

NC-17









430

W. Guebli and A. Belkhir

– Unify the vectors of all profiles: we add the elements that do not appear in the recommended-TV profile of each user, and which are in the others recommended-TV profiles. For each added element, the weight -1 is assigned to it. – Generate the Vcommon vector as follows: Vcommon ¼ fðelementk , new weightk Þ; k ¼ jEcommon jg & With: ( new weightk ¼

0 nw n

Pn 

i¼0

n

9ðelementk ,wk Þ 2 recommended TV profilei and wk ¼ 0; i ¼ 1; m wi

; k ¼ 1; m else

ð5Þ Where: m: number of elements in Ecommon . n: number of persons watching TV. wi : the weight of the elementi in the different recommended-TV profiles. nw: number of weight positive and not null. – Select the top nb values form the new Common vector V0common ; where nb ¼ m=n P – Calculate the new weights in such a way that i¼n i¼1 wi ¼ 1 Note that we take in consideration the unlike element (weight = 0), and we add it to the V0 common . we also attribute an age related to the common profile with: agecommon ¼ minðagei Þ; i ¼ 1; n: And a variable adult_present is used to notify if there is an adult in the group or not (1 if there is an adult; 0 else). Below an example of creation of the Vcommon profile of three persons P1, P2, P3 who are 20, 18 and 11 respectively: watching TV at the same time: Vfilm ðP1Þ ¼ fðDaniel Craig, 0:2Þ; ðaction, 0:25Þ; ðhorror, 0:15Þ; ðthriller, 0:2Þ; ðcomedy, 0:2Þ; ðwar, 0Þg Vfilm ðP2Þ ¼ fðDenzel Washington; 0:16Þ; ðdrama; 0:16Þ ðcomedy, 0:14Þ; ðspying, 0:14Þ; ðwar, 0:2Þ; ðhorror, 0.2Þg:

Vfilm ðP3Þ ¼ fðJacki Chan; 0:23Þ; ðcomedy, 0:23Þ; ðanimation, 0:3Þ; ðsuperhero, 0:23Þg:

– Calculate the variables: agecommon ¼ 11 and adult_present = 1 – Create the set Ecommon ¼ f Daniel Craig, Denzel Washington, Jackie Chan, drama; thrillers, comedy, war, horror, action, superhero, animationg. – Filter the Ecommon from each inappropriate word for children. Here, according to the forbidden-words, the word Horror is removed. – Unify the different profiles and select the top nb items (see Table 3)

TV Recommendation for Multiple Users Based on Movie Ratings

431

– Calculate the new common profile V0 common with respecting the constraints in the measure similarity Xi¼nb i¼1

wi ¼ 1 with nb ¼ 4 ðcomedy, animation; Jason Statham and Jacki ChanÞ

– Add the unlike element (war). The new common profile is: V0 common ¼ fðcomedy, 0:69Þ;ðanimation,0:11Þ, (Jason Statham, 0:1Þ;ðJacki Chan,0:1Þ;ðwar,0Þg

with agecommon = 11 and adult present ¼ 1 The Table 3 shows the common profile calculation. The measure of similarity is calculated as follows   8 0 if agecommon  13 and movierating 2 fR; NC  17g or > > > >   < 13\agecommon  17 and movirating 2 fNC  17g or and wXi ¼ 0Þ ðXi ¼ QSimTV R Vcategory ; Vprogram ¼ PYi i¼a > > wi > > Pi¼a a i¼1P else : i¼c a

i¼1

wi þ c

i¼1 þ a

ð6Þ

wi

Table 3 Common Profile for three users Ecommon element

Vaction (P1)

Vaction (P2)

Vaction (P3)

Vcommon

Comedy Animation action Jacki Chan superhero Thrillers Daniel Craig Denzel Washington Drama Spying War

0,2 −1 0,25 −1 −1 0,2 0,16 −1 −1 −1 0

0,14 −1 −1 −1 −1 −1 −1 0,16 0,16 0,14 0,2

0,24 0,3 −1 0.23 0,23 −1 −1 −1 −1 −1 −1

0,193,333,333 0,033,333,333 0,027,777,778 0,025,555,556 0,025,555,556 0,022,222,222 0,017,777,778 0,017,777,778 0,017,777,778 0,015,555,556 0

432

W. Guebli and A. Belkhir

Table 4 QSimTV vs QSimTV_R for multi users’ recommendation Movie Monsters, Inc The Pacifier Rush Hour 3 The Foreigner Hell 1911

Rating G PG PG-13 R NC-17 R

QSimTV Age  13

13 < age

✓ ✓ ✓ ✓ ✓ X

✓ ✓ ✓ ✓ ✓ X

 17

QSimTV_R Age  13 14 < age  17 ✓ ✓ ✓ X X X

✓ ✓ ✓ ✓ X X

The Table 4 shows the difference between the two recommendations approach QSimTV and QSimTV_R for multi users with the presence of at least on child or teenager and one Adult. In the following, the Common profile: Multi-users/agecommon = 11 and adult_present = 1 V0common ¼ fðcomedy, 0:69Þ; ðanimation, 0:11Þ; ðaction, 0:1Þ, ðJacki Chan, 0:1Þ; ðwar, 0Þg

Remark: some movies are not rated; if there is a person under 17 in the room, they are not recommended.

6 Conclusion Even though technology is becoming part of our daily lives, parental controls are still needed to protect children from inappropriate content. We considered that offering a TV recommendation system based only on user preferences was not enough. That is why we proposed QSimTV_R, an extension of the QSimTV recommendation system to orient the media recommendation for children towards age-appropriate content. For future, we will develop the relation between keywords describing movies like genre, sub-genre, casts, movie description…etc. using LOD for a better recommendation.

References 1. W. Guebli, A. Belkhir, TV home-box based IoT for smart home, in Proceedings of the Mediterranean Symposium on Smart City Application (SCAMS ‘17), Article 12 (Association for Computing Machinery, New York 2017), pp. 1–7. https://doi.org/10.1145/3175628. 3175634 2. Motion Picture Association of America, Inc, National Association of Theatre Owners, Inc., eds. Classification and Rating Rules, Effective as revised July 24, 2020 (2020) 3. V.K. Sejwal, M. Abulaish, CAMO: a context-aware movie ontology generated from LOD and movie databases. Multimed. Tools Appl. (2020). https://doi.org/10.1007/s11042-020-10076-4

TV Recommendation for Multiple Users Based on Movie Ratings

433

4. M.A. Hossainn, M.N. Uddin, A neural engine for movie recommendation system, in 4th International Conference on Electrical Engineering and Information & Communication Technology (iCEEiCT) (2018). https://doi.org/10.1109/ceeict.2018.8628128 5. F. Yin, S. Li, M. Ji, Y. Wang, Neural TV program recommendation with label and user dual attention. Appl. Intell (2021). https://doi.org/10.1007/s10489-021-02241-5 6. M.S. Kristoffersen, S.E. Shepstone, Z.H. Tan, The importance of context when recommending TV content: dataset and algorithms, IEEE Trans. Multimed (2020). https://doi.org/10. 1109/TMM.2019.2944214 7. A. Agarwal, S. Das, J. Das, S. Majumder, A Framework for Linear TV Recommendation by Leveraging Implicit Feedback (Springer, Computational Science and Technology, 2019), pp. 155–164 8. M.F. Alhamid, M. Rawashdeh A.E., Saddik, Towards context-aware recommendations of multimedia in an ambient intelligence environment, in 2013 IEEE International Symposium on Multimedia, Anaheim, CA (2013), pp. 409–414 9. R. Nagamanjula, A. Pethalakshmi, Novel scheme for movie recommendation system using user similarity and opinion mining. Int. J. Innov. Technol. Explor. Eng. (IJITEE), 8(4S2) (2019) 10. M. Gupta, A. Thakkar, V. Aashish, Gupta, D.P.S. Rathore, Movie recommender system using collaborative filtering, in 2020 International Conference on Electronics and Sustainable Communication Systems (ICESC) (2020). https://doi.org/10.1109/icesc48915.2020.9155879 11. T. Zhou, L. Chen, J. Shen, Movie recommendation system employing the user-based CF in cloud computing, in 2017 IEEE International Conference on Computational Science and Engineering (CSE) and IEEE International Conference on Embedded and Ubiquitous Computing (EUC) (2017). https://doi.org/10.1109/cse-euc.2017.194 12. J. Zhang, Y. Wang, Z. Yuan, Q. Jin, Personalized real-time movie recommendation system: practical prototype and evaluation. Tsinghua Sci. Technol. 25(2), 180–191 (2020). https://doi. org/10.26599/tst.2018.9010118 13. L. Uyangodage, S. Ahangama, T. Ranasinghe, User profile feature-based approach to address the cold start problem in collaborative filtering for personalized movie recommendation. in Thirteenth International Conference on Digital Information Management (ICDIM) (IEEE, 2018), pp. 24–28 14. http://www.imdb.com/ 15. M.V. Prohászková, The genre of horror. Am. Int. J. Contemp. Res. 2, 132–142 (2012) 16. F. Bauer, M. Kaltenböck, Linked Open Data: the essentials- Climate Knowledge Brokering Edition, 2nd edn. Edition mono/monochrome (2016)

Establishment of a Watch Platform of Public Sustainable Purchase in Morocco Tarik El Haddadi , Mohamed Ben Ahmed, and Taoufik Mourabit

Abstract Companies are evolving in an increasingly complex environment, characterised by frequent severe competitive pressure. Strong technological and economic evolution requires an acceleration of information flows, a transformation of operating methods, as well as an obligation to anticipate, innovate, and then make effective decision. In this article, we propose a business intelligence web portal called XEW 2.0 (Xplor EveryWhere). It acts as an effective system for all the treatments useful to the process of monitoring within client companies. Indeed, monitoring the technological environment of a company is one of the most promising methods that allow us to innovate and promote the development of a company. Keywords Sustainable Public Purchase intelligence XEW 2.0





Business intelligence
Competitive

1 Introduction Faced with the complexity of the technological world, companies need selective, sophisticated and up-to-date scientific and technical information in order to make the most of new techniques and to implement the essential innovations. It is necessary to have useful information for the company, so-called critical information. In a way, we need tools to absorb, structure information and make it usable for the decision maker. An important step in any project is to carry out a preliminary study. This study consists in examining the problem that is to be tackled in order to identify the T. El Haddadi (&)
T. Mourabit Natural Risks Research Team, Faculty of Science and Technology, Abdelmalek Essaadi University, Tangier, Morocco M. Ben Ahmed Computer Science Team, Faculty of Science and Technology, Abdelmalek Essaadi University, Tangier, Morocco © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_35

435

436

T. El Haddadi et al.

Fig. 1 Graphical interfaces of Xplor EveryWhere (XEW 1.0) [2]

failures and the shortcomings of the system. In general, the implementation of a project is due to a problem or lack. It is therefore necessary to study the existing to obtain effective solutions. In order to deepen our understanding of the subject and to have a clearer idea of our project and its expected functions, we conducted a study on two tools “Xplor EveryWhere” and “Tetralogy” which are part of the same framework as our work. “Xplor EveryWhere XEW 1.0”, a Business Intelligence Web portal [1], allows the users of support solutions to continue to seek, monitor, validate and repost strategic information on the move. They no longer need to sit in front of their computer to access useful data in the moment (preparation of a meeting, new agenda, information on a contact, a technology, a market, and an urgent request for an analysis or of a specific focus). Thanks to this Web Portal, they are permanently at the heart of relevant information. (See Fig. 1). “Tetralogy” is a tool well adapted to macroscopic analyses [3], it allows identification of strong signals, weak signals and trends from a corpus of documents collected on a specific subject. The resulting information is a synthesis obtained by various data analysis methods and disseminated via graphic visualizations (map geostrategic, evolutionary graph, PCA, CA, HAC, etc.) (See Fig. 2).

Establishment of a Watch Platform of Public …

437

Fig. 2 Tetralogy graphical interfaces [2]

Following this, the XEW Research Group is seeking, in addition to its existing monitoring services, to set up a “XEW 2.0” web platform of technological monitoring allowing: • Information retreival and in particular the interrogation of databases • Analysis of the corpus of data collected via various data mining and descriptive statistics algorithms… • Interpretation and refinement of the analysis results, which involves building a good tool mapping. Below, the block diagram of our XEW 2.0 system, which consists of three central entities: XEW-SS designating the Big Data collection service, XEW-SBDA which takes care of the analysis of the corpus of collected data and finally the XEW-SBDV providing a dashboard containing a rich palette of graphical visualizations (Fig. 3). Fig. 3 XEW 2.0 project overview

438

T. El Haddadi et al.

2 Critical Study of the Existing XEW 1.0 In this part we will make an objective judgment in order to reduce the shortcomings possible encountered during the study of the existing in order to offer a more reliable system than the old systems seen in the previous section. In Tables below, we present the advantages and disadvantages of the Xplor EveryWhere «XEW 1.0» and Tetralogy systems (See Table 1, Table 2) . This comparative study between these solutions was not random, in addition, it allowed us to establish the detailed characteristics, strengths and weaknesses of each solution.

Table 1 “Xplor EveryWhere «XEW 1.0»” Technical reviews Advantages

Disadvantages

- Provide easy-to-use features and to exploit - Good ergonomics and interactive exploration - Collaborative clear, readable and simple - Access to the system well secured

- Dependent on the old “Tetralogy” system

- Mobility aspect

- Absence of an internal search engine - Graphs still require adjustment - Absence of legends on static and evolving geostrategic maps - No help on the prototype

Table 2 “Tetralogy” technical reviews Advantages

Disadvantages

-

-

Very good adaptation to macroscopic analysis Diversity of analytical methods data Graphic and interactive visualization Diversity of collection sources data Access to the system well secured Mobility aspect

Failure to adapt to microscopic analysis Relatively fast response time Absence of an internal search engine Work process is not entirely automated

Establishment of a Watch Platform of Public …

439

3 New Sustainable Purchase Tools: XEW2.0 To avoid these drawbacks by taking into account the observations and findings, and possibly supplement them with other information which will be studied and discussed within of this article. A simpler and more practical application is needed that includes all tasks, and which use new techniques to facilitate data collection, analysis and visualisation while ensuring information security. Within this context, we will offer three major services which encompass all the fonctionalities of XEW 2.0: Sourcing Service (SS-XEW), Analysis Service (AS-XEW) and Visualisation Service (VS-XEW). • Implementation of a sourcing service (SS-XEW): provides a web editor to configure robots capable of extracting data in real time [4], they constantly scan the designated sources and know the data updates in different formats and languages. In our paper, we were asked to develop and validate an agent for the Sustainable Public Purchase database. • Redesign of a Big Data Analytics service (AS-XEW): The Data wrangling service is a rich library of Big Data Mining algorithm for fine-grained batch, streaming, real time and incremental analysis of the information environment [5], with the possibility of integrating open source analysis tools like Python, R, and other data analysis libraries… • Development of a Big Data Visualization service (VS-XEW): This service provides readable, understandable and interactive graphics that help XEW 2.0 users to better understand the data through an ergonomic and interactive visualisation. It facilitates navigation in the relational environment [6], of a known element (company, research center, researcher, inventor, keyword, etc.). Thus, it becomes possible to find, from this element, all or part of the information related to it (alliances, competitors, teams, emergencies, signals weak, etc.). The SBDV-XEW offers innovative methods for viewing large amounts of data, incremental clustering, time step comparison, temporal graphs, geographical maps, word clouds, etc. The section above detail our solution offers, which we will start with the realization part of our project. But before starting the implementation, first we will explain our conceptual study, which describes the concepts and the trades related to our theme.

3.1

Conceptual Study

After proposing a global solution that includes all the functionalities of XEW 2.0, in this section we begin the design phase. Thus, we focus on the design of an adequate structuring for the application. This step is essential to the smooth running of the project and aims to detail the tasks to be undertaken and to prepare the ground for the implementation phase [7]. In the first part, we start with the overall

440

T. El Haddadi et al.

architecture of our project. Then, in a second part, we will detail the static view, before finishing with a description of the dynamic view of the system. Nevertheless, conceptual study is an important step in the convergence of the notations used in the field of object design analysis since it represents a synthesis for our system [8]. Indeed, it is the initial phase of the creation and implementation of our project. It represents a primordial process of reflection in the software development cycle. During this activity, we will deepen our understanding of the system and refine the description already made in the analysis and specification, indicating the attributes and operations to our design classes.

3.2

Global Architecture

Our project consists of developing the functionalities of XEW 1.0 and proposing a new approach which manages different types of structured, semi-structured and unstructured web data. Thus offers a set of services accessible from the global network “internet” (Fig. 4). It’s a 3-level application, so the architecture is shared between: • A client, equipped with a user interface (usually a web browser), • A set of services, responsible for providing resources but calling on other servers, • Data (Sustainable Public Purchase in JSON format) which are supplied to the application for processing. This global architecture, we would like to explain how to comment on a design model in the light of architecture analysis and software couches. It allows us to move from object analysis to design.

Fig. 4 Overall architecture of the XEW 2.0 system

Establishment of a Watch Platform of Public …

3.3

441

Representation of the Class Diagram

The class diagram shows the internal structure of the system. It provides an abstract representation of the objects of the system that together will interact to realize the use cases [9]. It is a static view on one does not take into account the temporal factor in the behavior of the system. The class diagram established for the whole application aims at technology watch using Open Data. For this purpose, we will just integrate the class diagram of the part aiming at Sustainable Public Purchase in Morocco (See Fig. 5). to properly express the execution context of the relationships between these classes, we will deal in the following section with the transition state diagram which allow to describe the changes of states of an object or a component, in response to interactions (with other objects/components or with actors). A state is characterised by its duration and its stability, it represents an instantaneous conjunction of the values of object’s attributes. The Fig. 6, illustrates the transition state of an instance of the “Sustainable Public Purchase” class.

Fig. 5 Class diagram of the “Sustainable Public Purchase analysis” section

Fig. 6 Transition state diagram for class “Sustainable Public Purchase”

442

T. El Haddadi et al.

Within this section, we present the design phase which aimed to expose in a comprehensive and detailed manner, the functioning of our application in order to facilitate the implementation. To define the framework of our work and to prepare a favorable ground for the next stage, we built the global architecture of our application, then we described its static view and finally we exposed the dynamic view of the system. The next section will be the theme of the implementation phase, the latter focusing on the development of the application as well as the observation of the results obtained while basing on the detailed design of this section.

4 Experiments and Results In keeping with our proposed solution, we have organised the stages of construction of our system according to the three services (SS-XEW, AS-XEW, and VS-XEW) explained in Sect. 3. For the sourcing service, we have developed a search engine capable of collecting the different data sources and storing them in a MongoDB database. In order to test the proper functioning of the developed search engine, we executed a query with the keyword “smart-cites”. The figure below show the search result on XEW search engine (Fig. 7). Regarding the analysis service, we developed all the algorithms and functions for automatic extraction of strategic information available on MongoDB. Among the algorithms we have used, we cite text-mining, correlation calculator, document frequency identification, inventor classification, etc. The results of applying these

Fig. 7 Search results on the XEW search engine

Establishment of a Watch Platform of Public …

443

Fig. 8 Data Analysis of the sustainable public purchase in Morocco

Fig. 9 Keywords Semantic Network

algorithms to the corpus of collected data will be represented graphically in the visualisation service. Finally, the visualisation service allows us to translate the results of statistical analyses and text mining algorithms performed on the documents into a visual presentation (Fig. 8). To illustrate our visualisation, the figure below shows the relationship between the keywords in document source (Fig. 9).

444

T. El Haddadi et al.

5 Conclusion In this paper, we contributed to the establishment of a technology watch platform allowing the automatic analysis of Sustainable Public Purchase documents in Morocco. To carry out this work, we first presented the general context of our article where we presented the importance of anticipating sustainable purchase by sector of activity. Indeed, technological watch sources constitute essential strategic information for anticipating, developing, and exploiting sustainability. Then, we made a state of the art on our research subject. Next we proposed a global architecture of our platform, this conception was used to implement XEW 2.0 portal.

References 1. A. El Haddadi, A. El Haddadi, B. Dousset, A. Fennan, Le système d’intelligence économique XEW. E-TI E-Rev. En Technol. Inf. 1(8), 20–35 (2015) 2. A. El Haddadi, Fouille multidimensionnelle sur les données textuelles visant à extraire les réseaux sociaux et sémantiques pour leur exploitation via la téléphonie mobile, These de doctorat, Toulouse 3 (2011). https://www.theses.fr/2011TOU30175. Accessed 12 Oct. 2021 3. N. Guénec, E. Loubier, I. Ghalamallah, B. Dousset, Management and analysis of Chinese database extracted knowledge (20080 4. C. Biazzin, The role of strategic sourcing on global supply chain competitiveness (2020). https://doi.org/10.4018/978-1-5225-8157-4.ch008 5. B. Quinto, Big data visualization and data wrangling, in Next-Generation Big Data. ed. by B. Quinto (Apress, Berkeley, 2018), pp. 407–476. https://doi.org/10.1007/978-1-4842-3147-0_9 6. G. Andrienko, et al., Big data visualization and analytics: future research challenges and emerging applications (2020) 7. D. Adom, E. Hussein, J. Adu-Agyem, Theoretical and conceptual framework: mandatory ingredients of a quality research. Int. J. Sci. Res. 7, 438–441 (2018) 8. E. Moseholm, M. Fetters, Conceptual models to guide integration during analysis in convergent mixed methods studies. Methodol. Innov. 10, 205979911770311 (2017). https:// doi.org/10.1177/2059799117703118 9. S. Al-Fedaghi, Diagramming the class diagram: toward a unified modeling methodology, ArXiv (2017)

Smart E-Healthcare

Data Encryption for E-Health Service Karima Djouadi and Abdelkader Belkhir

Abstract In recent years, several efforts are being made to integrate IoT technology in various fields, including the health sector. Except that several problems arise, starting with the colossal amount of data generated that require significant computing resources where the need to use a combination of cloud and fog computing to optimize latency which is an important criterion for E-health applications. An even more important issue for such a system is the level of security and the mechanisms applied to ensure the confidentiality of the data as well as the preservation of the privacy of the users, because in a world where personal data is a manageable resource, the security and confidentiality of the data remain aspects of high importance. We have designed a platform for remote medical monitoring of patients by collecting, securing and protecting the end-to-end data transfer via encryption mechanisms, as well as user authentication. Keywords Internet of things (IoT) · E-health · Cloud computing · Fog · Security · Confidentiality

1 Introduction The deployment of infrastructure IoT-based is integrated into all sectors in an attempt to make people’s daily lives easier [1, 23]. One of the main sectors benefiting from this new technology is the health sector. Indeed, the rate of chronically ill people has largely increased in recent years as well as the rate of complicated and sometimes serious pathologies. Furthermore human and material resources dedicated to this sector may be insufficient especially in terms of monitoring patients. [22] The integration of IoT for remote monitoring systems of patients allows the doctor to monitor, treat and better diagnose patients. In e-health services, the amount of medical data collected by the various sensors is enormous and therefore a large storage K. Djouadi (B) · A. Belkhir Department of Computer Science, Computer Systems Laboratory, Department of Computer Science, Computer Systems Laboratory, USTHB BO 32, Bab Ezzouar, Algiers, Algeria e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_36

447

448

K. Djouadi and A. Belkhir

space is required, this is ensured by the cloud computing which provides means for the management and storage of data. [15] Nevertheless, latency and response time are still problems for this type of application that affects human lives and requires real-time processing. To overcome this problem, fog computing has been integrated into the architecture of IoT platforms [21]. However, latency is not the alone problem that this type of service must face. There is the security of medical data subject to several types of attacks that can cost human lives. Furthermore, their confidentiality and integrity remain aspects of high importance. We propose a secure platform of continuous remote monitoring of patients with the integration of medical data encryption mechanisms using RSA and AES protocols at the fog level and a backup procedure at the cloud level which will ensure the availability, traceability, integrity and confidentiality of data, which are, according to [15] the most important requirements to be considered in an e-health system. The remainder of this paper is organized as follows. Section 2 reviews related works on healthcare systems. Section 3 describes the proposed solution. Section 4 presents the experiments. Finally, Sect. 5 concludes the paper.

2 Related Works Several authors have implemented different E-health systems to collect medical data for remote patient monitoring such as [2, 6–8], all built in a way to incorporate a sensor to capture the medical parameters of patients (blood pressure, cardiac output, blood glucose level .....ect). However, all these solutions neglect the notion of privacy of patients and are subject to attacks that could harm human lives strong security mechanisms. While according to [9] ensuring user security and privacy is the most important key to meaningful use of IoT-based E-health applications. In addition, this needs to use high security protocols, which control access and identify identity. Other IoT-based Cloud works have integrated different security mechanisms to provide users with more secure solutions. In [10] authors proposed an IoT-based body sensor network technology utilizing lightweight wireless sensor nodes with IoT built-in to sending and receiving data via cloud computing by considering low security protocols. In [11]authors proposed A Secure IoT-Based Healthcare System with Body Sensor Networks, By ensuring the notion of confidentiality of the data by transmitting them on a secure channel and by ensuring the authentication of the LPU (Local Process Unit) without ensuring the authentication of the users which can harm their private life in case of malignant access to these devices. In [12], a robust secure solution is proposed based on privacy preserving biometrics scheme except that according to [14] several disadvantages arise such as need to trust the cloud service provider, centralized data storage and they estimate that expensive for real scale problems. Another more flexible solution that reduces the complexity of the encryption is proposed in [13], while ensuring an effective privacy preserving biometric identification scheme. However, according to [14] as the health records are extremely Sensitive and the data is exposed to the database owner, this scheme is less

Data Encryption for E-Health Service

449

acceptable in terms of security hence the need to create a secure tunnel between the user and the key generator. According to [15] E-health security system is quite tricky, in a way that the data travels to multiple destinations such as the patient, the hospital, and cloud service provider. Making it harder to keep the data secured at each step of the system, as it must be kept secured and have a safe path throughout the network. Recently, several researchers have also focused on blockchain technology to ensure the security of E-health systems without resorting to a trusted third party and thus avoiding the bottleneck. The previous medical data are also complete and consistent thanks to chaining. In this context, we quote [16] and [17] except that the main disadvantage of this technology is both the enormous computation time it requires and the expensive investment it requires All of these secure solutions mentioned above use cloud computing for data processing and storage, which can produce high latency and very long response times [18]. This is not tolerable in this type of application dedicated to the health sector. E-health is quite sensitive and requires a high quality of service adequate to systems where human lives are at stake [19]. We propose a secure E-health platform to ensure low latency, this by opting for architecture Fog enabled cloud computing allowing reducing the path between the user and the processing node. Our solution respects the security requirements of the e-health system that are adapted to it by ensuring: availability, reliability of the system, authentication of users, confidentiality and integrity [20]. We use an encryption service combining both symmetrical and asymmetrical modes, we also guarantee the rule of the unique identity using the NIN identifier (National Identifier Number) [1, 5].

3 E-Health Platform We present the architecture of our IoT system which is based on fog-enabled cloud computing as described in [3]. Then we will present the encryption protocol used for securing our system.

3.1 System Architecture As illustrated in Fig. 1, the architecture of our system [1] is mainly based on different components which are explained below: Medical Sensors, smartphones (Mobile Application), a fog node, an external server and a central server (cloud). – Medical sensors: it is planted in, on or around a human body to capture the patient’s health data and send it to the system periodically. – Smartphone (IoT equipment): on which a mobile application is installed, it’s used as a User Interface (UI). It authenticates users by submitting their identifiers (NIN) [5] with their relative passwords. It also allows the Patient/sensor association

450

K. Djouadi and A. Belkhir

procedure, which only the practitioner is authorized to do. It allows him to consult the medical records of his patients and make diagnoses. – Fog Node: it is an intermediate between Cloud datacenters and IoT sensors. Its role is to process the medical data collected by the sensors, and their encryption by generating the certificate and the session key, then sending the encryption elements to the external server. The request to generate a patient’s medical file is also processed at its level. – External Server: It is dedicated to the storage of encryption elements. – Cloud: A central server, which is responsible for the global storage of medical and personal data of Users.

Fig. 1 E-Health platform

3.2 Proposed Encryption Protocol Short time and security are essential factors in our system, that’s why we have chosen to combine the two types of encryption by using a session key. To ensure the encryption of the data circulating within our system two main entities are generated first, the digital certificate and the session key. We have generated a certificate specific to Fog to be used when encrypting the session key and the encryption elements via the RSA protocol. This is done by two steps: 1. Creation of a certification authority. 2. Generation of a digital certificate:

Data Encryption for E-Health Service

451

• Creation of an asymmetric key pair and the digital certificate. • Signature of the certificate with the key of the certification authority. The session key is a unique key generated randomly, which is valid for a single communication session. It is encrypted by the RSA protocol using the public key of the Fog. This is done by splitting the session key into two sub-keys allowing them to be encrypted independently due to size limitations. This key is used to encrypt the captured data using symmetric encryption (AES Protocol).

3.2.1

Security of the E-health Platform

After setting up the association of the sensor with the patient [1], the procedure for capturing medical data (heart rate) from the patient begins immediately. These captured data will be sent to the mobile application installed on the Smartphone which will transfer them in turn to the fog node for processing. In order to ensure the confidentiality, integrity and authentication of data exchanged in our system, we have chosen to use the SSL (Secure Socket Layer) protocol for this transfer. As illustrated in Fig. 1, data is encrypted as it travels through our system as follow: 1. Sends captured data in real time via a Bluetooth connection to the Mobile Application. 2. Sends received data to fog via a secure transmission channel using the SSL protocol. 3. Fog Node generates a session key and encrypt the data using the AES protocol and send it to the cloud. 4. Encryption of the session key and encryption files via the RSA protocol using the public key of the fog. Then send them to the external server for later use for decryption.

3.2.2

Use of AES Protocol for Data Encryption

To perform the encryption of the captured medical data, we use AES Protocol. For this purpose we have created a session key to be used as an encryption key, this by deriving it from a randomly generated password using the CSPRNG algorithm; and a hashing algorithm (PBKDF2WithHmacSHA1) [24]. We have also opted for salting (Salt) in order to reinforce the security of the keys and to prevent attacks by cryptanalysis among others: the dictionary attack, Rainbow tables, and frequency analysis. With AES Protocol, we use the CBC (Cipher Block Chaining) mode with an initiation vector (IV), associated with HMAC (an electronic fingerprint authentication code) to avoid attacks based on padding validation.

452

3.2.3

K. Djouadi and A. Belkhir

Encryption/Decryption Scenario

a-Encryption After the reception of the patient’s medical data (rate heart) by the mobile application, a socket connection is established between it and the FOG Node the date and time of the session creation are used as identifiers for the data files. A session key is then generated and the encryption of the data file begins Finally, this file is uploaded to the cloud. Once the encrypted file is uploaded to the cloud, the session key is split into two sub keys that are encrypted and the files containing the salt, password and initialization vector are sent to the external server. When these files are received, the server saves them in a directory dedicated to the communication session. At the end of the communication session, the files are deleted in the Fog. b-Decryption As illustrated in Fig. 2, upon request from the practitioner after logging in via the mobile application [1], for consultation of a file, specifying the date and time of its creation. The Fog looks for this file on the Cloud and downloads it. Then it extracts the session ID from the file name, and sends this ID to the external server to retrieve all the files used during the encryption during this session mentioned above. These files will be decrypted by the private key of the Fog, and the sub-keys concatenated to reconstruct the session key and decrypt the data file.

Fig. 2 Use case of choosing the decrypt file

4 Implementation and Results The experiments were conducted on an Intel Core i5 7th generation processor, 8.0 GB RAM. The system was built using java programming language. The user interface (UI) was built using an Android Studio V4.4+. We have used the cloud storage service offered by the cloud firebase platform (Paas). The objects are represented by files in a format chosen by the user, and they are grouped in hierarchically organized directories. We have chosen to assign a directory to each patient in which the files containing his medical data will be stored. For the capture of medical data we used

Data Encryption for E-Health Service

453

on an IoT sensor Xiaomi Mi Band3 [4], which is an intelligent IoT-based electronic bracelet that incorporates an HR (Heart Rate) heart rate sensor.

4.1 Simulation and Discussion of Results In order to test the performance of our system, and see how it will react to changes in the number of users and the impact of encryption on the transfer time of patient medical files from the fog to the cloud as well as on the sending of files dedicated to encryption to the storage server, we have carried out a Simulation based on a series of tests on the variation of the number of simultaneous users. The results of the 3 simulation tests are shown in the following table (Table 1): Table 1 Simulation’s results Number of users Times to upload data files to the cloud with encryption (Test 1)

1 10 20 30

4.565 s 8.489 s 17.31 s 23.88 s

Time to upload data files to the cloud without encryption (Test 2)

Time to upload encrypted data files to the cloud and transfer the files to the external server (Test 3)

4.363 s 7.757 s 12.34 s 15.89 s

20.67 s 237.6 s 483 s 731.4 s

From the table above we have made the following two graphs, allowing a better interpretation and analysis of these results. Fig. 3 Comparison diagram of test 1 and test 2

454

K. Djouadi and A. Belkhir

The series of three tests was performed by increasing the number of users: 1, 10, 20 and finally 30, this allows seeing the behavior of our system according to this number. Test1: calculation of the transfer time of the medical data of the patients from the fog to the cloud by applying the mechanisms of encryption. Test2: calculation of the transfer time of the medical data of the patients from the fog to the cloud by disabling the option of encryption. Test3: calculation of the transfer time of the encrypted patient medical data to the cloud as well as the transfer time of the encryption files from the fog to the external server.

Fig. 4 Comparison diagram of the three tests

As we can see on Fig. 3, the increase of the number of users, the two curves, take slightly ascending trajectories almost identical for a number of user varying between 1 and 10; while at the end of this figure, curve 1 takes the upper hand over the second one, which means that the influence of encryption on the time needed to transfer data to the cloud increases a little with the increase in the number of users, this does not constitute a major handicap for our system, as long as the latency time does not increase significantly. In Fig. 4 we can see that the two curves relating to test 1 and 2 become identical and straight and remain insignificant compared to that of test 3 which increases in parallel with the increase in the number of users, which is rather acceptable given that the values group together the time of transfer of data to the cloud and the transfer of the encryption files to the external server simultaneously. It remains to say that during our experiments, the policy of management of the file is a FIFO policy, which implies that the first file received will be the first loaded on the cloud, which is simple and practical to achieve, except that in the case of encryption, it remains less effective because of the latency time that can result, which is intolerable in the systems dedicated to the health field. Indeed, we will be obliged to wait for the transfer of all the files necessary for decryption to the storage server, before loading the next data file of the next user.

Data Encryption for E-Health Service

455

Comparison with a Other Systems: Several e-health solutions have been proposed in the literature, aiming to perform remote medical monitoring of patients. Some have implemented security procedures to protect users’ medical data and others do not. In the table below, we have made a comparison between some of them and our approach, in terms of latency and response time, the two key elements of e-health applications, and also in terms of the most important security criteria (privacy, confidentiality, authentication and integrity). We also compared these solutions with respect to the deployment cost and scalability of the approach. From this, we deduce that compared to the works cited in Table 2, our solution respects all the required security properties and presents a low latency and a low response time compared to [10, 12, 16, 17] on the one hand and it is not expensive on the other hand. Table 2 Comparison of our E-health platform with other systems Properties

Approaches Our approach

[2]

[6]

[7]

[8]

Low Latency











Low response time











Privacy

[10]

[11]

[12]

[16]

[17]

×



×

×

×

×



×

×

×



×

×

×

×



×







Confidentiality  of data

×

×

×

×











Authentication  of users

×

×

×

×

×

×







Scalability



×

×

×

×

×

×

×





Integrity



×

×

×

×



×







Expensive

×

×

×

×

×

×

×

×





5 Conclusion We have therefore improved our E-health IoT service platform dedicated to remote patient monitoring, by strengthening security measures, ensuring end-to-end data encryption and providing better privacy protection for patients, while trying to provide acceptable reliability and latency. Although in perspective, it remains to be improved and requires to review the priority management in order to provide a shorter latency and response time, this by giving the priority of treatment to the cardiac data files, so that they will be loaded first, then the transfer of the other files will be done during the collection of data by the sensor.

456

K. Djouadi and A. Belkhir

References 1. K. Djouadi, A. Belkhir, Secure E-health platform, in International Conference on Smart Homes and Health Telematics (Springer, Cham, 2020), pp. 240–248 2. M. Pasha, S.M.W. Shah, Framework for E-health systems in IoT-based environments. Wirel. Commun. Mobile Comput. 2018 (2018) 3. S.S. Gill, R.C. Arya, G.S. Wander, et al., Fog-based smart healthcare as a big data and cloud service for heart patients using IoT, in International Conference on Intelligent Data Communication Technologies and Internet of Things (Springer, Cham, 2018), pp. 1376–1383 4. J. Pino-Ortega, C.D. Gómez-carmona, M. Rico-gonzález, Accuracy of Xiaomi Mi Band 2.0, 3.0 and 4.0 to measure step count and distance for physical activity and healthcare in adults over 65 years. Gait Posture 87, 6–10 (2021) 5. A. Berbar, A. Belkhir, A universal identification code for e-health services, in 2019 Third World Conference on smart Trends in Systems Security and Sustainablity (WorldS4) (IEEE, 2019), pp. 327–332 6. N. Khatoon, S. Roy, P. Pranav, A survey on Applications of Internet of Things in Healthcare, in Internet of Things and Big Data Applications, Intelligent Systems, vol. 180, ed. by N. Khatoon, S. Roy, P. Pranav (Springer, Cham, 2020), pp. 89–106 7. P. Gupta, A. Pandy, P. Akshita, A. Sharma, IoT based healthcare kit for diabetic foot ulcer, in Proceedings of the ICRIC 2019, ed. by P.K. Singh, A.K. Kar, Y. Singh, M.H. Kolekar, S. Tanwar. Lecture Notes in Electrical Engineering, vol. 597 (Springer, Cham, 2019), pp. 15–22 8. B. Farahani, F. Firouzi, K. Charkabarty, Healthcare IoT, in Intelligent Internet of Thing, From Device to Fog and Cloud. ed. by F. Firouzi, K. Chakrabarty, S. Nassif (Springer Nature AG, Cham, 2020), pp. 515–537 9. M. Almulhim, N. Zaman, Proposing secure and lightweight authentication scheme for IoT based E-health applications, in 2018 20th International Conference on Advanced Communication Technology (ICACT) (IEEE, 2018), pp. 481–487 10. A.D. Shewale, S.V. Sankpal, IOT & Raspberry Pi based smart and secure health care system using BSN. Int. J. Res. Appl. Sci. Eng. Technol. 8, 506–510 (2020) 11. K.H. Yeh, A secure IoT-based healthcare system with body sensor networks. IEEE Access 4, 10288–10299 (2016) 12. L. Zhu, C. Zhang, C. Xu, X. Liu, C. Huang, An efficient and privacy preserving biometric identification scheme in cloud computing. IEEE Access 6, 19025–19033 (2018) 13. R. Charanya, M. Aramudhan, Survey on access control issues in cloud computing, in Proceedings International Conference on Emerging Trends in Engineering and Technology (ICETETS) (2016), pp. 1–4, Feb. 2016 14. S. Chenthara, K. Ahmed, H. Wang et al., Security and privacy-preserving challenges of e-health solutions in cloud computing. IEEE Access 7, 74361–74382 (2019) 15. M.A. Kamoona, A.M. Altamimi, Cloud E-health systems: a survey on security challenges and solutions, in 2018 8th International Conference on Computer Science and Information Technology (CSIT) (IEEE, 2018), pp. 189–194 16. P. Ray, D. Dash, K. Salah, N. Kumar, Blockchain for IoT-based healthcare: background, consensus, platforms, and use cases. IEEE Syst. J. 2020, 1–10 (2020) 17. G. Rathee, A. Sharma, H. Saini, R. Kumar, R. Iqbal, A hybrid framework for multimedia data processing in IoT-healthcare using blockchain technology. Multimedia Tools Appl. 2, 1–23 (2019) 18. B. Varghese, W. Blesson, N. Nan, D.S. Nikolopoulos, et al, Feasibility of fog computing, in Handbook of Integration of Cloud Computing, Cyber Physical Systems and Internet of Things (Springer, Cham, 2020), pp. 127–146 19. S. Yi, Z. Qin, Q. Li, Security and privacy issues of fog computing: a survey, in International Conference on Wireless Algorithms, Systems, and Applications (Springer, Cham, 2015), pp. 685–695

Data Encryption for E-Health Service

457

20. R. Sumathi, E. Kirubakaran, SCEHSS: secured cloud based electronic health record storage system with re-encryption at cloud service provider. Int. J. Comput. Commun. Eng. 2(2), 162 (2013) 21. P.H. Vilela, J.J.P.C. Rodrigues, P. Solic et al., Performance evaluation of a fog-assisted IoT solution for e-Health applications. Future Gener. Comput. Syst. 97, 379–386 (2019) 22. W. Zhao, C. Wang, Y. Nakahira, Medical application on internet of things, in IET International Conference on Communication Technology and Application (ICCTA 2011) (IET, 2011), pp. 660–665 23. S.G. Tzafestas, The internet of things: a conceptual guided tour. Eur. J. Adv. Eng. Technol. 5(10), 745–767 (2018) 24. S. Josefsson, PKCS# 5: password-based key derivation function 2 (PBKDF2) test vectors. Internet Engineering Task Force (IETF), RFC Editor, RFC, vol. 6070 (2011)

A Deep Learning Approach for the Diabetic Retinopathy Detection Riad Sebti, Siham Zroug, Laid Kahloul, and Saber Benharzallah

Abstract Diabetic retinopathy is a severe retinal disease that can blur or distort the vision of the patient. It is one of the leading causes of blindness. Early detection of diabetic retinopathy can significantly help in the treatment. The recent development in the field of AI and especially Deep learning provides ambitious solutions that can be exploited to predict, forecast and diagnose several diseases in their early phases. This work aims towards finding an automatic way to classify a given set of retina images in order to detect the diabetic retinopathy. Deep learning concepts have been used with a convolutional neural network (CNN) algorithm to build a multiclassification model that can detect and classify disease levels automatically. In this study, a CNN architecture has been applied with several parameters on a dataset of diabetic retinopathy with different structures. At the current stage of this work, obtained results are highly encouraging. Keywords Healthcare · Diabetes · Diabetic retinopathy · Artificial intelligence · Machine learning · Deep learning

1 Introduction Diabetes is one of the most common diseases in the world and in Algeria. One of its serious complications is diabetic retinopathy [11] which is a severe retinal disease, hence considered a leading cause of blindness in the world. Artificial Intelligence attracts, recently, lot of researchers and investments, thus its applications grow rapidly and touch all human activities. The rapid development of Machine learning (ML) and Deep learning and (DL) algorithms records successful applications in many fields R. Sebti · S. Zroug (B) · L. Kahloul LINFI Laboratory, Computer Science Department, Biskra University, Biskra, Algeria e-mail: [email protected] L. Kahloul e-mail: [email protected] S. Benharzallah LAMIE Laboratory, Batna 2 University, Fesdis, Algeria e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_37

459

460

R. Sebti et al.

[12, 17, 18]. In fact, using AI in healthcare [4] has shown a big success. In this field, AI has the ability to analyse medical data, to serve the better patient, and to help healthcare systems, etc. [15]. Vision and eye health are some of the most crucial organs in the human body. Eye disease such as diabetic retinopathy causes mainly severe damages to the retina. Diagnosing and predicting this Diabetic retinopathy at early stages can increase the chance of curing the patients and restore vision ability. Deep learning convolutional neural networks have proven very effective in many areas such as image recognition applications, including those of diabetic retinopathy detection. In this paper, a deep-learning approach is proposed and applied to detect and classify diabetic retinopathy based on retina images. After this general introduction, this paper is structured as follows. In Sect. 2, some of the previous related works are discussed. Section 3 presents the background of this work namely concepts of Machine learning, Deep learning, Artificial neuron, and convolutional neuronal networks are presented. Section 4 describes system design steps, datasets, pre-processing phase, the proposed CNN architecture, and the evaluation metrics. The results for different experimentation are presented in Sect. 5, and a comparison with other related works is exposed. Finally, Sect. 6 concludes the paper and identifies future perspectives.

2 Related Work In this section, some related work is presented. The authors in [14] proposed an automated detection and classification of fundus diabetic retinopathy (DR) images using a synergic Deep Learning model. They have exploited the Mesidor data set [5], and they achieved 99.28%, 98%, and 99% for the accuracy, sensitivity, and specificity, respectively. The work [6] used a known CNN, which is ResNet34, in their study to classify retina images of a dataset into normal or DR images. ResNet34 is one of the available pre-trained CNN architecture on the ImageNet database. The authors applied a set of image preprocessing techniques to improve the quality of images. The image preprocessing included the Gaussian filter, weighted addition and image normalization. The image number was 35000 images, and the size of the images was (512,512) pixels. The authors of [6] report an accuracy of 85% and a sensitivity of 86%. The authors in [3] propose an automatic system for diabetic retinopathy detection from colour fundus images. The proposed approach is based on the segmentation of blood vessels and the extraction of the geometric features, which are used in the early detection of diabetic retinopathy. The Hessian matrix, ISODATA algorithm and active contour are used for the segmentation of the blood vessels they have used. Finally, they have applied the decision tree CART algorithm to classify images into normal (NO-DR) or DR. The proposed system was tested on the DRIVE (https://drive.grand-challenge.org/) and Messidor (https://www.adcis.net/en/thirdparty/messidor/) datasets. It achieves an average sensitivity, specificity and accuracy of 89%, 99%, and 96%, respectively, for the segmentation of retinal vessels and 91%, 100%, and 93%, respectively, for the classification of diabetic retinopathy. Finally, the work reported in [13] proposes a CNN architecture inspired from the

A DL Approach for DR Detection

461

VGGNet architecture to classify retina images into normal, mild and moderate in the same level, severe and proliferate in the same level. The authors augmented data and resize images to (224,224) in order to improve the performance. In this last work, the authors report an accuracy of 92%.

3 Background Machine learning (ML) [10] is the part of artificial intelligence, which proposes and develops techniques that give the opportunity for a machine to learn, solve problems, and so achieve intelligent tasks. Depending on the nature of the data and the desired outcome of the algorithm, machine learning algorithms can be divided into three types [1]. Supervised learning, in this first type, the algorithm is fed with labelled data. The algorithm will be able to classify new data using a function that maps inputs to the desired outputs. In the second type, i.e., unsupervised learning, the algorithm is fed with unlabelled data. The algorithm will automatically create labels according to the similarities shared between samples; the algorithm will be able to classify new data using a function that maps inputs to the new labels. The third type is Reinforcement Learning. In this third type, it is about the ability of an agent to interact with the environment and find out what is the best outcome. It follows the concept of the hit and trial method. On the other hand, Deep learning (DL) [7] is an approach of machine learning, which deals with the training of neural networks. Deep learning is the most widely used technique in machine learning methods. It is also the most powerful technique in classification, which justifies its presence in almost all applications and fields that use machine learning. Neural network learning type depends on the requirement. However, the most exploited is the supervised one. An artificial neuron [16] is an elementary processor. It takes as input a number of variables X = {x1 , x2 , x3 , .., xn } called the input layer, where each individual input xi is associated to a weight wi representing the value of the connection. An activation function f transforms the weighted sum of the input variables and their weights n (wi × xi) to a value which will then be transmitted to the output layer to be i=1 compared with a threshold value, and then provide an output response. The multilayer artificial neural network consists of several layers of neurons, including the input layer, which is the first layer that receives data, a set of hidden layers where many treatments are done, and finally, the output layer, which produces the output of the network. A multi-layer artificial neural network is an improvement of classical neuronal networks to address more complex problems. n Indeed, an artificial neuron (wi × xi) + bias; this sum computes the sum of the inputs and their weight i=1 can take any value between ] − ∞, ∞[ so that the neuron can know the threshold value for its activation. The activation function is used to introduce a non-linearity in the functioning of the artificial neuron. This function has continuous values, allowing an infinity of possible values included in an interval of [−1,1] or [0, 1]. There are several forms of activation functions like ReLU function, sigmoid function, Softmax function, etc.

462

R. Sebti et al.

Fig. 1 Diagram of steps

Convolutional neural networks (CNN) [8] are a sub-category of multi-layer neural networks, which are specifically designed to process images. The CNN architecture is composed of two main blocks. The first block works as a feature extractor by applying convolution filtering operations to filter the image with several convolution kernels, and returns feature maps. In this first block, two types of layers can be found: convolutional layers and pooling layers. The second block, which is not specific to a CNN, is found at the end of all neural networks used for classification. Thus, the type of layers found in this block is “fully connected layers”.

4 The Proposed System IIn this section, we are interested in explaining the proposed system for diabetic retinopathy detection. We will start by showing the global system design, then describe the used data set, explain the applied preprocessing, the proposed convolutional neuronal architecture, and finally, the used evaluation metrics.

4.1 System Design The global architecture of the proposed system is depicted in Fig. 1. As depicted in Fig. 1, after the preprocessing of the data set, it is divided into three subsets (training, validation and test). The training and validation sets are used in the learning phase to construct the model. The test data set will be used to test the constructed model.

4.2 Datasets Descriptions The main two datasets that have been used in many studies are available in [2] [9]. The first one contains 3662 retina images, and the second one contains 35126 retina

A DL Approach for DR Detection

463

Fig. 2 Images from the dataset

Table 1 The original dataset classification

Level of DR

Class

NO DR Mild Moderate Sever Proliferate

Level 0 Level 1 Level 2 Level 3 Level 4

Table 2 The first proposed classification

Level of DR

Class

NO DR Mild Moderate Sever Proliferate

Class of NO DR Class of Yes DR

images; those images contain diabetic and healthy retina and are ranked into five different levels of diabetic retinopathy by specialists. Both of the datasets can be used in the processes of training, validation, and testing. Examples of images from the datasets are shown in Fig. 2. In this study, we have used the first dataset [2] with three different models: – The Original model: the original diabetic retinopathy datasets contain five classes which are the levels of diabetic retinopathy:(NO DR, Mild DR, Moderate DR, Sever DR, and Proliferate DR) (Table 1). – The first proposed classification (Binary classification): we have grouped all diabetic retinopathy levels (mild, moderate, sever, proliferate) in one class, “Yes DR”. Healthy ones are put in a second class, “NO DR”. Table 2 displays this first proposed model. – The second proposed classification (Multi-class classification): we have grouped diabetic retinopathy levels (mild with moderate) in a first class “Moderate DR”,

464 Table 3 The second proposed classification

R. Sebti et al. Level of DR

Class

NO DR Mild Moderate Sever Proliferate

Class of NO DR Class of Moderate DR Class of Sever DR

Fig. 3 Gaussian filter on retina image

(sever with proliferate) in a second class “Sever DR”, and finally healthy ones in a third class “NO DR”. We have three levels in this proposed model as presented in Table 3.

4.3 Pre-processing In data science, most of the time is spent generally in doing pre-processing. It is the phase that comes before the training. For this purpose, we have resized images to (224 × 224 × 3) and then applied Gaussian filtering on retina images. An example of pre-processing results is shown in Fig. 3.

4.4 The Proposed CNN Architecture In this work, a CNN architecture was proposed to detect diabetic retinopathy using the first structure (i.e., of the data set), then to classify diabetic retinopathy levels using the second structure (i.e., of the data set). The architecture is composed firstly of convolutional layers, each followed by a max pooling layer. Then, a first fully connected layer followed by a drop-out layer, and a second fully connected layer with Softmax as an activation function, are added. In CNN, the main objective is to reduce the amount of loss through weight modification, and this weight modification

A DL Approach for DR Detection

465

is done via the Adam optimization with a learning rate equals to 0.0001. Figure 4 displays the CNN architecture details.

4.5 Evaluation Metrics In order to evaluate the performance of a CNN model, true positives (TP), true negatives (TN), false positives (FP), and false negatives(FN) are used. These values are used to calculate variety of performance metrics including Accuracy, Precision, and F1score according to the three equations: (i) ACCU R ACY = (T P + T N )/(T P + T N + F P + F N ), (ii) P R EC I S I O N = (T P)/(T P + F P), and (iii) F1SC O R E = T P/(T P + 1/2(F P + F N )).

5 Results and Discussion In this section, the results of the proposed system are presented and discussed. Starting with the binary classification then the second multi-class classification is presented.

Fig. 4 The proposed architecture

466

R. Sebti et al.

Fig. 5 (a) Accuracy for the first model, (b) Confusion matrix first model

5.1 The First Classification The proposed model described in Sect. 4 was trained on Kaggle dataset [2] using Google Colab. The model was trained on the first structure (see Sect. 4.2) for 31 epochs before the early stopping. Training accuracy and training validation were found to be 96.93% and 95.08%, respectively. Then, the model was tested on the set of test images, and we have obtained: Accuracy test, Precision test, F1score test to be 95.65%, 95.67%, and 95.65% respectively. Figure 5(b) shows the confusion matrix in which the performance of the proposed model is shown. Row 0 Column 0 = The NO DR correctly diagnosed as NO DR and its value is 175 image out of 181. Row 0 Column 1 = The NO DR incorrectly diagnosed as YES DR and its value is 6 out of 181. Row 1 Column 0 = The YES DR is incorrectly diagnosed as NO DR and its value is 10 out of 187. Row 1 Column 1 = The YES DR is correctly diagnosed as YES DR and its value is 177 out of 187. The confusion matrix confirms that the proposed model is doing good on this first structure.

5.2 The Second Classification The CNN model was trained on the second proposed classification (see Sect. 4.2) for 23 epochs before the early stopping. Training accuracy and training validation were found to be 93.06% and 81.12%, respectively. The model was tested on a set of test images. Test accuracy, Precision test, F1score test were found to be 80.48%, 80.05%, and 80.23%, respectively. Figure 6(b) shows the confusion matrix in which the performance of the proposed model is shown. Row 0 Column 0 = The NO DR correctly diagnosed as NO DR and its value is 177 image out of 181. Row 0 Column 1 and Column 2 = The NO DR incorrectly diagnosed as Moderate DR or Sever DR and the values are 2,2 respectively out of

A DL Approach for DR Detection

467

Fig. 6 (a) Accuracy for the second model, (b) Confusion matrix second model Table 4 Comparison table Accuracy Our model (BC) Our model (MC) ResNet34 [6] CNN Model Inspired from VGGNet [13] Decision tree [3]

Validation accuracy

Dataset organization

Dataset

96.93% 93,06% 85% 92%

95.08% 81,12% – 89%

2 3 2 3

First First Second First

92%

89%

2

–

181. Row 1 Column 0 and Column 2 = The Moderate DR incorrectly diagnosed as NO DR or Sever DR and the values are 9,28 respectively out of 138. Row 1 Column 1 = The Moderate DR correctly diagnosed as Moderate DR and its value is 101 out of 138. Row 2 Column 0 and 1 = The Sever DR incorrectly diagnosed as NO DR or Moderate DR the values are 2,29 respectively out of 50. Row 2 Column 2 = The Sever DR correctly diagnosed as Sever DR its values is 19 out of 50. The confusion matrix confirms that the proposed model is finding difficulties in diagnosing the last class due the lack of data in this class.

5.3 Comparison with Other Related Works To compare the proposed models in this work to the previous works, we have chosen accuracy, validation accuracy and dataset organization as comparison criteria. We have also mentioned which dataset was used in each work (i.e., first dataset [2] or the scecond dataset [9]). The Table 4 illustrates the comparison.

468

R. Sebti et al.

After comparing the proposed architecture with other works, it can be observed that our architecture is doing good in binary classification. But not for the multiclassification due to the lack of data in the last class, which justifies why the accuracy has been decreased in this classification.

6 Conclusion Throughout this study, a deep learning architecture is proposed to detect and classify diabetic retinopathy levels. Indeed, several works and architectures are investigated to deal with images classification and diabetic retinopathy detection. Based on those proposed works, a variant architecture is proposed to improve the existing results. Practically, a dataset with different models is used to train our models and to validate their results. The results have shown that Convolutional Neural Networks (CNN) models can be used in medical diagnosis. Our model showed promising results in terms of classifying using the first dataset [2]. Finally, this work opens the field to several future works, such as improving the model using Deep learning techniques and other preprocessing methods. Acknowledgement This work is realized through the mixed team approved by the ATRSS (Agence Thématique de la Recherche en Science de Santé) on March 17th 2021, under the agreement of the ministry of high education and scientific research of Algeria.

References 1. Y.S. Abu-Mostafa, M. Magdon-Ismail, H.T. Lin, Learning from Data, vol. 4 (AMLBook, New York, 2012) 2. Asia Pacific Tele-Ophthalmology Society (APTOS). Aptos 2019 blindness detection. https:// www.kaggle.com/c/aptos2019-blindness-detection/overview 3. E.Z. Aziza, L.M. El Amine, M. Mohamed, B. Abdelhafid, Decision tree cart algorithm for diabetic retinopathy classification, in 2019 6th International Conference on Image and Signal Processing and their Applications (ISPA) (IEEE, 2019), pp. 1–5 4. T. Davenport, R. Kalakota, The potential for artificial intelligence in healthcare. Future Healthcare J. 6(2), 94 (2019) 5. E. Decencière, X. Zhang, G. Cazuguel, B. Lay, B. Cochener, C. Trone, P. Gain, R. Ordonez, P. Massin, A. Erginay, B. Charton, J.-C. Klein, Feedback on a publicly distributed database: the Messidor database. Image Anal. Stereol. 33(3), 231–234 (2014) 6. M.T. Esfahani, M. Ghaderi, R. Kafiyeh, Classification of diabetic and normal fundus images using new deep learning method. Leonardo Electron J. Pract. Technol. 17(32), 233–248 (2018) 7. I. Goodfellow, Y. Bengio, A. Courville, Deep Learning (MIT Press, Cambridge, 2016) 8. G. Jiuxiang, Z. Wang, J. Kuen, L. Ma, A. Shahroudy, B. Shuai, T. Liu, X. Wang, G. Wang, J. Cai et al., Recent advances in convolutional neural networks. Pattern Recogn. 77, 354–377 (2018) 9. ilovescience. Diabetic retinopathy (resized). https://www.kaggle.com/tanlikesmath/diabeticretinopathy-resized 10. T.M. Mitchell, et al., Machine learning, vol. 45, no. 37 (McGraw Hill, 1977), pp. 870–877

A DL Approach for DR Detection

469

11. M. Porta, F. Bandello, Diabetic retinopathy. Diabetologia 45(12), 1617–1634 (2002) 12. I. Remadna, L.S. Terrissa, S. Ayad, N. Zerhouni, RUL estimation enhancement using hybrid deep learning methods. Int. J. Prognost. Health Manag. 12(1) (2021) 13. M. Shaban, Z. Ogur, A. Mahmoud, A. Switala, A. Shalaby, H. Abu Khalifeh, M. Ghazal, L. Fraiwan, G. Giridharan, H. Sandhu et al., A convolutional neural network for the screening and staging of diabetic retinopathy. Plos One 15(6), e0233514 (2020) 14. K. Shankar, A.R.W. Sait, D. Gupta, S.K. Lakshmanaprabu, A. Khanna, H.M. Pandey, Automated detection and classification of fundus diabetic retinopathy images using synergic deep learning model. Pattern Recognit. Lett. 133, 210–216 (2020) 15. P. Szolovits, R.S. Patil, W.B. Schwartz, Artificial intelligence in medical diagnosis. Ann. Intern. Med. 108(1), 80–87 (1988) 16. B. Yegnanarayana, Artificial neural networks (PHI Learning Pvt. Ltd., 2009) 17. S. Zroug, L. Kahloul, S. Benharzallah, K. Djouani, A hierarchical formal method for performance evaluation of WSNS protocol. Computing 103(6), 1183–1208 (2021) 18. S. Zroug, I. Remadna, L. Kahloul, S. Benharzallah, S.L. Terrissa, Leveraging the power of machine learning for performance evaluation prediction in wireless sensor networks, in 2021 International Conference on Information Technology (ICIT) (IEEE, 2021), pp. 864–869

An Innovative Respiratory Rate Detection System Using Adaptive Filter with Speech Boundaries Detection Algorithm in Audio Signal Ahmet Reşit Kavsaoğlu

and Mohamed Elhashmi

Abstract The adaptive filter is a variation of the digital filtering technique. This form of filter, which does not resemble the classical filtering technique, consists of three basic elements. These elements are collector element, weighting (multiplication) element and a digital filter structure. A system, which has these elements, can make the necessary change in the filter characteristics depending on the environmental media by changing the filter coefficients. Breathing corresponds to the movement of the thorax and lungs and to volume and pressure changes that occur successive in these organs. Respiratory rate (RR) means the respiratory frequency per minute. Since the RR is used to detect and monitor the serious diseases, designing a respiratory rate detection system by means of using adaptive filtering is considered one of the important issues. In this study, a system that detects respiratory rate has been implemented. In this system, there are adaptive filtering, speech boundaries detection algorithm in the sound signal, two stethoscopes whose internal part is placed a microphone and a MATLAB GUI interface design. One of the microphones is placed to upper part of the trachea and the other is placed to upper part of the heart. The sound signals coming from the stethoscope on the heart are used as a noise source and a preprocessing is carried out by means of making free from this noise the sound signals received from the microphone placed on the trachea. After this pre-processing, a clean breathing sound signal is reached by means of making free the heart sounds from the stethoscope placed on the trachea at the adaptive filter output. Inhalation and exhalation time intervals can be determined by running the speech boundaries detection algorithm on this clean breathing sound signal. The respiratory rate is obtained by using these determined time intervals. Keywords Adaptive filtering


Speech boundaries detection
Respiration rate

A. Reşit Kavsaoğlu (&)
M. Elhashmi Karabuk University, Karabuk, Turkey e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_38

471

472

A. Reşit Kavsaoğlu and M. Elhashmi

1 Introduction The main objective of the respiratory system is to take in oxygen and get rid of carbon dioxide (CO2) from the human body. Inhaled oxygen enters the lungs through the airways and reaches the alveoli. The tissues surrounding the alveoli and capillaries have a thickness of only one cell and they are in close contact with each other. The thickness of the layers between air and blood is up to 1 micron (1/10000 cm). Oxygen rapidly passes through these layers into the blood flowing inside the capillaries. In the same way, carbon dioxide (CO2) passes from the blood into the alveoli and is then exhaled. [1]. Respiration occurs spontaneously but can be partially controlled. There is a short period of time between each inhalation and exhalation (one breath per four heartbeats). The amount of air that enters the lungs with each inhalation or leaves the lungs with each exhalation is called the tidal volume (respiratory air). This amount of air is about 500 cm3. At rest, each breath usually takes place at the same depth. Respiration depth is evaluated by observing respiratory movements. The depth of respiration is defined as deep and superficial according to whether the amount of air taken is below or above normal [2]. Observing respiratory system signals is one of the important things that detect many diseases such as sleep disorders and insomnia. The respiratory rate for adults is between 10–20 breaths per minute, while newborns have 44 breaths per minute. Many sounds affect the respiratory signal due to muscle movement and contractions. Adaptive filter is used to eliminate noise [3]. As the ability of digital signal processors increased, adaptive filters became much more common. Digital signal processors are wide used in devices such as mobile phones and other communication devices, video cameras, digital cameras, and medical monitoring equipment. In this study, the system block diagram shown in Fig. 1 is realized. The audio data obtained with a microphone placed into stethoscope is transferred to the MATLAB GUI interface via a USB audio adapter. In the MATLAB GUI interface the reference signal obtained from the microphone placed on the heart, and the noisy breath sound signal obtained from the microphone placed on the trachea are displayed separately in the time domain and frequency domain. After that, an

Fig. 1 System block diagram [4]

An Innovative Respiratory Rate Detection System …

473

adaptive filtering process is applied to these signals, and the noise represented by the heartbeat sound signals is removed, resulting in a clean and noise-free breathing signal. Inhalation and exhalation periods can be determined by using a speech detection algorithm with a filtered breathing signal. Respiratory rate can be obtained using inhalation/exhalation periods.

1.1

Similar Studies in the Scientific Literature

In 2005, Keenan and Wilhelm presented signal processing algorithms that allow the decomposition of the artifact range created by non-respiratory sensor movement and the acquisition of the respiratory component. They evaluated the performance of these techniques for better processing of human respiratory signals. In their work, in contrast to conventional low-pass filters, a soft thresholding technique applied to a set of wavelet filters effectively attenuated noise from the signal without losing breathing information. Their analysis shows that conventional high- and low-pass filters, although designed to reduce noise, can damage latent breathing components in the signal. In contrast, the presented linear and non-linear signal processing algorithms allow for the separation of several critical noises produced by non-respiratory sensor movement [5]. In 2007, Correa et al. proposed three adaptive filter methods based on the least mean squares (LMS) algorithms. The first filter cancels the line interference, the second filter eliminates ECG artifacts, and the third adaptive filter removes EOG spikes. Each of the three stages uses a finite impulse response (FIR) filter that changes its parameters and coefficients to generate an output similar to the artifacts found in the EEG signal. The proposed method was tested on five real EEG signals recordings from polysomnographic (sleep) studies. In all recordings, EOG artifacts, ECG and line frequency were attenuated. It was concluded that the proposed filtering method reduces artifacts found in EEG signals without damaging important information contained in these recordings [6]. In 2011, Pandey et al. presented an LMS-based adaptive filtering method for respiratory artifact suppression to improve the estimation of respiratory indices. The proposed technique uses a reference signal, which is obtained by an ECG. The technique was tested on signals with simulated noise and signals collected from nine healthy people and five people with cardiac disorders. The results of the experiments showed good agreement with the results obtained from Doppler echocardiography [7]. In 2020, Wu et al. used adaptive filtering algorithm to enhance the quality of breathing sounds and to check the accuracy of the Apnea detection algorithm using respiration sounds recorded from trachea. Tracheal sounds were obtained using a microphone placed in a plastic piece. Ambient sounds were obtained using a reference microphone placed outside the plastic case in quiet and noisy environments. Blood pressure signals and motions were considered as standard signals to identify apnea events. After that, NLMS algorithm was applied to respiratory sounds mixed

474

A. Reşit Kavsaoğlu and M. Elhashmi

with noises. Finally, they applied an apnea detection algorithm to tracheal respiratory sounds filtered by adaptive filter. As a result, the use of NLMS AF algorithm for apnea detection in noisy environments has proven to be accurate and reliable [8].

2 Materials and Methods 2.1

Adaptive Filters

A system is considered an adaptive system if it can change its parameters to achieve a defined purpose or goal depending on the state of the system and its environment. Therefore, the system adjusts itself to respond to some events in its environment. Adaptive filters are filters that use an algorithm that analyzes initial input statistics and monitors their changes over time. These filters predict the deterministic signal and remove noise that is not related to it [9]. The adaptive filtering principle is shown in Fig. 2. Figure 3 shows the adaptive noise canceling filtering principle.

2.2

Noise Cancelation Filter

A clean signal corrupted with noise is compared with a reference noise signal that is correlated with the corrupting noise. The adaptive filter solves for the component of the noisy signal that is correlated with the reference signal, and outputs an estimate of the clean signal. Many examples of noise cancellation adaptive filters are found in consumer electronics, such as noise cancelling headphones and mobile phones. This configuration is also common for improving the quality of acquired data in a

Fig. 2 FIR filter structure [10]

An Innovative Respiratory Rate Detection System …

475

Fig. 3 Adaptive noise canceling filter structure [11]

Fig. 4 Noise cancelation with adaptive filter [13]

noisy environment [12]. The principles of noise cancellation can be extended to image processing, as is demonstrated in adaptive mirror telescopes, designed to correct for turbulence in the Earth’s atmosphere. A noise cancellation example is illustrated in Fig. 4.

2.3

Detection Algorithm of Speech Boundaries in Audio Signal

The respiratory detection algorithm is based on Giannakopoulos Algorithm [14], although modified so that the statistics to threshold are short-term energy and spectral spread, instead of spectral centroid and short-term energy. The diagram and steps provide a high-level overview of the algorithm. Figure 5 shows respiratory detection algorithm. This algorithm detects boundaries of breathing signal in an audio signal. This method returns index pairs that correspond to the beginning and end of the speech in the audio signal. Specify audio input as a single or double precision one-channel (column) signal. Breathing indices is returned as an L-by-2 array, where L is the amount of individual speech regions identified [15]. From the flowchart in Fig. 5 the following steps can be understood: In the first step, the audio signal is converted to a time–frequency representation using fast Fourier transform and the specified Hanning window and Overlap the Length. Second, the short-term energy and spectral spread are calculated for each

476

A. Reşit Kavsaoğlu and M. Elhashmi

Fig. 5 The flow chart of speech boundaries detection algorithm [14]

frame. The spectral spread is calculated using (spectralSpread) MATLAB function. Third the histograms are created for both spectral speed distributions and the short-term energy. In fourth step, for each histogram, a threshold is determined using the following equation [15]. T¼

W  M1 þ M2 W þ1

ð1Þ

Where M1 and M2 are the first and second local maxima. In the fifth step both the short-term energy and the spectral spread are smoothed across time by passing

An Innovative Respiratory Rate Detection System …

477

through successive five-element moving median filters. In the sixth step the masks are created by comparing spectral spread and the short-term energy with their respective thresholds. To consider a frame as containing respiratory signal, a feature must be above its threshold [15]. The masks are combined in seventh step. For a frame to be considered as respiratory signal, both the short-term energy and the spectral spread must be above their respective thresholds. In last step the regions declared as respiratory signal are merged if the distance between them is less than Merge Distance [15].

3 Results and Discussion The purpose of including the noise cancellation technique from adaptive filter applications in this study is that when trying to measure the respiratory rate, the overlap of the respiratory signal with other biological signals such as heartbeat signals and some signals from the digestive system may cause some errors in the respiratory rate calculation. The effect of adaptive filtering on noise reduction is demonstrated by an application. The results showed that this goal was successfully achieved using this type of filters. Figure 6 shows the result of a test performed on an adult person using the purposed method. 30 s of breath sounds were recorded and used to calculate the respiratory rate over a full minute.

Fig. 6 A test result obtained using the detection system [4]

478

A. Reşit Kavsaoğlu and M. Elhashmi

Table 1 Finding respiratory rate with inhalation and exhalation records

An Innovative Respiratory Rate Detection System …

479

The clean signal obtained from the adaptive filtering process was used with a speech boundaries detection algorithm to determine the inhalation and exhalation periods. By detecting these periods, it is possible to calculate the rate of inhalation and exhalation by counting them during a certain period of time. Table 1 is shown an application for this. The respiratory rate calculation formula for the volunteer in Table 1 is shown in Eq. 2. Respiratory sounds were recorded for a period of 29.97 s and the result was 7 respiratory cycles (inhalation/exhalation), respiratory rate was calculated for a full minute according to Eq. 2, the resulting respiratory rate was 14.014 respiratory cycles per minute. 29:97seconds 60 second ¼ ; x ¼ 14:014 respiratory cycles 7cycles x

ð2Þ

The results obtained using this method are acceptable and accurate. In addition, the proposed system provides an easy-to-use graphical user interface and does not require much-specialized knowledge to use it.

4 Conclusion In this study, adaptive filtering and speech boundary detection algorithms are used as an innovative method to detect the respiratory rate. The adaptive filter reduced the effect of heart sounds, allowing us to obtain a clean respiratory signal. Thus, the algorithm of finding the sound limit in speech gave good results. In addition to that, the inhalation and exhalation intervals could be determined. Duration of this determined inhalation and exhalation intervals were calculated. As a result of this, the respiratory rate was calculated. All these processes are interacted and managed by an interface.

References 1. R. Dezube, Exchanging oxygen and carbon dioxide. Merck Manuals (2019) 2. Ministry of National Education, Hemşirelik yaşam (vıtal) bulguları. Ankara, 65-s41 (2012) 3. T.D. da Costa, M.D.F.F. Vara, C.S. Cristino, T.Z. Zanella, G.N.N. Neto, P. Nohama, Breathing monitoring and pattern recognition with wearable sensors, Wearable Devices - The Big Wave of Innovation (2019) 4. M. Elhashmi, Respiratory rate detection system design using adaptive filtering. M.Sc. thesis. University of Karabuk department of biomedical engineering, Karabuk (2021) 5. D. Barry Keenan, F.H. Wilhelm, Adaptive and wavelet filtering methods for improving accuracy of respiratory measurement, Tech. Papers of ISA, 455, 37–42 (2005) 6. A.G. Correa, E. Laciar, H.D. Patiño, M.E. Valentinuzzi, Artifact removal from EEG signals using adaptive filters in cascade. J. Phys: Conf. Ser. 90(1), 12081 (2007)

480

A. Reşit Kavsaoğlu and M. Elhashmi

7. V.K. Pandey, P.C. Pandey, N.J. Burkule, L.R. Subramanyan, Adaptive filtering for suppression of respiratory artifact in impedance cardiography, in 2011 Annual International Conference of the IEEE Engineering in Medicine and Biology Society (2011), pp. 7932–7936 8. Y. Wu, J. Liu, B. He, X. Zhang, L. Yu, Adaptive filtering improved apnea detection performance using tracheal sounds in noisy environment: a simulation study. BioMed Research International (2020) 9. R. Koteeshwari, Suhanya, Performance enhancement of adaptive filters using preprocessing technique by wavelet transform. Int. J. Adv. Res. Electr. Electron. Instrum. Eng. 3297 (2007) 10. A.R. Kavsaoğlu, Acoustic echo cancelation with adaptive filtering. M.Sc. thesis. Selçuk University, Konya (2005) 11. L. Tan, J. Jiang, Adaptive filters and applications, Digital Signal Processing (Elsevier, 2019), pp. 421–474 12. J.G. Avalos, J.C. Sanchez, J. Velazquez, Applications of adaptive filtering. Adapt. Filter. Appl. 1, 3–20 (2011) 13. N. Kehtarnavaz, Adaptive filtering. Digital Signal Processing System Design (2008), pp. 157–173 14. T. Giannakopoulos, A method for silence removal and segmentation of speech signals, implemented in Matlab. University Of Athens, Athens, 2 (2009) 15. MathWorks, Detect boundaries of speech in audio signal (2021). https://www.mathworks. com/help/audio/ref/detectspeech.html

Feature Extraction Methods for Predicting the Prevalence of Heart Disease Ivoline C. Ngong

and Nurdan Akhan Baykan

Abstract This paper presents an automatic classification technique for the detection of cardiac arrhythmias from ECG signals. With cardiac arrhythmias being one of the leading causes of death in the world, accurate and early detection of beat abnormalities can significantly reduce mortality rates. ECG signals are vastly used by physicians for diagnosing heart problems and abnormalities as a result of its simplicity and non-invasive nature. The aim of this study is to determine the most accurate combination of feature extraction methods and SVM (Support Vector Machine) kernel classifier that will produce the best results on ECG signals obtained from the MIT-BIH Arrhythmia Database. SVM classifiers with four different kernels (linear, polynomial, radial basis, and sigmoid) were used to classify different features extracted from the four feature selection methods; Random Forests, XGBoost, Principal Component Analysis, and Convolutional Neural Networks. The CNN-SVM classifier produced the best results overall, with the polynomial kernel achieving the maximum accuracy of 99.2%, the best sensitivity 92.40% from the radial basis kernel, and best specificity of 98.92% from the linear kernel. The high classification accuracy obtained is comparable to or even better than other approaches in literature.










Keywords Feature extraction Random Forest (RF) Boosted Trees (BT) Principal Component Analysis (PCA) and Convolutional Neural Networks (CNN) Support Vector Machines (SVM) Heart disease Arrhythmias ECG Classification
















I. C. Ngong (&) University of Vermont, Burlington, VT 05405, USA N. A. Baykan Konya Technical University, 42250 Konya, Turkey e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_39

481

482

I. C. Ngong and N. A. Baykan

1 Introduction An irregular heartbeat is called an “arrhythmia”—when the heart beats too slowly, too fast or irregularly. Some arrhythmias are harmless but if they are particularly abnormal or result from a weak or damaged heart, arrhythmias can cause serious and even potentially fatal symptoms. Coronary artery diseases have become a major cause of death globally. In the United States, annually, an estimated 610,000 people die from different types of heart disease, 370,000 of which die from coronary heart disease which is the most common heart disease [1]. In the United Kingdom, 154,639 people died from cardiovascular diseases in 2014 (25% of deaths in the UK in that year) [2]. It was also reported that in 2016, 3.4 million Turkish adults were living with cardiovascular diseases and the prevalence was projected to increase to 5.4 million by 2035 [3]. The electrical activity of the heart is represented by an ECG signal. ECG signals are measured using an electrocardiogram which uses electrodes placed on the skin to measure a trace of a person’s heartbeat. In the course of a heartbeat, the heart muscles contract and relax causing electrical depolarization and repolarization of these muscles. These events are recorded on the electrocardiogram as deflections on an ECG trace. The different sections of the ECG signals are described by its labels; P, QRS, and T as shown [4] in Fig. 1. The P wave represents atrial contraction that pumps blood to the ventricles, the QRS complex corresponds to ventricular contraction that pumps blood to the lungs and the rest of the body while the T wave represents ventricular repolarization [5]. The prevalence of many cardiovacular risk factors is increasing, especially in developing countries [6]. Therefore, early and accurate detection of disease is essential. Electrocardiogrphy (ECG) is a tool that can be used to study electrical abnormalities in patients with cardiac disease. Certain ECG abnormalities can be

Fig. 1 ECG signal representation

Feature Extraction Methods for Predicting the Prevalence …

483

used to predict adverse events in patients with documented disease, and among those without overt disease [7, 8]. Currently, the best practice for reducing human mortality rates caused by diseases is to detect their symptoms at early stages. Through the early recognition of symptoms one can get the most effective clinical treatment for the best outcome [9]. Previous research has focused primarily on processing ECG signals in order to identify cardiovascular disease in its early stages. Identifying arrhythmias from ECG recordings is therefore important for clinical diagnosis and treatment not just today but for future generations.

2 Related Work Identifying and classifying of ECG signals is a challenging task with a high likelihood of human error during the analysis due to distraction or fatigue. Therefore, computerized systems can be used to process ECG signals more accurately and quickly. Recently, substantial research has been carried out for automatically analyzing ECG signals. According to Karpagachelvi et al. [10] the majority of these studies are based on Fuzzy Logic Methods, Genetic Algorithm, Artificial Neural Networks, Support Vector Machines and other techniques in signal analysis. In their paper, they did a comparative study the different methods proposed by 10 research works in extracting features thereby giving an overview [10]. Ozcift [11] proposed a resampling strategy based Random Forests (RF) ensemble classifier for the diagnosis of cardiac arrhythmias. He used a feature selection algorithm based on correlation to select relevant features and then applied an RF machine learning algorithm to evaluate the performance of the selected features, obtaining a classification accuracy of 90%. In [12] the Cross Wavelet Transform (XWT) that is used to analyze and classify ECG signals achieved an accuracy of 97.6%. Shadmand and Mashoufi [13] trained a Block-Based Neural Network (BBNN) with a Particle Swarm Optimization (PSO) algorithm to classify 5 ECG heartbeats. They obtained a classification accuracy of 97% on the MIT-BIH arrhythmia dataset. Fei et al. [14] also used PSO to train SVM for the classification of arrhythmias. The proposed PSO-SVM method produced an accuracy of 95.65%. Principal Component Analysis (PCA) is one of the main methods used for data reduction and it has also been widely used in the medical studies. In [15], after applying several preprocessing and data normalization methods, the authors performed feature selection by PCA and then implemented 8 different classifiers on various data splits. In the study, a maximum accuracy of 89.74% was achieved with Support Vector Machines (SVM) and Naïve Bayes (NB) classifiers. Similarly, Martis et al. [15] obtained a classification accuracy of 94.52% using a Neural Network (NN) and PCA for feature extraction. Yang et al. [16] showed that applying PCA for feature extraction and a linear SVM on a noisy ECG signal can achieved an accuracy of 97.77% and it can therefore be applied on skewed data. Kamath [17] on the other hand attempted to analyze ECG beats using features

484

I. C. Ngong and N. A. Baykan

extracted from the Teager Energy Operator (TEO). A neural network was then applied on these features achieving 95% accuracy. Li et al. [18] implemented a parallel General Regression Neural Network (GRNN) to classify the heartbeat and obtained a classification accuracy of 95%. A combination of Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) was used in [19] for the diagnosis of cardiac arrhythmia. The proposed model processed variable-length data and achieved an accuracy of 98.10%. Asl et al. [20] presented a classification algorithm that is based on the Generalize Discriminant Analysis (GDA) feature extraction skin and the SVM classifier. Their proposed model achieved an accuracy of 98.94%. Though, some of the approaches mentioned above have produce relatively good accuracies, this performance could be improved. Also, the effect of different kernels in the SVM classifier has not been thoroughly explored and therefore choosing the right kernel for an SVM model can be confusing. The approach proposed in this study, which is an amalgamation of various feature extraction methods and different SVM kernels is developed to address these issues. Therefore, the goal of this paper is to determine the effectives of different SVM kernels in the automatic detection of 5 types of cardiac arrythmias using different feature extraction methods.

3 Proposed Method After the data is acquired, the signals are preprocessed and feature extraction is done. Each of the feature extraction methods is applied and finally the SVM classifier is applied on the dataset with the extracted features to get the results. Hence, the proposed system is composed of 4 major steps for classification as shown in Fig. 2.

Fig. 2 Steps of proposed method

Feature Extraction Methods for Predicting the Prevalence … Table 1 Properties of the dataset used

3.1

Number of samples Number of categories Number of features Sampling frequency Classes

485

109,446 5 187 125 Hz [‘N’: 0, ‘S’: 1, ‘V’: 2, ‘F’: 3, ‘Q’: 4]

Dataset

The MIT-BIH Arrythmia Dataset [21] is used in this study. The dataset contains ECG recordings obtained from 47 subjects gotten between 1975 and 1979 in the BIG Arrhythmia Laboratory. The recordings were digitized at 30 samples per second per channel with 11-bit resolution over a 10 mV range. Also, these recordings were annotated by 2 or more cardiologists. The five types of heartbeats classified according to the Association for the Advancement of Medical Instrumentation (AAMI) recommendation is N (normal beat), S (supraventricular ectopic beats (SVEBs)), V (ventricular ectopic beats (VEBs)), F (fusion beats) and Q (unclassified beats, including paced beats). The dataset’s content has been summarized in Table 1 and Fig. 3 shows sample signals for each class from the dataset.

3.2

ECG signal Preprocessing and Heartbeat Segmentation

Preprocessing is a very important step and should be done carefully as it influences the final results. The dataset used was preprocessed by [22] by first splitting the continuous ECG signal into 10 s windows. A 10 s window from the ECG signal is chosen and the amplitude values normalized to a range between 0 and 1. Following this, the set of all local maximums based on zero-crossings of the first derivative is found, from which the set of ECG R-peak candidates is found by applying a threshold of 0.9 on the normalized value of the local maximums. Subsequently, the median of R-R time intervals are found and for each R-peak a signal part with the length equal to 1.2 T is selected. Finally, each selected part is padded with zeros making it’s length equal to a predefined fixed length. The suggested preprocessing method is simple and effective in extracting the R-R intervals from signals with different morphologies. Moreover, all the extracted beats have identical lengths which are used as inputs to the subsequent processing parts. Amplification and stretching of the signals (was only done in CNN) and feature scaling (Normalization) were other preprocessing techniques that were carried out. Missing values did not pose a problem as no missing values were present in the dataset.

486

I. C. Ngong and N. A. Baykan

Fig. 3 Sample Signals for each class in Dataset

3.3

Feature Extraction/Selection

The feature extraction stage is a very crucial stage in the classification method. The least discriminative features can be found by various greedy feature selection approaches. However, in practice, many features depend on each other or on an underlying unknown variable. Therefore, a single feature can be used to represent a combination of multiple types of information and removing this feature can cause the loss of important information. PCA and a few machine learning algorithms are used as feature selection methods are used in this study: Principal Component Analysis (PCA): PCA is a mathematical procedure that uses linear transformations to map data from high dimensional spaces to low dimensional space [23]. Technically, PCA finds the eigenvectors of a covariance matrix that have the highest eigenvalues and then uses these values to project the data to a new subspace which can either have equal or less dimensions. Practically, it converts a matrix of n features into a new dataset of n or less than n features. It therefore reduces the number of features by building new and smaller number variables which capture a significant portion of the information found in the original features. PCA is carried out following the steps below: • • • •

The scale of the data is normalized and the mean vector of data calculated. The covariance matrix of the data is calculated. The eigenvalue and eigenvector of the covariance matrix is evaluated The principal components are formed using the eigenvectors of the covariance matrix as coefficients of the weights.

PCA may not perform well for all datasets but its low sensitivity to noise, increased efficiency and low memory capacity makes it advantageous to use. In this study, the number components parameter for PCA was set to 0.95 using the scikit-learn library. Therefore, PCA selects a minimum number of principal components such that 95% of the variance is retained. Random Forest (RF): Random Forest (RF) is a machine learning technique based on bagging that works by constructing a multitude of decision trees at training time. A decision tree is a tool that helps determine a course of action from all possible outcomes. In a decision tree structure, a leaf node represents the outcome of the event and the conditions along the path map the outcome to a specific possibility

Feature Extraction Methods for Predicting the Prevalence …

487

[14]. RF was brought forward to solve the some problems found in decision trees such as high variance and bias by computing the average between two boundaries. When implementing RF for classification, the mode of the classes is the final output of the algorithm [24]. It does this by creating a margin function which measures the degree to which the average number of votes for the actual class surpasses the average number of vote for any other class present in the outputs. Hence, in order to classify an example, the input vector is fed and the corresponding output class is got. The class with the most votes is chose as the final output of the random forest (Eq. 1). RF Ouput ¼

n 1X tree outputnth n 1

ð1Þ

Here, n represents each individual tree in the forest. Another advantage of random tress is that it corrects overfitting of the training set which is very common in decision trees. In this study, 100 decision trees have been used. Gradient Boosted Trees (XGBoost): Boosted Tree is an ensemble of classification or regression trees. It sequentially applies weak learners to the incrementally changing data then creates a series of decision trees that produce an ensemble of weak prediction models. Simply put, the boosted tree starts with a small tree and builds other tree models which it adds to the small tree. An observation is assigned a higher weight in the next iteration if it was misclassified. The weighted sum of the decisions made by the trees produces the final classifier. Boosting is a flexible nonlinear regression algorithm that improves the accuracy of trees by giving speed and clarity to the analysis. Gradient boosting tries to reduce these issues by generalizing the tree boosting. Consider a vector of input variables x, output variable y and the training set {(1, y1),…, (xn, yn)}. The goal is to find an approximation F0 ðxÞ to a function FðxÞ that minimizes the target value of a specified loss function Lðy; FðxÞÞ: 0

F ¼

argmin Ex;y ½Lðy; FðxÞÞ F

ð2Þ

In this study, XGBoost which stands for eXtreme gradient boosting was used with a learning rate of 0.1 and 100 decision trees. Convolutional Neural Networks (CNN): CNNs are a class of deep neural networks, a variation of multilayer perceptrons that have been designed to need very little preprocessing. They are regularly applied in analyzing visual images. A major advantage of CNNs is that unlike other image classification algorithms, a CNN based classification system automatically learns feature representation. Hence, no need for prior knowledge or for the filtering to be explicitly programmed which is one of the reasons why CNN can be good feature selector. CNN model used in this study is made up of three 1-dimensional convolutional layers each followed by a

488 Table 2 Size of feature vectors in input/output layers of the CNN model

I. C. Ngong and N. A. Baykan Layer

Input

Output

InputLayer Conv1D (128, 6, 6) RLU + MaxPool(2,2) Conv1D (64, 3, 3) ReLU + MaxPool(2,2) Conv1D (64, 2, 2) ReLU + MaxPool(2,2) Flatten Dense (200) ReLU + DropOut Dense (150) ReLU + DropOut Dense (100) ReLU + DropOut Dense (50) ReLU + DropOut Dense (5) Softmax

(None, 187, 1) (None, 187, 1) (None,182,128) (None,91,128) (None, 89, 64) (None, 45, 64) (None, 44, 64) (None, 22, 64) (1408)

(None, (None, (None, (None, (None, (None, (None, (1408) (200)

(200)

(150)

(150)

(100)

(100)

(50)

(50)

(5)

187, 1) 182, 128) 91,128) 89,64) 45, 64) 44, 64) 22, 64)

maxpooling layer and a ReLU activation function. The number of neurons in each convolutional layer (C) can be represented as C128-C64-C64 with kernel sizes of 6, 3, 2 respectively. The convolutional layers are followed by a flattening layer and 4 hidden dense layers with the ReLU activation function. DropOut is applied to all dense layers to avoid overfitting. The final layer is a dense layer with 5 neurons which represent the 5 classes in the data. Table 2 shows a summary of the CNN architecture and Fig. 4 shows the architecture. Learning/Classification: Support Vector Machines (SVM) has been used for classifying the extracted features in the system. SVM is a supervised learning algorithm that performs classification for 2 linearly separable classes by finding a hyperplane which separates the input space with a maximum margin (Fig. 5). To classify two classes labeled as 0 and 1, SVM distinguishes using Eq. 3. if w:xi þ b  1; if w:xi þ b\0;

yi ¼ 1 yi ¼ 0

ð3Þ

Where w is the weight vector and b the bias. According to the Eq. 1, any result above 1 is classified as belonging to class 1 while in Eq. 2 any value below 0 is classified as belonging to class 0. However, this works only for linearly separable data. The data used in this study is expected to be non-linearly separable; therefore a kernel function (K) is required to transform the problem into a linearly separable

Feature Extraction Methods for Predicting the Prevalence …

489

Fig. 4 CNN architecture

Fig. 5 (a) Non-linearly separable SVM, (b) Linearly separable SVM [27]

space [25]. A SVM classifier implicitly maps an input example X = 〖(x〗_i,… x_n) to a higher dimensional feature space which depends on a nonlinear function u(X). The kernel function is an inner-product between two points in a suitable feature space and is defined by Eq. 4. K ðxi ; yi Þ ¼ uðxÞuðxj Þ

ð4Þ

For a non-linearly separable data with 2 classes, the solution is as shown in Eq. 5 [26]. X f ð xÞ ¼ signð ai yj uð xÞuðxi Þ þ bÞ

ð5Þ

i

Linearly and non-linearly separable SVM are shown in Fig. 5. Several type of kernel functions have been implemented for classification with SVM and the type of kernel used can affect the performance of the SVM classifier.

490

I. C. Ngong and N. A. Baykan

In this study, 4 popular kernel functions was used and compared according to their performance. These kernel functions used in this study is given below. • Linear kernel function (Eq. 6) which is generally used when the data is linearly separable. It was used in experiments with this kernel to evaluate its performance on non-linear data.   K xi ; yj ¼ ðxi Þðyj Þ

ð6Þ

• Radial basis function (RBF) kernel is a general purpose function which is given Eq. 7:  2   K xi ; yj ¼ expðcxi yj  c [ 0

ð7Þ

• Polynomial kernel is defined in Eq. 8.   K xi ; yj ¼ ð1 þ xi :yj Þd

ð8Þ

Where d is the degree of the polynomial. • Sigmoid kernel function is given in Eq. 9   K xi ; yj ¼ tanhðcxi: :yj þ rÞ

ð9Þ

Where c and r are adjustable kernel values which are based on the data.

4 Experimental Results The models are evaluated using a confusion matrix. A confusion matrix can be used to summarize four possible outcomes. A True Positive (TP) and a True Negative (TN) are outcomes where the model correctly predicts the classes as positive and negative, respectively. Similarly, a False Positive (FP) and False Negative (FN) are outcomes where the model incorrectly predicts the positive class as negative or negative class as positive. Sensitivity and specificity are statistical measures of the performance of the binary classification test widely used in medicine and they use these four outputs from the confusion matrix. Sensitivity (True Positive rate) measures the proportion of correctly defined positives in each class (Eq. 10), while Specificity (True Negative rate) measures the proportion of correctly defined negatives in each class (Eq. 11). Similarly, accuracy is also a statistical measure that is the proportion of correct predictions (both true positives and true negatives) among the total number of cases examined (Eq. 12).

Feature Extraction Methods for Predicting the Prevalence …

491

Sensitivity ¼ TP/ðTP þ FNÞ

ð10Þ

Specificity ¼ TN/ðTN þ FPÞ

ð11Þ

Accuracy ¼ ðTN þ TPÞ=ðTN þ TP þ FN þ FPÞ

ð12Þ

Table 3 shows the number of features extracted from each feature extraction algorithm. CNN has the maximum number of features (150 features), followed by RF, XGBoost and PCA. For CNN, the features used in the SVM are extracted from the 2nd hidden layer that contains 150 neurons. The results of the feature extraction methods and SVM kernels are summarized in Table 4. The RBF Kernel attained the best accuracy for all feature extraction methods (the confusion matrices for each of these methods is given in Fig. 6). CNN and SVM with the polynomial kernel achieved the best accuracy of 99.22% while the CNN with RBF kernel produced the best sensitivity of 92.40% and finally the CNN with the Linear kernel produced the best Specificity. The CNN model is therefore the best feature extractor, followed by PCA, then RF and finally XGBoost. The results of the study were also compared with the literature as given in Table 5. Table 3 Number of features extracted

Algorithm

Numbers of features

PCA XGBoost RF CNN

35 41 72 150

Table 4 Comparisons of SVM results according to different kernels and feature extraction methods (%)

SVM

Kernel

PCA

XGBoost

RF

CNN

Accuracy

Linear RBF Polynomial Sigmoid Linear RBF Polynomial Sigmoid Linear RBF Polynomial Sigmoid

90.18 97.28 97.15 78.72 75.89 90.57 89.12 52.48 92.69 97.75 97.54 84.81

90.89 91.63 96.31 80.72 77.54 89.58 87.35 44.67 93.18 97.33 96.72 83.90

92.08 96.97 96.80 83.40 78.53 89.27 85.91 49.67 93.59 97.46 96.58 86.03

99.07 99.13 99.22 99.15 91.90 92.40 91.34 92.20 98.92 98.86 98.76 98.82

Sensitivity

Specificity

492

I. C. Ngong and N. A. Baykan

Fig. 6 Confusion matrices of classification results

Table 5 Comparison with other methods in literature Study

Feature selection

Classifier

Accuracy (%)

[28] [11] [13] [15] [17] [29] [16] [19] [20] Proposed method

PCA Correlation-based Hermit function PCA Teager energy function Time intervals PCA CNN and LSTM GDA CNN CNN CNN CNN

SVM RF NN (Neural Network) NN NN SVM SVM NN SVM SVM (Sigmoid) SVM (Linear) SVM (RBF) SVM (Polynomial)

89.74 90 97.00 94.52 95 95.65 97.77 98.10 98.94 99.15 99.07 99.13 99.22

Feature Extraction Methods for Predicting the Prevalence …

493

5 Conclusions Arrythmia ECG signals and classification was done with SVM. Our study compared the performance of several feature extraction methods and SVM classifiers based on four different kernel functions. Among the feature extraction methods, the CNN model performed better than any other model. The results show that the CNN-SVM model represents an initial version of a potentially useful tool. The performance measures of the classifier for each extraction or selection method were observed in terms of accuracy, sensitivity and specificity. The CNN-SVM classifiers that used the polynomial function produced the best accuracy (99.22%) while the classifier that used the radial basis kernel function achieved the best sensitivity (92.40%). Furthermore, a specificity of 98.92% is obtained from the SVM classifier with the linear kernel. Though these results show promise, there is much room for improvement. In future works, techniques to improve these models’ performances can be applied.

References 1. E.J. Benjamin, P. Muntner, A. Alonso, M.S. Bittencourt, C.W. Callaway, A.P. Carson, A.M. Chamberlain, A.R. Chang, S. Cheng, S.R. Das, Heart disease and stroke statistics—2019 update: a report from the American Heart Association. Circulation 139, e56–e528 (2019) 2. Office for National Statistics, Deaths registered in England and Wales. Off. Natl. Stat. (2015) 3. Y. Balbay, I. Gagnon-Arpin, S. Malhan, M.E. Öksüz, G. Sutherland, A. Dobrescu, G. Villa, G. Ertuğrul, M. Habib, Modeling the burden of cardiovascular disease in Turkey. Anatol. J. Cardiol. 20, 235 (2018). https://doi.org/10.14744/AnatolJCardiol.2018.89106 4. T. Barrella, S. McCandlish, Identifying arrhythmia from electrocardiogram data (2014) 5. H. Montazeri, Hybrid neuro-fractal analysis of ECG signals to predict ischemia (2008) 6. T. Ohira, H. Iso, Cardiovascular disease epidemiology in Asia–an overview. Circ. J. CJ-13 (2013) 7. R. Krittayaphong, A. Maneesai, V. Chaithiraphan, P. Saiviroonporn, O. Chaiphet, S. Udompunturak, Comparison of diagnostic and prognostic value of different electrocardiographic criteria to delayed-enhancement magnetic resonance imaging for healed myocardial infarction. Am. J. Cardiol. 103, 464–470 (2009) 8. P.M. Okin, L. Oikarinen, M. Viitasalo, L. Toivonen, S.E. Kjeldsen, M.S. Nieminen, J.M. Edelman, B. Dahlöf, R.B. Devereux, prognostic value of changes in the electrocardiographic strain pattern during antihypertensive treatment (2009) 9. M.R. Islam, S. Ahmad, K. Hirose, M.K.I. Molla, Data adaptive analysis of ECG signals for cardiovascular disease diagnosis, in Proceedings of 2010 IEEE International Symposium on Circuits and Systems (IEEE, 2010), pp. 2243–2246 10. S. Karpagachelvi, M. Arthanari, M. Sivakumar, ECG feature extraction techniques-a survey approach. arXiv Prepr. arXiv:1005.0957 (2010) 11. A. Özçift, Random forests ensemble classifier trained with data resampling strategy to improve cardiac arrhythmia diagnosis. Comput. Biol. Med. 41, 265–271 (2011) 12. S. Banerjee, M. Mitra, Application of cross wavelet transform for ECG pattern analysis and classification. IEEE Trans. Instrum. Meas. 63, 326–333 (2013)

494

I. C. Ngong and N. A. Baykan

13. S. Shadmand, B. Mashoufi, A new personalized ECG signal classification algorithm using block-based neural network and particle swarm optimization. Biomed. Signal Process. Control. 25, 12–23 (2016) 14. S. Fei, Diagnostic study on arrhythmia cordis based on particle swarm optimization-based support vector machine. Expert Syst. Appl. 37, 6748–6752 (2010) 15. R.J. Martis, U.R. Acharya, C.M. Lim, K.M. Mandana, A.K. Ray, C. Chakraborty, Application of higher order cumulant features for cardiac health diagnosis using ECG signals. Int. J. Neural Syst. 23, 1350014 (2013) 16. W. Yang, Y. Si, D. Wang, B. Guo, Automatic recognition of arrhythmia based on principal component analysis network and linear support vector machine. Comput. Biol. Med. 101, 22– 32 (2018) 17. C. Kamath, ECG beat classification using features extracted from Teager energy functions in time and frequency domains. IET signal Process. 5, 575–581 (2011) 18. P. Li, Y. Wang, J. He, L. Wang, Y. Tian, T. Zhou, T. Li, J. Li, High-performance personalized heartbeat classification model for long-term ECG signal. IEEE Trans. Biomed. Eng. 64, 78–86 (2016) 19. S.L. Oh, E.Y.K. Ng, R. San Tan, U.R. Acharya, Automated diagnosis of arrhythmia using combination of CNN and LSTM techniques with variable length heart beats. Comput. Biol. Med. 102, 278–287 (2018) 20. B.M. Asl, S.K. Setarehdan, M. Mohebbi, Support vector machine-based arrhythmia classification using reduced features of heart rate variability signal. Artif. Intell. Med. 44, 51–64 (2008) 21. O. Paiva, Helping radiologists to help people in more than 100 countries! coronavirus cases (2020) 22. M. Kachuee, S. Fazeli, M. Sarrafzadeh, ECG heartbeat classification: a deep transferable representation, in 2018 IEEE International Conference on Healthcare Informatics (ICHI) (IEEE, 2018), pp. 443–444 23. H. Abdi, L.J. Williams, Principal component analysis. Wiley Interdiscip. Rev. Comput. Stat. 2, 433–459 (2010) 24. Wikipedia contributors, Random forest. Wikipedia, Free Encycl. 25. R. Kybartas, N.A. Baykan, N. Yilmaz, S. Raudys, Multiclass mineral recognition using similarity features and ensembles of pair-wise classifiers, in International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems (Springer, 2010), pp. 47–56 26. K.S. Bayram, M.A. Kizrak, B. Bolat, Classification of EEG signals by using support vector machines, in 2013 IEEE INISTA. (IEEE, 2013), pp. 1–3 27. Wikipedia contributors, Support-vector machine. https://en.wikipedia.org/w/index.php?title= Support-vector_machine&oldid=1033188195 28. Pandey, S.K., Sodum, V.R., Janghel, R.R., Raj, A.: ECG Arrhythmia Detection with Machine Learning Algorithms. In: Data Engineering and Communication Technology. pp. 409–417. Springer (2020) 29. F. Shan, Y. Gao, J. Wang, W. Shi, N. Shi, M. Han, Z. Xue, Y. Shi, Lung infection quantification of Covid-19 in CT images with deep learning. arXiv Prepr. arXiv:2003.04655 (2020)

Quality Attributes for Evaluating IoT Healthcare Systems Loubna Chhiba, Abdelaziz Marzak, and Mustapha Sidqui

Abstract The Internet-of-Things (IoT) has taken over the business spectrum and its applications vary widely from agriculture, and healthcare, to transportation etc. A hospital environment can be very stressful, especially for senior citizens and children. With the ever-increasing world population, the conventional patient-doctor appointment has lost its effectiveness. Hence smart healthcare becomes very important. Smart healthcare can be implemented at all levels, starting from temperature monitoring for babies to monitoring vital signs in the elderly. The IoT healthcare application is a complex fusion of a variety of technologies such as wireless network, embedded, sensor and connectivity. This diversity leads to a variety of quality measurement model, which makes the process of measuring quality more challengeable, less accurate, and less applicable. In this research, different quality models for IoT systems have been studied and compared regarding the quality factors. Besides, we discussed the importance, requirements and applications of smart healthcare and we defined quality attributes for evaluating IoT healthcare applications by considering the impacts of the identified characteristics on the quality of IoT applications. Keywords Internet of Things evaluation Software quality





Smart healthcare
ISO/IEC 25,000
Quality

1 Introduction Smart healthcare empowers users to self-manage some emergency situations. It provides an emphasis on improving the quality and experience of the user. Smart healthcare helps in utilizing available resources to their maximum potential. It aids L. Chhiba (&)
A. Marzak Faculty of Science Ben M’sik, Laboratory of Technology of Information and Modelling, Hassan II University, Casablanca, Morocco M. Sidqui Faculty of Dental Medicine of Casablanca, Hassan II University, Casablanca, Morocco © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_40

495

496

L. Chhiba et al.

remote monitoring of patients and helps in reducing the cost of the treatment for the user. It also helps medical practitioners to extend their services without any geographical barriers. With an increasing trend towards smart cities, an effective smart healthcare system assures a healthy living for its citizens. IoT creates more flexible environments that are interconnected in a way that makes all processes and activities easier, despite all this; however, there are some challenges that may not seem straightforward to the user, and quality is one of them. Quality control is applied through the use of quality models, which are universally recognized models and standards so that each model contains a set of characteristics and sub-characteristics that are set according to the system. Therefore, the method of selecting the model is very important according to the type of system in order to verify that the system for user needs plus the expected things of the system, and the model used varies depending on the nature of the system. IoT healthcare because it contains different things in the nature of its composition such as sensors, devices, humans, animals and other things that can be linked to each other through the Internet, Therefore, the measurement method differs from other systems, so it must be dealt with in particular to avoid errors and unsatisfactory results through our research we proposed the work of a special model of the IoT to include all the important his characteristics. The rest of the paper is structured as follows: Related work in Sect. 2, requirements of smart healthcare is defined, in Sect. 3, characteristics required for smart healthcare system and discussion are presented, in Sect. 4, Comparison of IoT-Healthcare Applications, in Sect. 5 Finally, a conclusion and future works are given in Sect. 6.

2 Related Work After the study we did on research on the Internet of things and the use of quality models we concluded that in all these researches the ISO model was used, so that we did not find a specific model only for the internet of things, Specifically, it was used ISO/IEC 25,010 Which was made after the development and modification of ISO/IEC 9126. “Product quality model defined in ISO 25010 comprises eight quality characteristics: Functional Suitability, Reliability, Performance Efficiency, Usability, Security, Compatibility, Maintainability and Portability, and 31 sub-characteristics. Compared to ISO 9126, ISO 25010 is more comprehensive and complete. ISO 9126 provides 6 characteristics and 27 sub-characteristics, while ISO 25010 provides 8 characteristics and 31sub characteristics.” ISO 9126 has some limitations due to its generic nature. Some concepts presented in ISO 9126 need to be refined before they can be actually applied in a real project. In addition, elements of software metrics were not clear when defining the standard. New characteristics were inserted in ISO 25010 such as security and compatibility. Both characteristics

Quality Attributes for Evaluating IoT Healthcare Systems

497

were not defined in ISO 9126. In addition, the hierarchy of characteristics and sub characteristics was reorganized with the purpose of improving understanding of related concepts. To evaluate the quality of a software product, a quality model is required, and it defines quality goals for intermediate and final software products. An example of a product quality model is described in ISO/IEC 20,010; a possible strategy is the use of measures. According to ISO/IEC 25,010, a measure is the mapping of an entity to a number or a symbol in order to characterize an entity property. Measures can be part of a quality model. According to the SQuaRE standard, a quality model categorizes software quality into characteristics, which are subdivided into sub-characteristics and quality attributes [1]. In Table 1, all previous studies have used ISO 25010, Based on the studies we have compared with each other in terms of the model used, all the characteristics covered in each study were identified using the Quality model. Table 1 shows us in detail the reference and the characteristics that have been used in its quality model. In addition, Table 1 compares IoT applications in terms of the properties covered in each application and 20 characteristics selected for that comparison.

3 Comparative Smart Healthcare Architectures: Requirements and Components Requirements of smart healthcare can be broadly classified into functional requirements and non-functional requirements, as shown in Fig. 1. Functional requirements address specific requirements of a smart healthcare architecture. For example, if a temperature monitoring system is deployed, based on the application it is used for, the range of operation of the thermistor/thermometer, data collection mechanism, and frequency of operation might vary. Hence functional requirements are specific to each component used in that healthcare system based on their application. On the other hand, non-functional requirements are not very specific. Non-functional requirements refer to attributes based on which the quality of the healthcare system can be determined. On a broader perspective, non-functional requirements of smart healthcare can be classified into performance requirements and ethical requirements. Due to the large number of verticals involved in designing a complete smart healthcare system, performance requirements can be further classified into software and hardware requirements. Essential requirements for an efficient smart healthcare system are low power, small form factor, system reliability, quality of service, enriched user experience, higher efficiency, ability to interoperate across different platforms, ease of deployment, and popularity of the smart healthcare system to offer continuous support, scalability of the system to upgrade to newer versions and technologies, and ample connectivity since the very

G. M. Foody et al. 2013 [1]

✓

✓

Johan J.e. Tambotoh et al. 2016 [33]

Dr. Jay Kiruthika et al. ✓ ✓

✓

✓

2018 [20]

2017 [19]

✓

✓

✓

Gary White et al.

✓

✓

✓

✓

Emanuel Coutinho1 et al.

2015 [34]

✓

✓

A. E. Al-Fagih et al. 2013 [18]

L.R.Dominguez, 2014 [17]

2014 [16]

S. Lanziser et al.

2014 [15]

B. Kantarci et al.

2009 [14]

V.A.Reguera et al.

Y. Zhang et al. 2014 [13]

[12]

T.H.Szymanski, 2013

D. Wenrui et al. 2008 [11]

✓

✓

✓

✓

Compatibility

Correctness

S. Chen et al. 2014 [10]

✓

✓

✓

Effect iveness

✓

✓

✓

✓

Exten dibility

X. Mao et al., 2011 [9]

Jin Zhou et al. 2013 [8]

A. Zanella et al. 2014 [7]

F. Li, 2013 [6]

2014 [5]

S. G. Dellepiane et al.

2014 [4]

J. Kiljander et al.

R. Bose, 2013 [3]

2013 [2]

L. Hu, Z. Zhang et al.

Perform ance

✓

✓

Timeliness

✓

Reli ability

✓

Integrity

✓

Reusability

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

Robustness

✓

✓

Scalability

✓

✓

Security

✓

Cost efficient

Table 1 Analysis of quality attributes in the related work

✓

✓

✓

✓

✓

✓

✓

✓

✓

Maintainability

✓

✓

✓

✓

✓

✓

✓

✓

Usability

✓

✓

✓

✓

✓

✓

✓

✓

Effici ency

✓

✓

✓

✓

Porta bility

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

✓

Functi onality

✓

Verifi cation

✓

✓

Validation

498 L. Chhiba et al.

Quality Attributes for Evaluating IoT Healthcare Systems

499

prime motive of designing a smart healthcare is to ensure medical service promptly. In advanced applications, along with these requirements, the system also needs to have ambient intelligence to improve the quality of service. Figure 1 demonstrates the applications of smart healthcare, which start from fitness monitoring on one end of the spectrum to vital sign monitoring in hospitals. Based on the application, the quality of health care systems is improved with additional machine learning algorithms and artificial intelligence. The wide range of applications can be grouped into inter-body sensing, intra-body sensing and environmental management. Intra body sensing applications refer to those which help in monitoring multiple vital signs. Temperature sensors, ECG, blood pressure, blood glucose, EMG, heart rate, SpO2, gyroscope, motion sensors, and accelerometers, are the common sensors used in smart healthcare. Computing devices used in the present era range from smart phones, tablets, and PDAs to complex and advanced devices such as super computers and servers. Memories play a very important role in smart healthcare since storing the information is the most important function of these systems. Data storage components in the smart heath care network cover a broader spectrum starting from embedded memory on the sensing devices to big servers that are used to handle big data analytics. Networking components vary from link sensors to routers and base stations. Based on the severity of the problem addressed, the sophistication of the components varies. Wireless technologies are the backbone of the smart healthcare network.

4 Characteristics of IoT Healthcare and Discussion The most important characteristics required for smart healthcare system are shown in Fig. 2. Characteristics of smart healthcare can be broadly classified based on three categories: App-oriented, Things-oriented and Semantics oriented. App-oriented architectures need to ensure reliable transmission between the applications in smart phones and the sensors, establish a personalized network between the sensors and the user’s computing device and secure the information. Things-oriented architectures need to be adaptive based on the application, real time monitoring, on-time delivery, higher sensitivity, maintain higher efficiency at lower power dissipation, and embark on intelligent processing. Semantic-oriented systems need to be able to develop behavioral patterns based on the previously acquired information, process natural language processing techniques to enrich user experience and have ubiquitous computing capabilities. Adding to this list, other significant characteristics include heterogeneous computing, Interoperability, scalability, data confidentiality, dynamic networks which can accommodate a large number of devices as required, and resource constrained computing with higher efficiency. The new set of characteristics added is

500

L. Chhiba et al.

considering some features that is are special for IoT systems. The new set of characteristics focus on the nature of the IoT. Where IoT contains different heterogeneous objects connected with each other and configure as one system. It also contains lots of objects configured which must communicate with each other to achieve a particular goal in a specific time. The proposed set of characteristics of the IoT are as follows: • Intelligence: The aspect of intelligence as in the sensing capabilities in IoT devices and the intelligence gathered from data analytics (also artificial intelligence), knowledge extraction from the generated data. • Connectivity: Devices, sensors, they need to be connected: to an item, to each other, actuators, a process and to ‘the Internet’ or another network. • Sensing: IoT will not be applicable easily without sensors which will detect or measure any changes in the environment to generate data that can report on their status or even interact with the environment. • Heterogeneous: The devices in the IoT are heterogeneous, so IoT devices can interact and communicate on different types of networks and with other IoT device. • Interoperability: when different things cooperate in order to provide the desired service, at the right time. • Memory Limitations: Most IoT healthcare devices have low memory. Such devices are activated using an embedded operating system (OS), system software, and an application binary. Therefore, their memory may not be sufficient to execute complicated security protocols. • Energy Limitations: A typical IoT healthcare network includes small health devices of limited battery power (e.g., body temperature and BP sensors). Such devices conserve energy by switching on the power-saving mode when no sensor reading needs to be reported. In addition, they operate at a low CPU speed if there is nothing important to be processed. • Scalability: The number of IoT devices has increased gradually, and therefore more devices are getting connected to the global information network. Therefore, designing a highly scalable security scheme without compromising security requirements becomes a challenging task. • A Dynamic Topology: Because the IoT-based healthcare network must be anywhere and anytime, a new topology that may be based on the existent ones, but a dynamic topology • Data confidentiality: Because medical data are a sensitive subject, we need to build a stream access control or identity management system. In Table 2, the relationship of IoT proposed characteristics are mapped with the quality model characteristics. This mapping helps in understanding how the proposed characteristics of the IoT have an effect on the IoT quality. A new quality model is proposed that includes all IoT characteristics; the old and new ones addressed in Table 1 and Table 2 can covers all the quality factors of the IoT system. However, the proposed quality model covers all characteristics and their

Quality Attributes for Evaluating IoT Healthcare Systems

501

Fig. 1 Requirements and application domains in smart healthcare

quality factors, it still has some challenges to be applicable. These challenges mainly the cost of quality measurement process, tools, and personals. Moreover, there are some characteristics such as intelligence can be measured subjectively, which may lead to different measurement results from one to another.

5 Comparison of Architecture for IoT-Healthcare Applications The summary of the comparison with the existing proposed architectures is shown in Table 3. The parameters considered for comparison of different architectures are no. of layers in the architecture, the complexity of the architecture, data reliability at fog layer, real-time application support, and security. The possible values for the chosen parameters can be any one of the following: low, moderate, or high. The value for complexity is chosen based on the no. of layers and the functionality or modules carried out at each layer. The value for data reliability at the fog layer is chosen based on the availability of clusters or distributed computing at the fog layer. None of the existing architectures emphasized the data reliability at the fog layer, which is a major concern for emergency care applications. The value for real-time application support is chosen based on the existence of the fog layer and the amount of work done at this layer. The value for security is chosen based on the no. of layers. Fig. 2 Characteristics of smart healthcare

Transparency Connectivity

Responsive

Intelligence

Sensitive Adaptive

Reliability

✓

Security

Robustness

✓

Compatibility

Effectiveness

Extendibility

Performance

Verification

✓

✓

✓

Portability

Functionality

✓

Efficiency

✓

✓

Usability

Maintainability

Reusability

Integrity

Limited resources

✓

✓

✓

✓

✓

Scalability

✓

✓

✓

✓

✓

✓

✓

Heterogeneous

✓

✓

✓

Data confidentiality

✓

✓

✓

Dynamic Topology

Table 2 The relation between the IoT characteristics and software quality factors

✓

✓

✓

✓

Connectivity

✓

✓

✓

Sensing

✓

✓

✓

Intelligence

✓

Interoperability

✓

✓

Responsive

502 L. Chhiba et al.

[24] Azimi et al [25] Kumar N [26] Plageras et al [27] Villalba et al [28] Mahmud et al [29] Debauche et al [30] Paul et al [31] Awaisi et al [32] Abdelmoneem et al

[21] Cerina et al [22] Nandyala et al [23] Verma et al

Less Less Less

Less

Less

Moderate

Moderate Moderate Moderate

2

3

3

3 3 4

High

5

3 2 3

Moderate Moderate

Complexity

4 4

No. Oflayers

Low Moderate Moderate

Low

Moderate

Low

Low Low Low

Low

Low Low

Data Reliability (at Fog Layer)

Table 3 Comparison of the proposed with the existing architectures

Moderate Moderate Moderate

Moderate

Moderate

Low

Moderate Low Low

Moderate

Moderate Moderate

Real-time Application Support

Moderate Moderate Moderate

Moderate

Moderate

Moderate

High Low Moderate

Moderate

High High

Security

Telemedicine (ECG monitor, smart watches, etc.) Respiration, blood pressure, glucose level, etc Acute care Hospital ECG monitoring

Home care

Home care telemedicine (ECG sensor, blood pressure…) ECG monitoring telemedicine Hospital

Healthy living (disease prevention) Home care and hospital

Application domain

Quality Attributes for Evaluating IoT Healthcare Systems 503

504

L. Chhiba et al.

6 Conclusion and Future Work Through this study we explain the concept of the Internet of things (IoT) and its relationship to the quality models that are already identified in the previous works. Furthermore, a new group IoT characteristics have been proposed to bridge the gap in the literature. These characteristics focus on the quality factors that have not been discussed before; Robustness and Scalability. The proposed characteristics can be matched with all quality factors that may affect the IoT systems quality. However, the proposed characteristics and quality model still have some drawbacks, such as cost. In the future work, a new dynamic quality model, that gives weight to each IoT characteristics, depending on the nature of IoT, and create a model with quality factors that measure the highest weight.

References 1. G.M. Foody, D.S. Boyd, Using volunteered data in land cover map validation: mapping West African forests. IEEE J. Sel. Topics Appl. Earth Observ. 6(3), 1305–1312 (2013) 2. L. Hu, Z. Zhang, F. Wang, K. Zhao, Optimization of the deployment of temperature nodes based on linear programming in the internet of things. Tsinghua Sci. Technol. 18(3), 250–258 (2013) 3. R. Bose, A paradigm shift in internet-based remote experimentation. IEEE Access 1, 718–725 (2013) 4. J. Kiljander et al., Semantic interoperability architecture for pervasive computing and internet of things. IEEE Access 2, 853–873 (2014) 5. Dellepianeand, S.G., Angiati, E.: Quality assessment of despeckled SAR images. IEEE J. Sel. Topics Appl. Earth. Observ. 7(2), 691–707 (2014) 6. F. Li, P. Xiong, Practical secure communication for integrating wireless sensor networks into the internet of things. IEEE Sens. J. 13(10), 3677–3684 (2013) 7. A. Zanella, N. Bui, A. Castellani, L. Vangelista, M. Zorzi, Internet of things for smart cities. IEEE Internet Things J. 1(1), 22–32 (2014) 8. J. Zhou, Hu. Liang, F. Wang, Lu. Huimin, K. Zhao, An efficient multi-dimensional fusion algorithm for IoT data based on partitioning. Tsinghua Sci. Technol. 18(4), 369–378 (2013) 9. X. Mao, C. Zhou, Y. He, Z. Yang, S. Tang, W. Wang, Guest editorial: Special issue on wireless sensor networks, cyber-physical systems, and internet of things. Tsinghua Sci. Technol. 16(6), 559560 (2011) 10. S. Chen, H. Xu, D. Liu, B. Hu, H. Wang, A vision of IoT: applications, challenges, and opportunities with china perspective. IEEE Internet things J. 1(4), 349–359 (2014) 11. Y. Zhang, Z. Zheng, M.R. Lyu, An online performance prediction framework for service-oriented systems. IEEE Trans. Syst. Man Cybern. Syst. 44(9), 1169–1181 (2014) 12. V.A. Reguera, E.M.G. Fernández, F.A. Paliza, W. Godoy Jr., VoIP quality of service in networks with active queue management, IEEE Latin America Trans. 7(5), 603–611 13. B. Kantarci, H.T. Mouftah, Trustworthy sensing for public safety in cloud-centric internet of things, IEEE Internet Things J. 1(4), 360–368 14. S. Lanziser, A.R. Weber, A. Liao, D. Pajak, A.K. Meier, Communicating power supplies: bringing the internet to the ubiquitous energy gateways of electronic devices, IEEE Internet Things J. 1(2), 153160

Quality Attributes for Evaluating IoT Healthcare Systems

505

15. L.R. Dominguez, D.M. Rodriguez, C.V. Rosales, D.T. Roman, J.R. Pacheco, RTT prediction in heavy tailed networks, IEEE Commun. Lett. 18(4), 700–703 16. E. Al-Fagih, F.M. Al-Turjman, W.M. Alsalih, H.S. Hassanein, A priced public sensingframework for heterogeneous IoT architectures, IEEE Trans. Emerg. Topics Comput. 1(1), 133–147 17. G. White, V. Nallur, S. Clarke, Quality of service approaches in; IoT: a systematic mapping. J. Syst. Softw. 132, 186–203 (2017) 18. E. Coutinho, M. Neto, W. Sales, C.I.M. Bezerra, J. Souza, Research opportunities in quality assessment of internet of things, software defined networks and network function virtualization environments (2018) 19. A. Bader, H. Ghazzai, A. Kadri, M.S. Alouini, Front-end intelligence for large-scale application-oriented internet-of-things. IEEE Access 4, 3257–3272 (2016) 20. A. Banerjee, S.K.S. Gupta, Analysis of smart mobile applications for healthcare under dynamic context changes. IEEE Trans. Mobile Comput. 14(5), 904–919 (2015) 21. L. Cerina, S. Notargiacomo, M.G. Paccanit, M.D. Santambrogio, A fog-computing architecture for preventive healthcare and assisted living in smart ambients, in 2017 IEEE 3rd International Forum on Research and Technologies for Society and Industry (RTSI) (2017) 22. C.S. Nandyala, H.-K. Kim, From cloud to fog and IoT based real-time U-healthcare monitoring for smart homes and hospitals. Int. J. Smart Home 10(2), 187–196 (2016) 23. P. Verma, S.K. Sood, Fog assisted-IoT enabled patient health monitoring in smart homes. IEEE Internet Things J. 5(3), 1789–1796 (2018) 24. I. Azimi et al., HiCH: hierarchical fog-assisted computing architecture for healthcare IoT. ACM Trans. Embed. Comput. Syst. 16, 1–20 (2017) 25. N. Kumar, IoT architecture and system design for healthcare systems, in 2017 International Conference on Smart Technologies for Smart Nation (SmartTechCon) (2017) 26. A.P. Plageras, K.E. Psannis, Y. Ishibashi, B.-G. Kim, IoT-based surveillance system for ubiquitous healthcare, in: IECON 2016-42nd Annual Conference of the IEEE Industrial Electronics Society (2016) 27. M.T. Villalba, M. Teresa Villalba, M. de Buenaga, D. Gachet, F. Aparicio, Security analysis of an IoT architecture for healthcare, in Internet of Things. IoT Infrastructures (2016), pp. 454–460 28. R. Mahmud, F.L. Koch, R. Buyya, Cloud-fog interoperability in IoT-enabled healthcare solutions, in Proceedings of the 19th International Conference on Distributed Computing and Networking - ICDCN ’18 (2018) 29. O. Debauche, S. Mahmoudi, P. Manneback, A. Assila, Fog IoT for health: a new architecture for patients and elderly monitoring. Procedia Comput. Sci. 160, 289–297 (2019) 30. A. Paul, H. Pinjari, W.-H. Hong, H.C. Seo, S. Rho, Fog computing-based IoT for health monitoring system. J. Sens. 2018, 1–7 (2018) 31. K.S. Awaisi, S. Hussain, M. Ahmed, A.A. Khan, G. Ahmed, Leveraging IoT and fog computing in healthcare systems. IEEE Internet Things Mag. 3(2), 52–56 (2020) 32. R.M. Abdelmoneem, A. Benslimane, E. Shaaban, S. Abdelhamid, S. Ghoneim, A cloud-fog based architecture for IoT applications dedicated to healthcare, in ICC 2019 - 2019 IEEE International Conference on Communications (ICC) (2019) 33. J.J. Tambotoh, S.M. Isa, F.L. Gaol, B. Soewito, H.L.H.S. Warnars, Software quality model for internet of things governance, in 2016 International Conference on Data and Software Engineering (ICoDSE) (2016). pp. 1–6 34. J. Kiruthika, S. Khaddaj, Software quality issues and challenges of internet of things, in Distributed Computing and Applications for Business Engineering and Science (DCAB)

Early Prediction of ICU Admission Within COVID-19 Patients Using Machine Learning Techniques Ikram Maouche, Sadek Labib Terrissa, Karima Benmohammed, Noureddine Zerhouni, and Safia Boudaira

Abstract The COVID-19 pandemic has become a great challenge for healthcare systems due to the urgent need for ICUs that exceeded their capacity. Determining critical patients that require ICU transfer early will be valuable in optimising ICU resources and triage the patients. We propose a ML-based approach to predict ICU requirement within COVID-19 patients based on clinical data. A Mexican dataset of 7078737 cases and 38 attributes was considered in this paper. We trained four models MLP, DT, RF, and GB, on 70% of the data with five fold cross-validation and tested using the remaining 30%. Classification accuracies obtained were 97.72%, 97.14%, 99.06%, and 99.28%, respectively. Feature importance analysis based on GB model showed that age, hypertension, diabetes, obesity, pneumonia, days between symptoms onset and hospitalisation, location of the care unit, and private and public insurance are the main important factors. The latter factors highlight the importance of rapid and good quality care. Keywords Prediction · COVID-19 · SARS-CoV-2 · Intensive care unit · ICU admission · Machine learning · Clinical data · Feature importance · Prognosis

1 Introduction With the beginning of 2020, the World Health Organisation (WHO) declared the COVID-19 disease caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) as a global pandemic that transmits rapidly between individuals. In Italy, one of the most affected countries by the first wave of the pandemic, the number I. Maouche (B) · S. L. Terrissa LINFI Laboratory, University of Biskra, Biskra, Algeria e-mail: [email protected] K. Benmohammed · S. Boudaira Endocrinology Department, Faculty of Medicine, Salah Boubnider University 3, Constantine, Algeria N. Zerhouni FEMTO-ST institute, Univ. Bourgogne Franche-Comte, CNRS, ENSMM, Besançon, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_41

507

508

I. Maouche et al.

of cumulative cases recorded until March 31st , 2020 have exceeded 100000 cases, among which 28000 were hospitalised. More than 4000 of the hospitalised patients required Intensive Care Unit (ICU) admission [1]. The Italian hospitals have struggled to manage the allocation of ICU and announced a saturation of resources that led to increased deaths. Due to the novelty of this disease and the unclear progression behaviour among the infected individuals from simple asymptomatic illness to a severe state that requires hospitalisation and critical care. The affected countries had no choice but to proclaim preventive measures as a result of the growing issues associated with the COVID-19 infection. The preventive measures included social distancing policies, travel restrictions, and postponing non-urgent care and surgeries [2]. Given this urgency, adoption of rapid and reliable clinical predictive models toward predicting ICU requirements within COVID-19 patients became an obligation to: (i) optimise the use of hospital resources and staff, especially in hospitals with limited resources; (ii) improve the social distancing measures by targeting the population who are at higher risk; (iii) make more efficient vaccination road-maps by prioritising high-risk individuals, especially in the first stages of vaccination where the produced amount may not be sufficient for all the population. Recently, machine learning (ML) techniques have been adopted in predictive tasks of many domains. In the industry field for predicting the Remaining Useful Life (RUL) prediction for engine degradation prognosis [3]. It has been used for networking protocols performance prediction [4]. For the medical field, machine learning has been applied for both diagnosis and prognosis tasks [5, 6]. During this pandemic, several predictive machine learning models have been developed. Some of these works focus on predicting the need for ICU admission using laboratory data obtained from blood test results such as C-reactive Protein, Leukocytes, Lymphocytes, and Creatinine [7–10] to make the prediction. In the same way, [11] used time-series data of laboratory and vital signs of COVID-19 patients to predict ICU transfer within 24 h. Moreover, other works [12, 13] incorporated Computerised Tomography (CT) images with laboratory data to improve the predictive performance of the models that rely on clinical data only. These works showed high predictive performance. However, laboratory-data-based models may be inconvenient during this pandemic where the time and resources to acquire this information is restricted. To enhance the predictive performance and the understanding of COVID-19 prognosis, we attempt through this study to design a clinical-data-driven method to address two questions. Will the patient need intensive care admission? and what are the clinical factors that lead to ICU risk?. The rest of the paper is organised as follows. Section 2 describes the proposed method. Section 3 reports the experimental results and a discussion of the results. Finally, we conclude this paper in Sect. 4.

Early Prediction of ICU Admission Within COVID-19 Patients...

509

Fig. 1 Flowchart of our proposed method.

2 Proposed Method Figure 1 illustrates the flowchart of our proposed method. It is composed of three parts: data pre-processing and feature engineering, model training, and prediction. The output of the prediction is whether the patient requires ICU admission or not. Each part will be separately explained in the upcoming sub-sections.

2.1 Data Description The dataset used in this paper is a public dataset provided by the Mexican government, which is daily updated [14]. By the 2nd of June 2021, it contained 7078737 cases that have been tested for COVID-19 in Mexico and 38 attributes. The dataset includes information on socio-demographics and lifestyle characteristics, medical conditions, and COVID related characteristics. The positive cases included in the dataset are confirmed by the Mexican Ministry of Health when tested positive and have shown at least one of COVID main symptoms such as cough, fever, and chest pain. Data details are provided below in Table 1.

2.2 Feature Engineering and Data Pre-processing In this sub-section, we will describe the operations applied to the data to improve its predictive quality. Data Cleaning The dataset was cleaned from instances that are incomplete or irrelevant to the outcome. We eliminated instances where: (i) the outcome of ICU admission is unknown, (ii) the laboratory or antigen test results came negative, (iii) the final classification of COVID was not positive. Feature Selection The dataset contained many attributes that seem redundant, have a strong correlation, or have many missing values. The attributes that had missing values superior to 99%

510

I. Maouche et al.

Table 1 Data description. Category

Attribute

Description

Socio-demographic and lifestyle characteristics

EDAD

age

SEXO

sex

Medical conditions

COVID related characteristics

ENTIDAD_NAC

patients’ birth state

ENTIDAD_RES

patients’ state of residence

MUNICIPIO_RES

patients’ city of residence

NACIONALIDAD

Mexican or foreign

INDIGENA

indigenous individual

HABLA_LENGUA_INDIG

use of indigenous language

MIGRANTE

migrant

PAIS_NACIONALIDAD

nationality

PAIS_ORIGEN

country of origin

TABAQUISMO

smoking

DIABETES

diabetes

HIPERTENSION

hypertension

EPOC

chronic obstructive pulmonary disease (COPD)

ASMA

asthma

INMUSUPR

immunosuppression

RENAL_CRONICA

chronic kidney disease (CDK)

CARDIOVASCULAR

cardiovascular

OBESIDAD

obesity

EMBARAZO

pregnancy

OTRAS_COM

having other comorbidities

ORIGEN

monitored by USMER (Healthcare Monitoring Unit for Respiratory Diseases)

SECTOR

healthcare sector that provided the care

OTRO_CASO

having contact with another SARS-CoV-2 case

NEUMONIA

pneumonia

TOMA_MUESTRA_LAB

laboratory test performed

RESULTADO_LAB

laboratory test result

TOMA_MUESTRA_ANTIGENO

antigen test performed

RESULTADO_ANTIGENO

antigen test result

CLASIFICACION_FINAL

classification of COVID diagnosis

TYPO_PACIENTE

type of care (inpatient or outpatient)

ENTIDAD_UM

location of the medical unit that provided the care

FECHA_SINTOMAS

date of symptoms appearance

FECHA_INGRESO

date of hospitalization

FECHA_DEF

date of death

INTUBADO

tracheal intubation

UCI

ICU admission

Early Prediction of ICU Admission Within COVID-19 Patients...

511

Fig. 2 Correlation Matrix heatmap for the dataset.

were dropped; these attributes are MIGRANTE and PAIS_ORIGEN. In addition, the attributes I N T U B AD O and F EC H A_D E F which come after the ICU outcome were removed. Figure 2 shows the correlation heatmap between the attributes. We fixed the threshold of high correlated attribute pair to 0,7. The high correlated pairs are (ENTIDAD_NAC, ENTIDAD_UM), (ENTIDAD_RES, ENTIDAD_UM), (ENTIDAD_RES, ENTIDAD_NAC), (INDIGENA, HABLA_LENGUAINDIG) and one attribute of each pair was eliminated. The medical conditions also showed a correlation superior to the threshold between them, but this was expected due to the interactive influence of these diseases [15–18] so their correlation was neglected. In addition, we eliminated MUNICIPIO_RES attribute due to its large cardinality that exceeds 500 values which could impact the performance of the prediction models. RESULTADO_LAB, TOMA_MUESTRA_ANTI-GENO, RESULTADO_ANTIGENO, and CLASSIFICATION FINAL are also removed since after data cleaning and taking only positive covid cases all these attributes contain one value. Finally, we generated a new attribute DaystoAdm which represents the duration between symptoms onset (FECHA_SINTOMA) and hospitalisation (FECHA_INGRESO), then these latter attributes were dropped.

512 Fig. 3 Class distribution before (blue) and after (red) using SMOTE technique.

I. Maouche et al. 4,957 Original Generated 4,000

2,000

2,000

189 0 Non-ICU

ICU

Encoding of Categorical Variables The categorical attributes present in the dataset were represented using one-hot encoding or label encoding depending on their cardinality. Handling Missing Values The missing data were imputed with K-nearest neighbour imputer using the Euclidean distance measure, and we fixed the number of neighbours to k = 2. Handling Unbalanced Data The unbalanced distribution of the instances as depicted in Fig. 3 was solved using the SMOTE technique. We choose to oversample the minority class in the training dataset than under-sampling to avoid losing important information; however, oversampling works by duplicating instances which could lead to overfitting in the classification. To this end, we tested different overs-sampling proportions and we found that augmenting the size of the minority class around half the size of the majority class gave the best results.

2.3 Classification Models The present study focuses on the prediction of the need for ICU admission within COVID-19 patients. This problem is formulated as a classification task. We select four machine learning models that are well-known for their high predictive performance in the classification tasks. Decision Tree Decision tree (DT) is a classification algorithm that constructs a tree-like model to describe the relationship between the attributes and a target class. An internal node represents a test on a single feature, and a terminal node (leaf) is assigned to a target class. The algorithm works by a binary recursive splitting of the instances using splitting criteria to decide if an attribute should be placed at the root or as an internal node at different levels of the tree [19].

Early Prediction of ICU Admission Within COVID-19 Patients...

513

Random Forest Random Forest (RF) is an ensemble learning classification method that works by constructing multiple simple decision trees in the training stage using bootstrapping, and in the classification stage, an unweighted majority voting between those trees is made, and the class voted by most trees is chosen [20]. Multi Layer Perceptron Multilayer Perceptron (MLP) is a feed-forward artificial neural network characterised by an input layer that receives the input data, hidden layers responsible for exploring the relationship between the input and the output, and an output layer that assigns the input to a class. The training is achieved using a back-propagation algorithm to adjust the weights of the connection between the neurons. The neurons of hidden and output layers use an activation function that maps the weighted inputs to the output of each neuron. Gradient Boosting Gradient boosting (GB) is an ensemble learning classification algorithm that produces a set of decision trees called weak learners in an additive manner. During each step, the newly created tree considers the errors made by the previous trees and tries to minimise the loss function by parameterising the tree. The tree that reduces the loss is chosen to be added to the model.

3 Results and Discussion We ran the experimentation on Python. The hyper-parameters of each model were adjusted using the Grid Search method. Table 2 denotes the best hyper-parameters values chosen for each model based on accuracy. The dataset was split into 70% training set and 30% testing set. The validation of the models during the training step was made using five fold cross-validation. The results of classification represented in Table 3 shows that Gradient Boosting and Random Forest achieved the highest prediction accuracies, 99.06% and 99.28%, respectively. Due to the limit of the accuracy metric for unbalanced classification problems, we extended our evaluation of the performance to other metrics: precision, F1-score, sensitivity, and specificity. The precision results show that all classifiers had high precision in predicting the patients that do not require ICU admission (class 0) ranging from 99% to 100% with MLP, RF, and GB predicting 100% of the cases positively. For the patients requiring an ICU admission (class 1), the precision ranges from 93% using DT to 98% using GB. Both GB and RF achieved 100% sensitivity, which means all patients that require intensive care were predicted correctly. On the other hand, their specificity reached 98.58% and 98.92%, respectively. i.e. the majority of the ICU safe patients were classified as non-ICU. These results conclude that Gradient Boosting overcame the rest of the classifiers and can be used accurately to predict ICU requirement within COVID-19 patients.

514

I. Maouche et al.

Table 2 Adjusted hyper-parameters of the classification models using Grid Search. Model Hyper-parameter Adjusted value MLP

hidden_layer_sizes activation learning_rate max_depth n_estimators max_depth n_estimators max_features subsample max_depth

DT RF GB

Table 3 Performance results. Model Class Accuracy % Precision MLP DT RF GB

0 1 0 1 0 1 0 1

97.72 97.14 99.06 99.28

1.00 0.94 0.99 0.93 1.00 0.97 1.00 0.98

112 tanh adaptive 20 38 26 36 sqr t 0.8 29

F1-score

Sensitivity %

Specificity %

0.98 0.97 0.98 0.96 0.99 0.99 0.99 0.99

99.46

96.84

98.93

96.23

100

98.58

100

98.92

The identification of the important risk factors of ICU risk is critical for a better triage of COVID-19 patients. To this goal, We have used the permutation feature importance [21] method which is a global post-hoc explainability method. The results of this method with gradient boosting as denoted in Fig. 4 shows that the age of the patient is the most important risk factor in the socio-demographic characteristics. While, hypertension, diabetes, and obesity are the most critical risk factors in the medical conditions. These results conform with the previous studies [22–24]. However, surprisingly, cardiovascular disease is not an essential risk factor contrary to [22, 23]. As for the COVID related characteristics, we found that the location of the care unit, days from symptoms onset to hospitalisation, public and private insurance sector, contact with other covid cases and pneumonia are considered risk factors for ICU need. The work in [25] indicated that patients who live in municipalities with greater marginalisation are at greater risk of ICU admission, which corroborates our findings on the location of the care unit. Whereas they found that private sector patients have high chance of admission to ICU. While, our findings suggest that both public and private insurance patients have more chance of admitting to ICU

Early Prediction of ICU Admission Within COVID-19 Patients...

515

Fig. 4 Variable importance based Gradient Boosting model.

than uninsured patients. An explanation for this result is that insured patients in both public and private sectors are more privileged than non insured patients in receiving critical care. The work in [23] indicated the significance of duration between the beginning of symptoms to hospitalisation with increasing ICU requirement, which matches those observed in our study. In conclusion, this study’s results indicate that socio-demographic and comorbidities are not the only causes of poor COVID-19 prognosis. The health infrastructure, health coverage, and early medical care play an essential role in ICU requirement.

4 Conclusion A machine learning-based approach was proposed to predict ICU admission within COVID-19 patients based on clinical data. The classification using the Gradient Boosting model achieved the highest performance among all models. The explanation of the predictive performance of this model showed that age, hypertension, diabetes, obesity, days from symptoms onset to admission, pneumonia, public and private health insurance, and the location of the unit that provided the care are the most critical features in ICU prediction. These results indicate that the cause of COVID-19 poor outcome of ICU is not related to the socio-demographic and medical characteristics of SARS-CoV-2 only but also with the type of medical care. This study has gone some way towards enhancing the understanding of COVID-19 poor outcome of ICU admission by highlighting the importance of early medical care, the upgrade of the health infrastructure in marginalised areas, and the availability of health coverage.

516

I. Maouche et al.

References 1. Dati covid-19 Italia. https://github.com/pcm-dpc/COVID-19/blob/master/schede-riepiloga tive/regioni/. Accessed 30 June 2021 2. M. Chinazzi, J. Davis, M. Ajelli, C. Gioannini, M. Litvinova, S. Merler, A. Piontti, K. Mu, L. Rossi, K. Sun, C. Viboud, X. Xiong, H. Yu, M. Halloran, I. Longini, A. Vespignani: The effect of travel restrictions on the spread of the 2019 novel coronavirus (COVID-19) outbreak. Science, 368, eaba9757 (2020) 3. I. Remadna, L.S. Terrissa, S. Ayad, R. Zemouri, An overview on the deep learning based prognostic (2018) 4. S. Zroug, L. Kahloul, S. Benharzallah, K. Djouani, A hierarchical formal method for performance evaluation of WSNS protocol. Computing 103, 06 (2021) 5. A. Belaala, L.S. Terrissa, N. Zerhouni, C. Devalland, Computer-aided diagnosis for spitzoid lesions classification using artificial intelligence techniques. Int. J. Healthcare Inf. Syst. Inform. 16, 22 (2021) 6. R. Bojana, A. Cirkovic. Machine learning approach for breast cancer prognosis prediction. Computational Modeling in Bioengineering and Bioinformatics (2020), pp. 41–68 7. J. Heo, D. Han, H. Kim, D. Kim, Y. Lee, D. Lim, S. Hong, M. Park, B. Ha, W. Seog, Prediction of patients requiring intensive care for COVID-19: development and validation of an integerbased score using data from Centers for Disease Control and Prevention of South Korea. J. Intensive Care 9(1), 16 (2021) 8. P. Podder, M.R.H. Mondal, Machine learning to predict COVID-19 and ICU requirement, in 11th International Conference on Electrical and Computer Engineering (ICECE) (2020), pp. 483–486 9. E.C. Gök, M.O. Olgun: SMOTE-NC and gradient boosting imputation based random forest classifier for predicting severity level of COVID-19 patients with blood samples, Neural Computing and Applications (2021) 10. T. Dan, Y. Li, Z. Zhu, X. Chen, W. Quan, Y. Hu, G. Tao, L. Zhu, J. Zhu, Y. Jin, L. Li, C. Liang, H. Wen, H. Cai, Machine learning to predict ICU admission, ICU mortality and survivors’ length of stay among COVID-19 patients: toward optimal allocation of ICU resources, in 2020 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) (2020), pp. 555–561 11. F. Cheng, H. Joshi, P. Tandon, R. Freeman, D.L. Reich, M. Mazumdar, R. Kohli-Seth, M.A. Levin, P. Timsina, A. Kia, Using machine learning to predict ICU transfer in hospitalized COVID-19 patients. J. Clin. Med. 9(6), 1668 (2020) 12. Q. Xu, X. Zhan, Z. Zhou, Y. Li, P. Xie, S. Zhang, X. Li, Y. Yu, C. Zhou, O. Gevaert, G. Lu, AI-based analysis of CT images for rapid triage of COVID-19 patients. npj Digit. Med. 75, 1–11 (2021) 13. H. Estiri, Z.H. Strasser, S.N. Murphy, Individualized prediction of COVID-19 adverse outcomes with MLHO. Sci. Rep. 11(1), 5322 (2021) 14. D. Abiertos, Dirección General de Epidemiología. https://www.gob.mx/salud/documentos/ datos-abiertos-152127. Accessed 11 Aug 2021 15. E. Wouters, Obesity and metabolic abnormalities in chronic obstructive pulmonary disease. Ann. Am. Thoracic Soc. 14, S389–S394 (2017) 16. A. Berbudi, N. Rahmadika, A.I. Cahyadi, R. Ruslami, Type 2 diabetes and its impact on the immune system. Curr. Diabetes Rev. 16(5), 442–449 (2020) 17. C.Y. Chen, K.M. Liao, Chronic obstructive pulmonary disease is associated with risk of chronic kidney disease: a nationwide case-cohort study. Sci. Rep. 6, 25855 (2016) 18. B. Rodriguez-Iturbe, H. Pons, R.J. Johnson, Role of the immune system in hypertension. Physiol. Rev. 97(3), 1127–1164 (2017) 19. J.Q. Quinlan, Simplifying decision trees. Int. J. Man Mach. Stud. 27(3), 221–234 (1987) 20. L. Breiman, Random forests. Mach. Learn. 54(1), 5–32 (2001) 21. A. Altmann, L. Tolo¸si, O. Sander, T. Lengauer, Permutation importance: a corrected feature importance measure. Bioinformatics 26(10), 1340–1347 (2010)

Early Prediction of ICU Admission Within COVID-19 Patients...

517

22. S. Wollenstein-Betech, C.G. Cassandras, I.C. Paschalidis, Personalized predictive models for symptomatic COVID-19 patients using basic preconditions: hospitalizations, mortality, and the need for an ICU or ventilator. Int. J. Med. Inform. 142, 104258 (2020) 23. F.D. Martos-Benítez, C.D. Soler-Morejón, D. García-del Barco, Chronic comorbidities and clinical outcomes in patients with and without COVID-19: a large population-based study using national administrative healthcare open data of Mexico. Intern. Emerg. Med. 16(6), 1507–1517 (2021) 24. O.Y. Bello-Chavolla, J.P. Bahena-López, N.E. Antonio-Villa, A. Vargas-Vázquez, A. GonzálezDíaz, A. Márquez-Salinas, C.A. Fermín-Martínez, J. Naveja, C.A. Aguilar-Salinas, Predicting mortality due to SARS-CoV-2: a mechanistic score relating obesity and diabetes to COVID-19 outcomes in Mexico. J. Clin. Endocrinol. Metab. 105(8), 2752–2761 (2020) 25. L. Ortiz Hernández, M. Pérez-Sastré, Social inequalities in the progression of COVID-19 in the Mexican population. Revista Panamericana de Salud Pública, 44, e106 (2020)

Agent-Based Model for Analyzing COVID-19 Infection in the Campus Using AnyLogic Software W. X. Gan and S. Amerudin

Abstract COVID-19 is a fatal global pandemic that have been spread throughout the world rapidly. Based on the global statistics, the confirmed cases has reached 662 million cases at the mid of June 2021. Typically, COVID-19 is transmitted when a healthy person is closed contact with the infected person via the respiratory droplet or saliva. With the reopening of the academic institution in Malaysia, the formation of new clusters become more seriously. As until 21st April 2021, there are total of 83 COVID-19 clusters is reported that related to the education sector since early January 2021. The students may come back to the campus for conducting academic activities. Therefore, this study is proposed to analyze the COVID-19 infection inside the campus using ABM. The designed ABM model has three different settings, which are lecture room, laboratory and office. Students are the agents of the simulation. The factors of COVID-19 infection are social distance, ventilation condition of the room and exposure time of contact. The ABM model allows the users to analyze the effect of number of people and social distance towards COVID-19 infection. Based on the preliminary analysis, office has the highest risk, followed by lecture room and laboratory. For generating less than 25% of new infected people, the students should maintain at least 1.8 m of social distance. Through the model, the administrators can use to plan the classroom and laboratory to the students. This paper suggests to extend the research by analyzing other rooms in the campus. Keywords COVID-19


Social distance
Agent-Based Modelling (ABM)

W. X. Gan
S. Amerudin (&) Universiti Teknologi Malaysia, Johor Bharu, Malaysia e-mail: [email protected] W. X. Gan e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_42

519

520

W. X. Gan and S. Amerudin

1 Introduction Coronavirus (COVID-19) has been recognized as the main issue for worldwide since year 2020. It is a fatal global pandemic that is caused by Severe Acute Respiratory Syndrome Coronavirus 2 (SARSCoV2) [1]. From the global statistics of mid of June 2021, the cumulative confirmed COVID-19 cases has been reached 176 million, whereas 662 thousand cases in Malaysia [2]. The virus is transmitted when contacting with the infected person via their saliva or respiratory droplets. World Health Organization (WHO) advises people to stay at least 1.0 m of physical distance because an estimation of 92% of cases will be lowered if the proper social distancing practices are conducted by the community [3]. For the indoor environment, the COVID-19 infection risk is affected by social distance, ventilation condition and exposure time of contact [4]. Computational model like Agent-Based Modelling (ABM) is the recent popular effective tool to analyze diseases spread in terms of geographical and demographic aspects [5]. An ABM model is comprised of interacting agents and their autonomous behaviors within an environment. The repetitive dynamic processes are allowable to show the agent interaction over the time [6]. Comparing with other techniques such as Ordinary Differential Equation (ODE) model and Cellular Automata (CA) model, ABM has more advantageous with its high ability and flexibility for simulating the individual movements and also the interactions among the individuals [7]. In Malaysia, there are total of 83 COVID-19 clusters is reported that related to the education sector from 1st January 2021 to 21st April 2021 [8]. Through this statistics, it shows that educational area is a high risk place to form the new COVID-19 clusters. In this study, an agent-based simulation model of COVID-19 infection that consider the social distance, ventilation condition and exposure time within building is developed. It is used to simulate the students movement into the room spaces of campus during COVID-19 pandemic and predict the spread of the virus infection based on social distance using ABM.

2 Literature Review The infection risk of COVID-19 is affected by social distance, ventilation and exposure time [4]. In a confined space, only considering the social distance is not enough to prevent the COVID-19 infection.

Agent-Based Model for Analyzing COVID-19 Infection …

2.1

521

Social Distance

Social distance describes the physical distance that maintained between the individuals. With the outbreak of COVID-19, WHO advises people to stay at least 1.0 m of physical distance to each other to reduce the contact between people [9]. Table 1 summarizes the social distancing rules for some countries [10].

Table 1 Social distancing rules by countries [10]

Country

Social distancing rule

Australia Brazil Canada China Japan Malaysia South Africa United Kingdom United State

Keep 1.5 m At least 2.0 m At least 2.0 m More than 1.0 m At least 1.8 m At least 1.0 m At least 1.0 m At least 2.0 m Maintain 2.0 m

The review of the studies of the comparative meta-analysis with COVID-19, SARS and MERS shows that the current policies that staying at least 1.0 m social distancing is just contributed on large reduction in infection [11]. Among of the social distance, 2.0 m maybe more effective than 1.0 m to control the spread of influenza disease. There is strong association of proximity between the exposed individual with the risk of infection. The relative effect might decreased 2.02 times for every 1 m away of distancing [11] (Fig. 1).

Fig. 1 Relationship of social distance and COVID-19 risk [11]

522

2.2

W. X. Gan and S. Amerudin

Ventilation

COVID-19 is one of the airborne disease which can be transmitted easier in a poor ventilation [12]. However, only maintaining the social distance is not enough to control the spread of COVID-19 especially in the indoor. There is necessary to increase the amount of air into the building from outdoor to dilute the indoor air. Wells-Riley Model is a popular model to predict and examine the infection risk of disease (Eq. 1). The purpose of this model is to examine the relationship of ventilation rate and the probability of infection. PI ¼ 1  expð

Iqpt Þ Q

ð1Þ

where PI is the probability of infectious risk, I is number of infected persons, q is quantum generation rate produced by an infector, p is pulmonary ventilation rate of individual, t is exposure time and Q is the ventilation rate. Since both ventilation and social distance affect the infection risk, [13] have modified the Wells-Riley Model to model the infected probability of the COVID-19 in the confined space by adding the consideration of social distance probability (Pd) and effectiveness of the ventilation factor (Ez) (Eq. 2). Through this model, the relationship of social distance, exposure time and effectiveness of ventilation factors towards the COVID-19 infection rate can be analyzed. PI ¼ 1  expðPd

Bqpt Þ E z  Q=N

ð2Þ

where Q/N is the minimum ventilation rate per person and B is initial infection rate. Q/N is referenced to the standard of American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) based on the occupancy settings (Table 2). Table 2 Minimum requirement of ventilation rate [14] Occupancy category

People outdoor air rate (L/s/person)

Lecture Classroom Office Space University/College Laboratories

3.8 2.5 5

Pd is associated with the social distance and calculated from Eq. 3 [13]. Pd ¼ 18:19lnðdÞ þ 43:276

ð3Þ

Agent-Based Model for Analyzing COVID-19 Infection …

2.3

523

Exposure Time

[15] stated that the close contact increases the chances of getting infection. It describes close contact happened when the persons are contacted for at least 15 min within six feet (1.8 m) over 24 h. If a healthy person is closed contact with an infected person, it is more likely infected by the COVID-19.

3 Methodology 3.1

Study Area

The study area is located at the Block C02, Faculty of Built Environment and Surveying (FABU) in Universiti Teknologi Malaysia (UTM), Johor, Malaysia (Fig. 2). Different room settings is represented by different floor level, just as shown in Table 3. The room is assigned their respective parameter of ventilation condition based on the ASHRAE standard (Table 2). Fig. 2 Location map of study area

Table 3 Representative of study area Floor level

Room type

Room name

1 2 4

Laboratory Administrator office Lecture room

GIS Lab 2 Postgraduate office (Reception Area) Lecture room 2

3.2

Data Source

The shapefile of the FABU rooms and the interior layout of the rooms of the buildings are referred to the building layout map from Office of Asset Development

524

W. X. Gan and S. Amerudin

(PHB) UTM. The non-spatial data is composed of the number of FABU students and staffs that entering the UTM campus during the semester 2 session 2020/2021. The statistics of students and staffs are collected from FABU UTM.

3.3

Design of Agent-Based Model

In the model scenario, the agents will walked into the rooms and stayed inside the room. The time of staying is associated to the “Time of Contact” in input interface. The agents will leave the room when the exposure time is ended. Model Interface There are two types of model interface, which are input interface and result interface. The input interface is displayed before the simulation, whereas the result interface is shown during the simulation. In the input interface, the users are allowed to input the tested parameters such as number of people, fraction initially infected probability, social distance, time of contact and probability of infection through slider or typing the value in the edit box (Fig. 3). The result interface consists of a simulation map, graph showing the number of infected people over the time and agents. During the simulation, the number of infected number is updated with the simulation time in seconds by graph (Fig. 4).

Fig. 3 Input interface of ABM model

Agent For this model, the agents are represented by FABU students and staffs. At the start of the simulation, the model will generated some agents with susceptible state and infectious state. The “Susceptible” agents will changed to “Exposed” agents when they are contacted with “Infectious” agent. The “Exposed” agent is then changed to “Infectious” agent if the social distance and exposure time are not favored. The state chart of agent is shown in Fig. 5.

Agent-Based Model for Analyzing COVID-19 Infection …

525

Fig. 4 Result interface of ABM model

Fig. 5 State of agent

3.4

Model Development

The ABM model is developed using AnyLogic software. The spatial data of FABU UTM is imported into the software and acts as the base map for the simulation. Statechart The state of the agents is constucted using Statechart tool in AnyLogic (Fig. 6). At the start of simulation, the model will simulate the movement of “Susceptible” agent and few “Infectious” agent. The number of “Infectious” agent is associated with the value of “Fraction Initially Infected” in the input interface. The transition of agent state from “Susceptible” to “Exposed” is triggered when the agent is contacted with the “Infectious” agent and received the message “Infection”. This message is sent by the “Infectious” agent. Then, the “Exposed” agent will changed to “Infectious” agent after some time out. The Eq. 2 is rearranged to find the time (t) for becoming infectious (Eq. 4).

526

W. X. Gan and S. Amerudin

Fig. 6 Statechart of ABM model

t¼

Inð1  PI Þ
E z
Q=N Pd
Bqp

ð4Þ

In the Statechart of the model, t is represented by TimeSecond and the values of the parameters are considered the time unit of “seconds”. Some of the parameter values that applied in the model are constant, while others can be changed and tested with different value through input interface (Table 4).

Table 4 Parameters of TimeSecond Parameters

Value

Reference

p q Ez

[16] [13] [14]

Pd B

0.000083 0.238 1.0 Ceiling supply of cool air 5 Laboratory 3.8 Lecture room 2.5 Office Calculated by Eq. 3 0 to 1.0

PI

0 to 1.0

Q/N

[14] [14] [14] “Social Distance” in Input Interface “Fraction Initially Infected” in Input Interface “Probability of Infection” in Input Interface

Pedestrian Library Pedestrian library in AnyLogic software is used to stimulate the human behaviors. The blocks such as PedSource, PedService, PedGoTo and PedSink are used (Fig. 7). PedSource defines where and how the agents started to appear in the model. The maximum of people arrival is linked to the “Number of People” in the input interface. The agents will entered the room and stay in the seats. The delay time of PedService is linked to the “Time of Contact” in the input interface. In PedGoTo, the agents will walked into the exiting door after the delay time reached. PedSink removes the agents in the simulation model.

Agent-Based Model for Analyzing COVID-19 Infection …

527

Fig. 7 Pedestrian library blocks

Table 5 Parameter used for the analysis

3.5

Parameter

Value

Number of people Fraction initially infected Social distance (m) Time of contact (minutes) Probability of Infection

25 0.2 (20%) 1.0, 1.5, 1.8, 2.0 10 0.02 (2%)

Analysis Using Developed ABM Model

Table 5 shows the value of the parameter for the preliminary analysis. These parameters value can be edited in the input interface before the simulation.

4 Preliminary Results During the simulation, the infectious agents will transmitted the disease when they are closed contact with other agents. As shown in Fig. 8, the number of infected people with Infectious state is increased with the time. Based on the preliminary result in Fig. 9, the number of new infected people in the Postgraduate Office (reception area) is the highest, followed by Lecture Room 2 and GIS Lab 2. This differences may due to the different in ventilation requirement based on ASHRAE standard. When comparing the social distance, the number of new infected people is decreased with the increased of social distance. 1.8 m and 2.0 m of the social distance can produced less than 25% of new infected people after 10 min of exposure time. Social distance of 1.0 m and 1.5 m have generated more than 60% of the total new infected people in Postgraduate Office (reception area) and Lecture Room 2. Among the settings, GIS Lab 2 shows the most favorable result as there is no added infected people in 10 min if the students have maintained 1.8 m and 2.0 m.

528

W. X. Gan and S. Amerudin

Fig. 8 ABM model (a) during early simulation and (b) after 10 min simulation

Fig. 9 Comparison of analysis with 10 min of exposure time

5 Conclusion ABM is a technique that can be used to analyze the spread of COVID-19 through agent’s interaction. The model is hopefully useful for the administrators to plan the social distance and occupancy capacity for laboratory, lecture room and office during pandemic. The preliminary result shows that practising 1.8 m and 2.0 m of social distance would be safer than 1.0 m and 1.5 m. This study may further applied to analyze other rooms in the UTM. It can be a reference tool for improving the current Standard of Procedures (SOP) of the campus during the COVID-19

Agent-Based Model for Analyzing COVID-19 Infection …

529

pandemic. However, as this ABM model is developed using AnyLogic software in Personal Learning Edition, the maximum simulation time for the analysis is only one hour. The research is suggested to consider another parameters for the future studies and development. These include the effect on the COVID-19 infection by different room size (dimension), seat arrangement and etc. Acknowledgements We would like to express our gratitude to the MOHE and UTM for their financial support in the research grant Vot Number Q.J130000.3852.19J13.

References 1. World Health Organization (WHO). Naming the coronavirus disease (COVID-19) and the virus that causes it. World Health Organization (2020). https://www.who.int/emergencies/ diseases/novel-coronavirus-2019/technical-guidance/naming-the-coronavirus-disease-(covid2019)-and-the-virus-that-causes-it 2. Center for Systems Science and Engineering of Johns Hopkins University (JHU CSSE). CSSEGISandData (2021). GitHub. https://github.com/CSSEGISandData/COVID-19 3. K. Prem et al., The effect of control strategies to reduce social mixing on outcomes of the COVID-19 epidemic in Wuhan, China: a modelling study. Lancet Public Health 5(5), e261– e270 (2020) 4. H.C. Burridge et al., The ventilation of buildings and other mitigating measures for COVID-19: a focus on wintertime. Proc. Royal Soc. A 477(2247), 20200855 (2021) 5. F. Dignum (ed.), Social Simulation for a Crisis: Results and Lessons from Simulating the COVID-19 Crisis (Springer, Cham, 2021) 6. C. Macal, M. North, Introductory tutorial: Agent-based modeling and simulation, in Proceedings of the Winter Simulation Conference 2014 (IEEE 2014), pp. 6–20 7. N.M. Gharakhanlou, M.S. Mesgari, N. Hooshangi, Developing an agent-based model for simulating the dynamic spread of Plasmodium vivax malaria: a case study of Sarbaz, Iran. Ecol. Inf. 54, 101006 (2019) 8. Bernama. Covid-19: 83 clusters in education sector recorded, 49 still active. Bernama (2021). https://www.bernama.com/en/news.php?id=1954443 9. World Health Organization (WHO). Coronavirus Disease 2019 (COVID-19) Situation Report–83. World Health Organization (2020). https://www.who.int/docs/defaultsource/ coronaviruse/situationreports/20200412-sitrep-83-covid-19.pdf 10. K. Xie, B. Liang, M.A. Dulebenets, Y. Mei, The impact of risk perception on social distancing during the COVID-19 pandemic in China. Int. J. Environ. Res. Public Health 17 (17), 6256 (2020) 11. D.K. Chu et al., Physical distancing, face masks, and eye protection to prevent person-to-person transmission of SARS-CoV-2 and COVID-19: a systematic review and meta-analysis. The Lancet 395(10242), 1973–1987 (2020) 12. L. Morawska et al., How can airborne transmission of COVID-19 indoors be minimised? Environ. Int. 142, 105832 (2020) 13. C. Sun, Z. Zhai, The efficacy of social distance and ventilation effectiveness in preventing COVID-19 transmission. Sustain. Cities Society 62, 102390 (2020). https://doi.org/10.1016/j. scs.2020.102390 14. ASHRAE. Standard 62.1–2019: Ventilation for Acceptable Indoor Air Quality. American Society of Heating, Refrigerating and Air-conditioning Engineers (2019)

530

W. X. Gan and S. Amerudin

15. Centers for Disease Control and Prevention (CDC). Contact Tracing Appendix. Centers for Disease Control and Prevention (2021). https://www.cdc.gov/coronavirus/2019-ncov/php/ contact-tracing/contact-tracing-plan/appendix.html 16. X. Duan, Exposure Factors Handbook of Chinese Population. China Environmental Science Press, Beijing (2013)

Smart GIS and Earth Management

Smart Prediction System for Territorial Resilience at the Large-Scale Level. Case Study of the Seasonal Forest Fires Risk in Northern Morocco Hicham Mharzi-Alaoui, Jean-Claude Thill, H. Bahi, H. Hajji, F. Assali, and S. Moukrim Abstract Predicting forest fire risks constitutes a significant component of territorial risk management and combat strategies. It plays a major role in resource allocation, in mitigation and recovery efforts as well as in anticipating landscape deterioration around urban areas that create an ecological balance at the territorial scale. The purpose of this study is to develop a smart predictive system of seasonal forest fire risk using a machine learning approach. To achieve this aim, data related to 2,130 forest fire events that occurred between 1997 and 2014 were used. Furthermore, biophysical characteristics over the study area were completely processed and retrieved from in-situ measurements; and from time series of MODIS and Landsat (TM, ETM+ and OLI/TIRS) satellite imagery. These data sources served to represent 5 groups of variables, namely Rainfall, Wind, Evapotranspiration, Normalized Difference Vegetation Index (NDVI) and Water Balance; in total, these variable groupings were structured into 39 elemental variables according to the month of the year. The Random Forests algorithm was used to find the best-fit link between theses predictors and the target variable of seasonal forest fire risk. The trained model exhibited a good predictive ability (83% of accuracy, p-value = 0.013). It showed that precipitations, mainly those of the wintry period, have a strong influence on fire occurrence and seasonal severity in the fire season. Accordingly, the developed model allows to predict seasonal risk according to the winter precipitation and to anticipate forest fire risks at a very early stage as well as their impact on improving socio-ecosystem services and territorial resilience. H. Mharzi-Alaoui (&)
H. Bahi School of Architecture, Planning and Design, Mohammed VI Polytechnic University, Benguerir, Morocco J.-C. Thill University of North Carolina at Charlotte, Charlotte, NC, USA F. Assali National Center For Climatic and Forest Risk Management, DEF, Rabat, Morocco H. Hajji Hassan II Institute of Agronomy and Veterinary Medicine, Rabat, Morocco S. Moukrim Department of Forest and Water, Rabat, Morocco © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_43

533

534

H. Mharzi-Alaoui et al.










Keywords Forest fire Machine learning Predictive modeling Random forest Territorial resilience Seasonal risk







1 Introduction One of the current challenges, perhaps the most important of modern times, is climate change, which affects countries in different ways. Climate change manifests itself primarily by an upward trend in temperature and by the higher frequency of extreme events, such as heat waves [1]. This results in a negative impact on forest ecosystems that become highly vulnerable to forest fires, especially in Mediterranean forest ecosystems. Morocco is not immune to this scourge, with 3,000 ha of forests that go up in smoke annually [2]. This situation is aggravated by the difficulties of post-fire regeneration. The increase in the frequency of drought periods, as it occurred during the summers of 2003, 2004 and 2006 in the Rif region of northern Morocco, causes serious fires that make all efforts in forest management and development much more problematic [3]. Furthermore, forest fire brings about considerable negative impacts in the cities located in the vicinity of forests plagued by fires. Indeed, smoke exacerbates the deterioration of the quality of the urban environment and there may be a heightened risk that fire would spread to human living quarters and property, with important economic loss and human loss when the urban perimeter and forest fuel come into contact. Recently, in the summer of 2021, many parts across the northern hemisphere have experienced the worst wildfires in years of recorded history. In Turkey, for example, 55,000 hectares of forest have been burnt in a single district (Mugla) more than twice the area burnt across the whole of Turkey in 2020 - and 36,000 people were evacuated. In Greece, more 25,000 hectares of forested area were burned in only four days, and many houses and vehicles were damaged. In order to avoid this kind of situations, or at least reduce the likelihood of wildfires, it has become imperative to be able to predict the prevalence and scope of forest fires in the next year or in the next season. This may help city and region managers to better anticipate fire-fighting management actions and the allocation of necessary resources to mobilize. Predictive modeling of seasonal forest fire risk seems to be a tool currently missing in Morocco [2] to increase the resiliency to future threats. Overall, the challenge of predictive modeling is how to combine the different indicators/predictors in order to make a decision and how to minimize errors due to bias and variance. In practice, a trade-off between these values is the norm because it is easy to devise a method with extremely low bias but high variance, or a method with very low variance but high bias. The challenge lies in finding a method for which both the variance and bias are low enough. This leads analysts to look for models with a deliberately limited number of variables; the “best” predictive model can give slightly biased estimators in favor of

Smart Prediction System for Territorial Resilience …

535

a compromise for a lower variance. A good model is no longer the one that best explains the data in the sense of minimal bias at the cost of a large number of variables inducing high variance. The good model then is the one that gives the most reliable predictions. This paper deal with the prediction problem; it presents a predictive model for seasonal forest fire risk classification in the Mediterranean context- case of Morocco - based on a machine learning approach in order to offer decision-makers a reliable tool to protect forests against fire.

2 Materials and Methods 2.1

Study Area

This study concerns the Rif Region, located in Northern Morocco, which extends over an area of 12,426 Km2 (Fig. 1). Climate of the Rif region is influenced by sea expanses, both the Atlantic Ocean to the West and the Mediterranean Sea to the North. The Atlantic side of the Rif is affected by heavy precipitation, with an average annual rainfall over 1000 mm, increasing to more than 2,000 mm in mountainous parts. On the Mediterranean coastline, the precipitation is usually less than 500 mm annually, decreasing to less than 200 mm in certain areas. The temperature range in the northern zone, which is influenced by the Mediterranean Sea, is lower than in the southern zone, which has a more continental climate [4]. According to the last Moroccan census of High Commission for Planning, this region represents 10.6% of population of the country and encompasses several large cities such as Chefchaouen, Larache, Ouezzane, Tanger, Assilah and Tétouan that can be highly vulnerable to fire risk.

Fig. 1 Location of the study area, with the forest cover depicted in green

536

H. Mharzi-Alaoui et al.

Given the favourable soils characteristics and climatic conditions, the Rif region has significant floristic potential. The forest cover represents around 35% of the total area of the region. This cover is dominated by cork oak and matorral formations followed by holm oak, pine, eucalyptus, zeen oak, cedar, etc. The natural forest encompasses 84.47% of the total forest area, against 15.53% covered by afforestation [4]. In addition to their ecological importance, forest ecosystems of the Rif contribute significantly to the socio-economic well-being at the national, regional or local levels; this appears obviously through the wood production annual average (from 2006 to 2009) which is 256 m3 for timber, 31,574 m3 for industrial wood, 51,728 m3 for timber service and 64,928 st for fuelwood. Besides, forest ecosystems provide society with multiple non-wood forest products (resins, mushrooms, hunting…), as well as a variety of other ecosystem services (soil and water protection, amenities, scenic beauty, recreation and tourism, etc).

2.2

Data Collected and Processed

Forest fires are a major problem in the Rif region (Fig. 2). Their frequency and magnitude vary considerably from year to year, and largely depend on two main factors, interacting directly or indirectly with each other, namely the vegetation status and climatic conditions [5].

Fig. 2 Fire event locations in the Rif region in 1997–2014

Smart Prediction System for Territorial Resilience …

537

In term of forest fire, any prediction mechanism has to take into account a continuous stream of observational data of a number of specific features informing these two parameters to understand and predict seasonal fire risk. The following data series are used. – Climatic Data Series Weather parameters such as precipitation, temperature, air humidity, wind, and sunshine affect the moisture content of vegetation and are therefore natural factors of wildfire fire spread [3]. Among these parameters, precipitation plays a predominant role in the water content of plants. It effect on vegetation varies significantly according to its quantity, duration and time during the calendar year. Wind supplies the fire with additional oxygen, provides even more dry potential fuel and increases the fire ignition and spread risk [3, 5]. Thus, during a live fire period, wind has the biggest impact on the fire behavior. In this study, the climatic data, spanning the period from 1997 to 2014, were extracted from the interpolated grids database based on the ALBACHIR’ Model coupled with the global model ARPEGE (Méteo-France). The chosen features are the ones that have a high correlation with the risk of fire and include the following: • • • •

Cumulative monthly precipitation, Average of monthly wind speed, Monthly evapotranspiration and Monthly water balance.

– Fire Events In addition to the climatic parameters, data on 2,130 forest fire events that occurred between 1997 and 2012 were sourced from the National Center for Climatic and Forest Risk Management. These data cover the territory of the Rif region, presenting information on geographic coordinates, burned area, species affected and date of ignition.

325 300 275 250 225 200 175 150 125 100 75 50 25 0 1997

1999 2001 Nombre Incendie

2003

2005 2007 2009 Linear (Nombre Incendie)

Fig. 3 Evolution of the number of fire events (1997–2011)

538

H. Mharzi-Alaoui et al.

6,000.00

Burned Area (ha)

5,000.00 4,000.00 3,000.00 2,000.00

y = -2.3451x + 1656.3 R² = 3E-05

1,000.00 -

(1,000.00)

19971998199920002001200220032004200520062007200820092010

Years

Fig. 4 Evolution of the burned area (1997–2011)

The number of fires in the study area is characterized by a very significant interannual variability (Fig. 3) with a very significant upward trend at the 5% significance level (r = 0.44, p = 0.05). The minimum was recorded in 1997 with 77 fires and the maximum is 303 fires recorded in 2005. The average of the burned area, from 1997 to 2014, is around 1,543 ha annually. While the number of fires trends upward longitudinally, the area affected by fire over the study period has a downward trend, but this relationship is not significant at the 5% significance level (Fig. 4). To understand why for some years we record vast expanses of burned land and vegetation cover, we have classified the data according to the total area burned each year. Specifically, three types of years were recognized to better years with large areas affected compared to those that experienced less burned area. Our approach then introduces a seasonal fire risk index on a scale of 1 to 3, where 1 corresponds to the ‘severe years’ when we have recorded a high burned area (area recorded is higher than the median) (>1000 ha) and 3 corresponds to the ‘weak years’ ( NIR] Kappa

0.8333 (0.5159, 0.9791) 0.5 0.01929 0.75

From the summary results reported in this table, the following main points emerge. The accuracy of prediction is 83%, which means that the model is able to predict severe years in 83% of cases with the predictors used in the RF model. The Kappa index is 0.75, which means that the model is 75% better than a random classification. These validation results indicate our model has strong external validity and is robust. This places us in a very strong position to advocate for the use of this model and its results for policy making and planning for future emergencies created by wild fires, especially in the vicinity of urban areas, where the potential consequences in terms of loss of lives and economic impacts are quite substantial.

546

H. Mharzi-Alaoui et al.

4 Conclusions and Recommendations The paper presented a smart prediction system of seasonal forest fire risk, based on the use of Random Forests as a machine learning ensemble model, and 39 elemental variables derived from meteorological parameters (precipitation, wind, evapotranspiration …) and vegetation status (NDVI). The results demonstrate the ability of RF to predict seasonal forest fire risk with only three variables related to precipitation. In fact, RF has shown that the precipitation is the variable that have the highest power to predict the severity of forest fires (83% of accuracy, p-value = 0.013). In addition, the study showed that the more precipitations in the months of February and March are abundant, the more likely the fires during the year are to be severe, and vice versa. Such relationship can be explained by the contribution of the rainfall to primary build-up and for fuel dryness. Furthermore, and through the recent development of seasonal climate forecasts in Morocco, it become easier for the forest manager to predict, since the previous year, the seasonal forest fire risk of the next period, and then lead safely all prevention strategies and best plan for rational uses of fires combatting means. Land use planning remains a very powerful tool to fight against wildfires. Indeed, it is well known that the presence of a wildland urban interface contributes significantly to the production of a large burned area. The presence of useful information about wildland-urban interface can be considered as a very powerful input for forest fire risk prevention surrounding urban areas. Consequently, spatial planning is one of the principal components that should be more developed in the coming studies in order to better design and plan urban and extra urban areas and increase the sustainability and territorial resilience at the large-scale level. The societal benefits of this study are highly promising since the study opens the possibility to preserve the lungs of residents in all surrounding areas at large scale level by anticipating forest fire risk and to improve human wellbeing especially in the urban zones by upgrading territorial resilience. Acknowledgements The authors would like to thank the Department of Forest and Water in Morocco and specially the Forest Protection Service for their support and assistance and for their appreciation of the benefits to be gained from this research.

References: 1. IPCC. AR5 Synthesis Report: Climate Change 2014—IPCC (2014). https://www.ipcc.ch/ report/ar5/syr/, Accessed 13 Aug 2021 2. HCEFLCD. Rapport annuel d’incendies de forêts et bases de données du Service de la Protection des Forêts. Bilan interne. Haut-Commissariat aux Eaux et Forêts et à la Lutte contre la Désertification, Rabat-Chellah, Maroc (2014)

Smart Prediction System for Territorial Resilience …

547

3. H. Mharzi Alaoui, et al., Mapping of fire intensity and of sensibility to crown fires in Mediterranean forests. Case of the province of Chefchaouen in northern Morocco (2017). http://hdl.handle.net/2042/62682 4. A. Benabid, Description de la composition et de la structure des types de combustibles. Appui à la mise en oeuvre du programme forestier national. Elaboration des cartes de risques aux incendies de forêts du Nord du Maroc. Projet UTF/MOR, FAO-Rabat, Maroc (2007) 5. F. Assali, H.M. Alaoui, M. Rouchdi, M. Badraoui, Modélisation et Cartographie du Risque d’éclosion d’incendie de forêt dans le nord-ouest du maroc (région de chefchaouen-ouazzane) (2016), p. 18 6. S. Goswami, J. Gamon, S. Vargas Zesati, C. Tweedie, Relationships of NDVI, Biomass, and Leaf Area Index (LAI) for six key plant species in Barrow, Alaska (2015). https://doi.org/10. 7287/PEERJ.PREPRINTS.913V1. 7. S. Goward, D. Dye, A. Kerber, V. Kalb, Comparison of North and South American biomes from AVHRR observations (1987). https://doi.org/10.1080/10106048709354079. 8. S.A. Sader, J.C. Winne, RGB-NDVI colour composites for visualizing forest change dynamics. Int. J. Remote Sens. 13(16), 3055–3067 (1992). https://doi.org/10.1080/ 01431169208904102 9. C.J. Tucker, Red and photographic infrared linear combinations for monitoring vegetation. Remote Sens. Environ. 8(2), 127–150 (1979). https://doi.org/10.1016/0034-4257(79)90013-0 10. E. Chuvieco, El factor temporal en teledetección: evolución fenomenológica y análisis de cambios (1998), p. 9 11. L. Breiman, Random forests. Mach. Learn. 45(1), 5–32 (2001). https://doi.org/10.1023/A: 1010933404324 12. Cutler, Random Forests For Classification In Ecology - Cutler - 2007 - Ecology - Wiley Online Library (2007). https://esajournals.onlinelibrary.wiley.com/doi/abs/https://esajournals. onlinelibrary.wiley.com/doi/abs/10.1890/07-0539.1, Accessed 13 Aug 2021 13. G. Biau, E. Scornet, A random forest guided tour. TEST 25(2), 197–227 (2016). https://doi. org/10.1007/s11749-016-0481-7 14. G. Biau, Analysis of a random forests model. J. Mach. Learn. Res. 13(38), pp. 1063–1095 (2012). Accessed 13 Aug 2021, http://jmlr.org/papers/v13/biau12a.html 15. A.M. Prasad, L.R. Iverson, A. Liaw, Newer classification and regression tree techniques: bagging and random forests for ecological prediction. Ecosystems 9(2), 181–199 (2006). https://doi.org/10.1007/s10021-005-0054-1 16. M. Wiesmeier, F. Barthold, B. Blank, I. Kögel-Knabner, Digital mapping of soil organic matter stocks using Random Forest modeling in a semi-arid steppe ecosystem. Plant Soil 340 (1), 7–24 (2011). https://doi.org/10.1007/s11104-010-0425-z 17. S.H. Welling, H.H.F. Refsgaard, P.B. Brockhoff, L.H. Clemmensen, Forest floor visualizations of random forests. CoRR, vol. abs/1605.09196, 2016, Accessed 13 Aug 2021. http:// arxiv.org/abs/1605.09196 18. J. Paul, M. Verleysen, P. Dupont, Identification of Statistically Significant Features from Random Forests (2013). https://www.semanticscholar.org/paper/Identification-of-StatisticallySignificant-from-Paul-Verleysen/e408c17a67dd4b194a2f4a1789607708662ff585

Mapping of the Study Area with GIS a Tool for the Description of Study Sites in Epidemiology Hajar El Omari, Abdelkader Chahlaoui, Fatima Zahra Talbi, Abdelkarim Taam, and Abdelhakim El Ouali Lalami

Abstract A scientific research project in epidemiology is a work that aims to present a solid and relevant research idea and in most cases, the analysis of the results requires a good study of the environment studied. Spatial epidemiology can be used to describe spatial variations in the study area and the risk factors for a disease in a given population, such as the rate of poverty or urbanization. With this in mind, the purpose of this research was to determine the value of GIS in describing the study area by creating a database containing several of the parameters in a geographic information system (GIS). This database was then exploited by a spatial and thematic analysis of some risk factors in the selected study area (prefecture of meknes), which allows an interpretation of the results, especially for epidemiological studies. The results obtained show the undeniable place of GIS in the representation of data containing geographical (prefectures), socio-economic (population and urbanization) and environmental (rivers) parameters on easy to interpret thematic maps. Keywords GIS


Epidemiology
Mapping
Study area

H. E. Omari (&)
A. Chahlaoui Natural Resources Management and Development Team, Laboratory of Health and Environment, Faculty of Sciences, Moulay Ismail University, Meknes, Morocco F. Z. Talbi Hassan First University of Settat, Faculty of Sciences and Technologies, Laboratory of Biochemitry, Neurosciences, Natural Resources and Environment, 577, Settat, Morocco A. Taam Laboratory of Engineering Sciences, National School of Applied Sciences (ENSA), Ibn Tofail University, Kenitra, Morocco A. E. O. Lalami Higher Institute of Nursing Professions and Health Techniques Fez, Regional Health Directorate, EL Ghassani Hospital, Fes, Morocco © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_44

549

550

H. El Omari et al.

1 Introduction Many diseases are linked, more or less directly, to the environment, and this link is particularly strong for human or animal diseases whose distribution is linked to environmental parameters [1, 2]. Geographic information systems (GIS) and the maps they produce are useful for strengthening the entire process of managing and analyzing epidemiological surveillance data. A GIS can store, manipulate, and geographically integrate large amounts of information from different sources, programs, and sectors as a starting point for eventual integration of surveillance activities. Standardized geo-referencing of data facilitates a rigorous approach to data management and analysis. Indeed, spatial epidemiology can be defined as the description and analysis of geographic variations in disease in relation to demographic, environmental, and socioeconomic risk factors [3]. It also allows the data to be structured in a logical way to address the issues. Geographic analyses have begun to emerge in different countries [4–6], in order to characterize the distribution of diseases and the risk factors that might explain it. [7–9]. Indeed, knowledge and control of the area of the diseases allow to highlight the causes and possible consequences of an extension to areas at risk [10–12]. With this in mind, the present study provides a spatial analysis of a study area through the creation of various maps. The maps produced can play an important role in guiding researchers to analyze their results in correlation with the environment.

2 Material and Methods 2.1

Study Area

To achieve our goal, we chose to work in the prefecture of Meknes, located in south-central Morocco in the Meknes-Fes region, on the Sais plateau between the Middle Atlas Mountains to the south and the Pre-Rifa Hills to the north. It is 140 km from the administrative capital Rabat and 60 km from the spiritual capital Fez. Its geographical coordinates are: Longitude: 5° 33', Latitude: 33° 52' and Altitude: 530 m.

Mapping of the Study Area with GIS …

2.2

551

Acquisition of Data

The socio-economic data used in this chapter are from the publications of the population censuses by the High Commission of the Plan [13].These include the number of the population, the poverty rate and urbanization.

2.3

Software Used

The formalization and use of a GIS require documents in digital format and computer equipment. Indeed to process images and maps for the elaboration of thematic maps, the following software was used: 3.0.2 QGIS, because of the availability of the product and the functions offered by the software (open source).

2.4

Map Creation

The methodology consists of crossing, by means of the geographic information system, the data plans relating to the number of the population, type of environment and the water network which makes it possible to obtain thematic maps. To convert the geographic data into computer format we used the tool ‘‘digitalization table’’. The digitization of the municipalities and water network associated with the prefecture of Meknes was done using the gisMap software. A database created containing the digital information was used to integrate the data relating to this study at the level of the attribute table of the Qgis.

3 Results and Discussion 3.1

Map of Study Area

The prefectural territory is divided into 21 communes: six urban communes and fifteen rural communes representing respectively 18.2% and 9.3% of the total of the same type of commune at the level of the Fez-Meknes region [14] (Fig. 1).

552

H. El Omari et al.

Fig. 1 Geographical situation of the study area

3.2

Population Map

The legal population of the Prefecture of Meknes reached 835,695 inhabitants in 2014 compared to 715,284 in 2004. Thus, it recorded an average annual growth rate of 1.6% for the period 2004–2014. The prefectural legal population represented 19.7% of the total regional legal population in 2014 [13, 15]. Figure 2 presents the number of inhabitants per commune in Meknes.

3.3

Urbanization Map

The urbanization rate in our study area is 75%, the demographic growth of the urban population is attributed in addition to the natural balance (birth deaths), to the extension of the urban perimeters of the major cities, the emergence of new urban centers and also the migration of rural people to the cities. In general, the prefecture of Meknes consists of urban communes (Meknes, Tulle, Ouislane, Boufkrane) and rural communes (Fig. 3).

Mapping of the Study Area with GIS …

Fig. 2 Number of inhabitants per commune in the prefecture of Meknes

Fig. 3 Type of environment (urban/rural) in the prefecture of Meknes

553

554

H. El Omari et al.

Fig. 4 Map of poverty rate by commune

3.4

Poverty Map

As for poverty in the prefecture of Meknes, it is 7.9% in rural communes and 1.5% in urban communes [14, 15], according to the 2014 poverty map, the poverty rate in the prefecture of Meknes stood at 2.6%, almost half of what is recorded at the regional 5.1% and national 4.8%. Indeed, poverty remains much more prominent in rural areas. The poverty rate in this area range from 4.6% to 28%. Figure 4 represents the distribution of the poverty rate of communes.

3.5

Map of the Water Network

The irrigation of urban and peri-urban agriculture in the city of Meknes is supported by three rivers (Fig. 5): Bouisshak in the west, Boufekrane in the center and Ouislane in the east. These wadis are characterized by a relatively low flow rate, which leads to a dependence on other resources for irrigation, in particular: raw or mixed wastewater. These different maps, as well as others that can be made using geographic information systems, could be a tool for analyzing the risk factors of a disease by explicitly including spatial information in our research. Indeed, spatial data analysis differs from classical data analysis [16], because ‘‘each phenomenon is related to all the others, but phenomena close in space will tend to be more related than distant

Mapping of the Study Area with GIS …

555

Fig. 5 Water network map

phenomena’’. This notion of spatial dependence may be less intuitive than temporal dependence, which is based on the notion of the past (a natural order relationship) and in which anteriority often implies causality. Nevertheless, many ecological and epidemiological processes may be responsible for the existence of spatial structure in health data [17]. The value of a variable in a given geographical position is thus often correlated with those in neighboring positions [18]. GIS allows a graphical representation of information whose objective is to provide an understanding of the content of information, to produce visual representations of raw data, which allows a good analysis of the study area and subsequently a good interpretation of the results. Indeed, in many scientific studies, the multitude of variables analyzed can make understanding difficult. GIS is therefore an effective tool for synthesizing information and facilitating the overall understanding of the study. GIS is particularly effective for illustrating the descriptive steps and impact analysis. Several recent researches have used GIS to represent their study environment and their results in different fields such as epidemiology [19–21]. Geographic Information Systems (GIS) have created new opportunities for epidemiologists to take into account the spatial dimension of diseases, to identify risk areas or to study the role of environmental factors [19, 20].

556

H. El Omari et al.

GIS facilitates the reconciliation of multi-sectoral data such as epidemiological surveillance data, demographic data, environmental information, and other resources into a common analysis base [20–22]. The Geographic Information System (GIS) is used to characterize the health-environment relationship and social inequality. The use of GIS is emerging as a relevant choice when it comes to processing spatialized data. The integration of data across different information layers allows for rigorous spatial analysis. This analysis by crossing information, if it can be done visually (like layers superimposed on each other) often requires crossing with alphanumeric information. In our case, we crossed the Geographic data (limits of municipalities) and the number of the population, the type of environment (urban or rural) and the water network this is an example of sophisticated analysis that allows the use of a GIS. This tool allowed us to create thematic maps that facilitate the analysis and interpretation of the results.

4 Conclusion The mapping of risk factors by computer tools (geographic information systems (GIS) provides epidemiologists with new opportunities to take into epidemiologists new opportunities to take into account the spatial dimension of diseases, identify risk areas or study the role of environmental factors.

References 1. S.I. Hay, R.W. Snow, D.J. Rogers, S.E. Randolph, From predicting mosquito habitat to malaria season using remotely sensing data: practice, problems and perspective. Parasitol. Today 14(8), 306–312 (1998) 2. P.J. Curran, P.M. Atkinson, G.M. Foody, E.J. Milton, Linking remote sensing, land cover and disease. Adv. Parasitol. 47, 37–80 (2000) 3. P. Elliott, D. Wartenberg, Spatial epidemiology: current approaches and future challenges. Environ. Health Perspect 112, 998–1006 (2004) 4. F. Talbi et al., Thematic maps of the impact of urbanization and socioeconomic factors on the distribution of the incidence of cutaneous leishmaniasis cases in Sefrou Province, Central North of Morocco (2007–2011). Interdisc. Perspect. Infect. Dis. 2020, 1–8 (2020) 5. F. Rakotomanana et al., Approche géographique dans la lutte contre le paludisme dans la région des Hautes Terres Centrales à Madagascar. Archives de l’Institut Pasteur de Madagascar 67, 27–30 (2001) 6. C.W. Kabaria et al., Mapping intra-urban malaria risk using high resolution satellite imagery: a case study of Dar es Salaam. Int. J. Health Geograph. 15(1), 1–12 (2016) 7. J.F. Guégan, M. Choisy, Introduction à l’épidémiologie intégrative des maladies infectieuses et parasitaires, LMD.Licence Maîtrise Doctorat.Cours Biologie. De Boeck, Bruxelles (2009) 8. H. El Omari, A. Chahlaoui, A. EL Oualialami, M. Khafou, The contribution of geographic information systems in the fight against parasitic diseases: the case of Leishmaniasis, in SCA '18: Proceedings of the 3rd International Conference on Smart City Applications, Tetouan, Morocco (2018), pp. 1–5

Mapping of the Study Area with GIS …

557

9. H. El Omari, A. Chahlaoui, A. Taouraout, A. El Ouali Lalami, Geographical information systems (GIS) and epidemiology of vector diseases: case of leishmaniasis in the Fez-Meknes, region of Morocco, in SCA '19: Proceedings of the 4th International Conference on Smart City Applications, Casablanca, Morocco (2019), pp. 1–4 10. G.L. Brody, A. Aschengrau, W. McKelvey, R.A. Rudel, C.H. Swartz, T. Kennedy, Breast cancer risk and historical exposure to pesticides from wide-area applications assessed with GIS. Environ. Health Perspect. 112, 889–897 (2004) 11. J.K. Rodgers, K.J. Bergmann, V.L. Salak, D.T. Lackland, V.K. Hinson, Geographic distribution of parkinson’s disease and stroke in South Carolina. Ann. Epidemiol. 17, 723 (2007) 12. C.B. Ghetian, P.R. Ghetian, J.E. Volkman, E.J. Lengerich, Cancer registry policies in the United States and geographic information systems applications in comprehensive cancer control. Health Policy 87, 185–193 (2008) 13. Haut Commissariat au Plan (HCP). Recensement général de la population et de l’habitat (2014). http://rgph2014.hcp.ma/downloads/Publications-RGPH-2014_t18649.html, Accessed 10 Aug 2021 14. Haut Commissariat au Plan (HCP). Recensement général de la population et de l’habitat (2014). Annuaire Statistique de la Région Meknès-Tafilalet (2014) 15. Haut Commissariat au Plan (HCP).Recensement général de la population et de l’habitat (2007). https://www.hcp.ma/Demographie-population_r142.html, Accessed 10 Aug 2021 16. T.P. Robinson, Spatial statistics and geographical information systems in epidemiology and public health. Adv. Parasitol. 47, 81–128 (2000) 17. R.S. Ostfeld, G.E. Glass, F. Keesing, Spatial epidemiology: an emerging (or reemerging) discipline. Trends Ecol. 20, 328–336 (2005) 18. A. Banos, A propos de l’analyse spatiale exploratoire des données. Cybergeo Eur. J. Geogr. Systèmes Modélisation Géostatistiques (2001) 19. F. Talbi et al., Cartography and epidemiological study of leishmaniasis disease in Sefrou Province (2007–2010), Central North of Morocco. Interdisc. Perspect. Infect. Dis. 2020, 1–8 (2020) 20. F.Z. Talbi, A. Janati Idrissi, A. Sandoudi, A. El Ouali Lalami, Spatial distribution of incidence of leishmaniasis of different communes of Sefrou Province (2007–2010), central north of Morocco, in Proceedings of the 4th International Conference on Smart City Applications (SCA’19), Casablanca, Morocca (2019), pp. 1–8 21. H. El Omari, A. Chahlaoui, F. Talbi, K. Ouarrak, A. El Ouali Lalami, The contribution of cartography in risk management of vector-borne diseases: cas of leishmaniasis in the Fez-Meknes, Region of Morocco, in Innovations in Smart Cities Applications Edition 3. ed. by M.B. Ahmed, A.A. Boudhir, D. Santos, M. El Aroussi, I.S. Karas (2020), pp. 1192–1201 22. H. El Omari, A. Chahlaoui, A. El Ouali Lalami, The geographic information systems are a lever for fighting parasitic diseases: case of Leishmaniasis, in Innovations in Smart Cities Applications Edition 2: The Proceedings of the Third International Conference on Smart City Applications. ed. by M.B. Ahmed, A.A. Boudhir, A. Younes (Springer International Publishing, Cham, 2019), pp. 1204–1213. https://doi.org/10.1007/978-3-030-11196-0_98

Geodesign – a New Approach for Rapid Development of Planning and Carbon Sequestration Scenarios Fred Barış Ernst , Abdullah İzzeddin Karabulut , and Mehmet İrfan Yeşilnacar

Abstract Turkey is one of the countries that is already greatly affected by climate change. Geodesign is an emerging planning approach based on GIS that can address climate. It uses a participatory approach and is very adaptable to changed requirements. In this study, development scenarios for Harran district located in the Southeastern Anatolian Project (GAP) one of the world's biggest irrigation projects are presented. These scenarios show how the big potential in the tourism sector building on the region's huge archaeological heritage and the agricultural sector can be deployed to the benefit of a rapidly growing population and carbon sequestration can be achieved. During the Geodesign process, different scenarios are created by the different interest groups. The scenarios are compared with each other and using tools of the web based GeodesignHub software until a solution was found that satisfied the needs of all groups. The design of three different scenarios for a time horizon until 2035 and 2050 was required: 1) A non-adopter scenario, which presumed that current unsuitable practices would continue, 2) An early-adopter scenario, in which new environmentally friendly technologies would replace unsuitable practices immediately and 3) A late-adopter scenario, in which such technologies would be deployed with a delayed start. Results show that for 2050 under optimum conditions 23 386 tons of carbon sequestration in Harran district and 12 400 000 tons at the national level could be achieved for the agricultural sector alone. Keywords Geodesign


GIS
Scenarios
Carbon sequestration

F. B. Ernst (&) Department of Geomatics Engineering, Harran University, Şanlıurfa, Turkey A. İ. Karabulut Graduate School of Natural and Applied Sciences, Harran University, Remote Sensing and Geographic Information Systems, 100/2000 CoHE PhD Scholarship, Sanliurfa, Turkey M. İ. Yeşilnacar Department of Environmental Engineering, Faculty of Engineering, Harran University, 63050 Sanliurfa, Turkey e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_45

559

560

F. B. Ernst et al.

1 Introduction Climate change has become a topic on the world-wide political agenda that is gaining momentum day by day. For example, in a recent questionnaire conducted by the Munich based Security Conference interviewed persons of countries like China, Germany and the UK voted for “Climate Change general” being the highest overall risk for their country (Bunde et al. 2021). Turkey is one of the countries that is already greatly affected by climate change (MOEU 2019). According to the climate change calculator “Climate Change Scenarios” hosted by the National Center for Atmospheric Research (UCAR 2021) that has contributed its data to IPCC” s reports certain parts of Turkey could expect a decrease of precipitation up to 35 mm in the coming years. Since more than 30 years, Geographic Information Systems (GIS) have been successfully deployed for planning and management of cities as part of Urban Information Systems (UIS). However, when used as a tool in the planning process they have some limitations: 1) They require highly skilled human resources from the field of Geomatics. 2) One of their most important outputs, maps, are usually hardly understood by those that are meant to implement the planning – decision-makers. 3) As a result of point 1 and 2, participation of all stakeholders in the planning process has been very limited in the past. Geodesign is an emerging planning approach based on GIS that aims to overcome these shortcomings. It is a recent extension of the early works of McHarg (1969) who has been a pioneer in the field of landscape architecture and regional planning. Today, the analytical power of GIS and its huge underlying scientific databased with the intuitive tools of the designer, architect or planner can be combined. In 2012, Carl Steinitz laid the theoretical framework with his book “A Framework for Geodesign” (Steinitz 2012). Geodesign projects have implemented all around the world. In Singapore, the Urban Redevelopment Agency charged AECOM with reevaluating its long-term land use strategy based on the company’s Sustainable Systems Integration Model. The calculation of impacts on GHG emissions was a fundamental part of this reevaluation. To prepare for Florida’s future, its Regional Planning Council used the Geodesign methodology to develop four different scenarios and synthesize them into the Regional Growth Vision 2050 (MacEllvany 2012). In Turkey, a new masterplan for Harran University’s main campus was drawn with the active participation of the university’s higher management (Ernst et al. 2019). In addition, more than 100 projects with scientific background have been completed under the umbrella of the International Geodesign Collaboration (IGC) whose purpose is to foster research and education in the field of Geodesign (https://www.igc-geodesign. org/). The “TRILLION TREES INITIATIVE” (http://trilliontreesinitiative.com/) is one of the most important international initiatives trying to mitigate the impact of climate change. To support this initiative, at IGC summit 2021, sequestering of carbon had to be estimated for all submitted projects. Harran University presented a

Geodesign – a New Approach for Rapid Development …

561

research undertaken in the Southeastern Anatolia Project (known as GAP in Turkey) region where identifying the land resources is crucial for agricultural and water management purposes (Yesilnacar and Cetin 2005). Development scenarios have been designed showing how the big potential in the tourism sector deploying this glorious past and the agricultural sector can be deployed to the benefit of a rapidly growing population (Karabulut et al. 2021) and at the same time achieving a maximum amount of Carbon sequestering.

2 Methodology The research area covers an area of 1600 km2 and according to the requirements of IGS forms an exact square. However, due to the partial unavailability of some data the estimation of carbon sequestration was carried out only for Harran district itself. The methodology used in this study followed closely the framework of Geodesign according to Steinitz (2012) as summarized in the figure below (Fig. 1). The whole framework is made up of 6 models that are worked on three times during three different phases. During the first phase, a general understanding of the project area was reached and project goals were defined as development of scenarios for Harran district with focus on agricultural, archaeological and tourism systems. During the second phase, the exact tools and ways to conduct the study had to be decided on. They can be summarized as follows: 1) Study of previous development and master plans for this region especially: HARRAN YÖNETİM PLANI 2016 - 2021. AnaDOKU. Şanlıurfa. 2016 and GAP Bölgesi Turizm Master Planı –Planlama Kararları. 2) Collection of data and building a comprehensive GIS database. 3) Creation of evaluation models (suitability maps). For this, different GIS software (ArcGIS and QGIS) was used.

Fig. 1 Geodesign framework by Carl Steinitz (2012)

562

F. B. Ernst et al.

4) Creation of changes models, computation of impact models and representation of decision models. For this, the online software GeodesignHub was used. During the third phase, based on the results of the first and second phases, the real project works were carried out. In the reminder of this paper, only steps three and four of the third phase will be explained in detail.

2.1

Evaluation Models

During this step, it is evaluated how well the research area is performing currently. Usually, this is done by conducting a multi-criteria analysis (MCA) that deliver suitability maps for the systems that are relevant for a specific area. Whereas eight systems had been predefined by the standards of IGC (agriculture, energy infrastructure, green infrastructure, industry and commerce, institutional, residential mixed, transport infrastructure, water infrastructure) two additional ones could be added to represent local conditions (in our case, tourism and archaeology). As an example, the results for the system “agriculture” are shown in Fig. 2. The main objectives for changes in this system was to reach a percentage of 20 for the conversion of crops to orchards on irrigated lands.

2.2

Change, Impact and Decision Models

In contrast to the previous works the setting up of change, impact and decision models refer to the envisioned future state of the research area. These works were

Fig. 2 Suitability map for the system “agriculture”. The research area is shown as a rectangle against the background of a Google Earth image. Şanlıurfa is located in the upper left corner and the sharp line at the bottom represents the border to Syria

Geodesign – a New Approach for Rapid Development …

563

realized using the web-based software GeodesignHub (Ballal 2015). It can be categorized as a Public Participation GIS (PPGIS) that by means of a user-friendly interface can be handled by anyone who has some basic understanding of maps. Web based technology allows for the interaction of many users simultaneously. In fact, due to the pandemic all works for these three models had to be performed online in a simulated workshop during a master class. Change models consist of planned changes (projects) to be made for each of the ten mentioned systems like e.g. planting of new orchards for the system “agriculture”. The following change requirements that should be met: 1) Convert 20% of irrigable land into orchards as originally planned (USIAD 2008), 2) manage water resources including agricultural drainage water carefully, 3) enhance the touristic infrastructure (boutique hotels, hiking and biking), 4) preserve the archeological heritage, 5) reach 3 million of tourists/year by 2050 with 500,000 overnight stays, 6) build a satellite city outside of the fertile Harran plain for 50,000 inhabitants, 7) connect Harran and its satellite city by means of light rail with Şanlıurfa,and 8) make Harran district energy self-sufficient based on photovoltaic plants. For the first requirements, suggested that pecan trees (Carya illinoinensis) should be planted because they are well adapted to the climatic conditions of the region. According to our estimations they could generate revenues 30 times higher than cotton, one of the main crops currently cultivated, and can sequester a high amount of carbon in a relatively short period of time (3.5 kg/year). Suitability maps serve the selection of appropriate sites for new projects. In a second step, these projects are evaluated in terms of their impact on their own

Fig. 3 Selected new projects in the surrounding of Harran city and their cross-impact analysis

564

F. B. Ernst et al.

system and across other systems. Such a “cross system impact analysis” will reveal whether the planned orchards will have a negative or positive impact on the other systems (Fig. 3). Based on the instant feedback of this impact analysis, proposed changes can be edited or removed entirely if the results are not satisfactory. During decision modeling, different scenarios were created by two simulated interest groups working under the guidance of several teachers (developer and an ecologist group). Using online tools for comparison it was tried to come up with a solution that satisfied the needs of both groups. Adhering to the framework of IGC three different scenarios for the time horizons of 2035 and 2050 were designed: 1) A non-adopter scenario, which presumed that current unsuitable practices would continue, 2) an early-adopter scenario, in which new environmentally friendly technologies would replace unsuitable practices immediately and 3) A late-adopter scenario, in which such would be deployed with a delayed start.

2.3

Estimation of Carbon Sequestration

The methodology to estimate carbon sequestration for the above-described different scenarios consisted of the following five steps: 1. Development of baseline measures. The baseline is made up of the amount of carbon storage in the absence of any new projects, which equals the Non-Adaptor scenario. To calculate this baseline, different methods concerning forests, orchards or single trees in urban areas have to be applied. For working in urban areas, tools like i-Tree Canopy (https://canopy.itreetools.org/) are available that can speed up this process. For our research area, we used data from Turkish Statistical Institute (2020). Interpolating from a provincial to a district level resulted in about 2000 walnut trees for the Harran district. Walnut trees (Juglans regia) were used because their characteristics are similar to those of pecan trees in terms of carbon sequestering. 2. Determination of change in tree numbers for different scenario. Changes in tree numbers are determined by foreseen changes in the respective land use classes (in this research only “agriculture”) for the different scenarios. The goal of having 20% of the irrigated land converted to orchards would only be reached under optimum conditions in the EA scenario. For any scenario the combination of effects like conservation, tree felling (as a result of occurred fires, pests, etc.) without replacement, tree felling with replacement and addition of new tees has been shown. Tree felling was not in issue for the given time span. These facts have been taking into consideration when tables like the table for the NA scenario shown below have been created (Table 1).

Geodesign – a New Approach for Rapid Development … Table 1 Tree count for the Non-adopter scenario (project level) (in thousands)

565

Non-adopter scenario

2020

2035

2050

Conserved Lost/harvested Replaced Added Count

2 0 0 1 3

3 0 0 0 3

4 0 0 0 4

3. Interpolation of results to the national level. During this step the results of this project are scaled-up to the national level. Only changes in land use classes that have been worked during this project have been addressed (i.e. “agriculture”). The goal of this estimation is to understand the dimension of the possible impacts of such changes and precision is not required. The calculation under this step has been performed as follows: • Instead of using pecan trees walnut trees have been chosen because they can be cultivated nearly in all provinces of Turkey. Only provinces with a high potential (more than 1000 tons) (Cografya Harita 2019) have been included. • In order to attain a high productivity only irrigated land has been included. • As a baseline, data from Turkish Statistical Institute (2020) have been used. 4. Calculation emission changes at the project level. As recommended by IGC carbon sequestration at the project level has been calculated using a guide of U.S. Department of Energy (1998). Older trees reaching greater heights and deeper roots have a much higher biomass than younger trees thus, holding a much higher carbon stock. Whereas younger trees show a much higher growth rates and accumulate carbon much faster than older ones resulting in the following values: • • • •

22 kg/year carbon sequestered by 1 mature tree(older than 15 years) 3% carbon increase per year by 1 mature tree 3.5 kg/year carbon sequestered by 1 small tree (up to 15 years) 14% carbon increase per year, young trees

5. Assessment of carbon contribution at the national level. Here again, the goal of this assessment is to understand the dimension of the contribution at the national level and precision is not required. The above-mentioned values were used for this assessment Then, for each scenario the achieved carbon sequestration has been divided by the population numbers for the respective time horizon. To compare these savings with CO2 emissions per capita data have been retrieved from EDGAR (2021), which indicated 5.01 t CO2/cap. for Turkey in the year 2019. A population growth of 1,39% for Turkey has been used (Turkish Statistical Institute 2020).

566

F. B. Ernst et al.

3 Results and Discussion Following the Geodesign approach three different scenarios (NA, LA, EA) for 2035 and 2050 for Harran district have been generated. They are the result of the selection from a pool of a total of 123 projects and their combination according to the assumption for these scenarios and the above-mentioned requirements. In the worst-case scenario (NA 2050), none of these requirements would be met as bad practices would continue. Consequently, the map in Fig. 4a with exception of unplanned urban development does not show any change. This will cause an occupation of fertile lands and an encroachment of the archaeological site thus, limiting its potential for tourism severely. As Harran has the highest population growth rate in Turkey (Karacadağ Development Agency 2018) doubling of the population without adequate economic development the migration to megacities will continue. In the most optimistic scenario (EA 2050), all the above-mentioned requirements would be met. Besides agriculture tourism will become the major source of income, which is in accordance with the “Harran Management Plan” (Anadoku 2016). As can be seen in Fig. 4b, the proposed changes have a striking impact on the future spatial pattern characterized by conversion to pecan orchards, management of rangelands, building of a satellite city and a new infrastructure for tourism. In this scenario, no migration would become necessary. Here, only the two most extreme scenarios could be addressed. All other scenarios can be found at the IGC home page under https://www.igc-geodesign.org/igc-2020-21-projects. Naturally, the above-described changes in land use patterns would have a very different impact on carbon storage. Figure 5a shows the numbers for carbon conserved, lost, replaced, added and its total accumulation for NA 2050 and Fig. 5b for EA 2050 scenarios. The huge increase in accumulation for EZ 2050 can be attributed to the plantation of 1 million pecan trees. The total storage of carbon per person interpolated for whole Turkey for the most optimistic scenario has been computed as 0,1 of carbon T/capita equaling 12 400 000 tons in total compared with the current number of 5 T/capita of carbon emission.

Fig. 4 a and 4 b: NA and EA scenarios for 2050. For better display classes “tourism” and “archaeology” have been combined and a separate class “orchards” has been added

Geodesign – a New Approach for Rapid Development …

567

Fig. 5 a and 5 b: Carbon (tons/year) in the research area for NA 2050 and EA 2050 scenarios

The most important advantage of the Geodesign methodology used in this study is its fast adaptability to changed conditions. If requirements are changed (e.g. 25% of new orchards instead of 20%) and therefore, projects have to be edited in terms of size or location and consequently an impact analysis can be run on the spot. It qualifies this methodology to be utilized during workshops with the participation of decision-makers who expect to learn about new results immediately. One of the shortcomings was that until now, is has been mainly an academic endeavor because due to the pandemic the planned workshops to be hold in cooperation with the municipality of Harran and all other stakeholders could not be realized yet. Workshop are a crucial part of the Geodesign process because successful implementation of plans can only be attained by means of participation and its acceptance by these stakeholders. Another issue is the enacting of regulations for water consumption and subsidies for agricultural crops and their enforcement. Pecan trees do not generate income during the first years. Therefore, as long as the cultivation of cotton will be highly subsidized (up to 9600 TL/ha) (Turkish official gazette 2019) and no financial and educational support is given for farmers that are interested in planting pecan trees the proposed scenarios cannot put into practice. And, although recently new plantations of walnuts have been supported still, a great gap between production and demand exists (TEPGE 2020). Becoming a walnut exporter would mean a big increase in carbon sequestration. The number of 0,1 T/cap potential carbon savings compared with 5 current emission appears to be very low. However, one has to keep in mind that this saving is accomplished by using only 6% (Republic of Turkey Ministry of Agriculture and Forestry 2021) of Turkey's area. In this study, the Geodesign methodology has been applied to a rural area. Of course, the described methodology can be used for bigger cities as many related studies have been proven. It has to be questioned, how effective the concept of carbon sequestration by means of planting trees can be for Turkish cities if current policy priorities will not change. For example, with a percentage of 2, 2% public

568

F. B. Ernst et al.

green space Istanbul ranges 40th compared with other big cities in the world (WORLD CITIES CULTURE FORUM 2021). Akyel and Kenar (2021) concluded that in all Turkish urban forests 41,148 tons of carbon are stored. Compared with the amount of 12 400 000 tons that could be sequestered in the agricultural sector the contribution of urban green spaces to carbon sequestration seems to be very limited.

4 Conclusion Climate change is progressing and is already showing it's negative effects on urban and rural environments in Turkey in the form of heat waves, flooding and water shortages. At the same time, in places like the GAP region inappropriate agricultural techniques cause unnecessary highwater consumption leading to the salinization of soils. In order to overcome this situation smart solutions are required that support the planning at the city, regional and national level. Such solutions must be based on spatial information systems allowing for the participation of all stakeholders and offering tools that show the impacts of different planning scenarios immediately. Only in this way, there exist a chance that development plans are really implemented and ambitious goals for CO2 reduction can be achieved in time. The Geodesign methodology that has been demonstrated in the study of creating development scenarios for Harran district has at its disposal all the required features of an advanced spatial planning tool. While especially municipalities and regional councils all around the world have made very positive experiences with Geodesign in planning efforts addressing climate change in Turkey, it has been introduced to academic and local administration community just recently. Because a Geodesign study addressing carbon sequestration has been carried out in Turkey for the first time further research is required. Especially, the underlying assumption for the computation of carbon sequestration have to be analyzed carefully and changed if necessary. Until now, this study has been mainly an academic endeavor. To complete the Geodesign based planning process workshops with all stakeholders especially the local administration will have to be conducted. Experience shows that such workshops, in which some unpopular measures are suggested, cannot be conducted online. So that these workshops will be more appealing to all interest groups an in-depth financial analysis of all five scenarios is currently carried out at Harran University.

References Akyel and Kenar 2021 E. Akyel, F. Kenar, Analysis of carbon absorption amount of urban forests by spatial interpolation

Geodesign – a New Approach for Rapid Development …

569

methods, in 2nd Intercontinental Geoinformation Days (IGD), Mersin, Turkey, 5–6 May 2021 (2021), pp. 183–185 Ballal 2015 H. Ballal,, Collaborative Planning with Digital Design Synthesis. Doctoral Dissertation. University College London, United Kingdom (2015) Harita 2019 Cografya Harita, Türkiye Ceviz Üretim Haritası (2019). http://cografyaharita.com/haritalarim/ 4cturkiye-2019-ceviz-uretim-haritasi.png, Accessed 12 June 2021 Emissions Database for Global Atmospheric Research (EDGAR) Emissions Database for Global Atmospheric Research (EDGAR): Country Fact Sheet Turkey. https://edgar.jrc.ec.europa.eu/country_profile/tur Ernst, et al. 2019 F.B. Ernst, et al., Geodesign for urban planning: a case study from Harran University’s Campus master plan. Int. J. Environ. Trends (IJENT)3(1), 17–30 (2019). ISSN: 2602–4160. Anomalies and Variability, and Uncertainty in Space and Time with The Climate Inspector Explore Climate Anomalies, Variability, and Uncertainty in Space and Time with The Climate Inspector | GIS Climate Change Scenarios (Ucar.Edu) Yönetim and Plani 2016 Harran Yönetim Plani 2016–2021. Anadoku, Şanlıurfa (2016) Şanlıurfa 2018 İstatistiklerle Şanlıurfa. Karacadağ Development Agency. Diyarbakır (2018) Karabulut et al. 2021 A.I. Karabulut, S. Benek, F.B. Ernst, Planning of eyyübiye district center (şanliurfa) using the geodesign method. J. Geogr. 42, 251–269 (2021). https://doi.org/10.26650/JGEOG2021-897149 Mcelvaney 2012 S. Mcelvaney, Geodesign – Case Studies in Regional and Urban Planning (ESRI Press, Redlands, 2012) Mcharg 1969 I. Mcharg, Design with Nature (Doubleday & Co., Garden City, 1969) MOEUTurkey's Fourth Biennial Report 2019 MOEU Turkey's Fourth Biennial Report, Ministry Of Environment And Urban Planning(Report), December, Ankara, Turkey (2019) Republic Of Turkey Ministry Of Agriculture And Forestry 2021 Republic Of Turkey Ministry Of Agriculture And Forestry. https://corinecbs.tarimorman.gov.tr/ corine, Accessed 20 June 2021 Steinitz 2012 C.A. Steinitz, Framework for Geodesign (ESRI Press, Redlands, 2012) Ekonomi and ve Politika Geliştirme Enstitüsü (TEPGE) 2020 Tarimsal Ekonomi ve Politika Geliştirme Enstitüsü (TEPGE). Tarım Ürünleri Piyasaları - Ceviz, Temmuz-2020, Tarım Ürünleri Piyasa Raporu.Pdf (Tarimorman.gov.tr) (2020)

570

F. B. Ernst et al.

Tobias 2021 T. Bunde, et al., Munich security report (2021) Between states of matter – competition and cooperation, in Munich Security Conference (MSC), Munich (2021) Turkish official gazette 2019 Turkish official gazette (2019). https://www.resmigazete.gov.tr/eskiler/2019/10/20191024-3.pdf, Accessed 01 June 2021 Statistical and Institute 2020 Turkish Statistical Institute. https://data.tuik.gov.tr, Accessed 12 June 2020 U. S.Department Of Energy 1998 U.S. Department Of Energy. Method For Calculating Carbon Sequestration by Trees in Urban and Suburban Settings (1998) USIAD. GAP Raporu GAP’ta Ne Oldu 2008 USIAD, 2008, GAP Raporu GAP’ta Ne Oldu. Bölgede Ekonomik, Stratejik ve Siyasal Gelişmeleri Yayınlayan: USİAD Ulusal Sanayici Ve İşadamları Derneği Editör: Dursun YILDIZ Hazırlayan: ADA Strateji 0312. 417 0041 (2008). ISBN 978–975–98399–3–2 % Of Public Green Space (Parks and Gardens) 2021 World Cities Culture Forum, % Of Public Green Space (Parks and Gardens) (2021). http://www. worldcitiescultureforum.com/data. Accessed 05 June 2021 Yesilnacar and Cetin 2005 M.I. Yesilnacar, H. Cetin, Site selection for hazardous wastes: a case study from the GAP area, Turkey. Eng. Geol. 81(4), 371–388 (2005)

Assessment of Rapid Urbanization Effects with Remote Sensing Techniques Nur Yagmur , Adalet Dervisoglu , and B. Baha Bilgilioglu

Abstract Istanbul is the most populous city in Turkey. The population, which was approximately 5.5 million in 1985, has reached 15.5 million in 2020. Population growth is the most important factor behind human activities that put pressure on the environment. An increasing population means depletion of limited resources, increasing environmental problems, and rapid urbanization. In parallel with the increase in population and urbanization, there has also been an increase in demand for housing, leading to new residential areas in almost every district of Istanbul. This study examined the transformation from vegetation areas to residential areas between 1985 and 2020 in a selected region in Buyukcekmece, one of the 39 districts of Istanbul. The relationship between land use and land cover (LULC) change in the area and Land Surface Temperature (LST) change caused by urbanization was analyzed. It is seen that the built-up area has increased from 57.1 ha to 781.4 ha in 35 years. In every five years, an increase in surface temperatures was determined in parallel with increasing urbanization, and this increase was determined as about 5.4°C from 1985 to 2020. Also, when the temperature data of the Buyukcekmece Meteorological station is analyzed, it is seen that there has been an increase of approximately 2 ºC in air temperatures in the last five years. In addition, movements were observed in the stability of structures in rapid urbanization areas after analyzing with the PSI time series InSAR method. The main causes were determined as construction sites around the buildings and geological conditions of the ground, which are triggered by urbanization. Keywords Rapid urbanization


NDVI
LST
PSI
Istanbul

N. Yagmur (&)
A. Dervisoglu
B. B. Bilgilioglu Civil Engineering Faculty, Geomatics Engineering Department, Istanbul Technical University, 34469 Istanbul, Turkey e-mail: [email protected] N. Yagmur Engineering Faculty, Geomatics Engineering Department, Gebze Technical University, 41400 Kocaeli, Turkey B. B. Bilgilioglu Faculty of Engineering and Natural Sciences, Geomatics Engineering Department, Gumushane University, 29000 Gumushane, Turkey © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_46

571

572

N. Yagmur et al.

1 Introduction Environment and natural resources are the most basic requirements for the continuation of life on Earth and are adversely affected as consumption needs increase and demands change due to the increasing population. The world has been changing very rapidly in recent years due to human activities. One of these activities is rapid urbanization, which destroys natural resources such as water bodies [1], green spaces [2], air quality conditions [3] and changes the weather conditions (e.g., an increase in air temperature, decreased precipitation) [4]. Especially with the increase in the popularity of megacities, these cities receive immigration, and the need for shelter arises due to the increasing population. The need for shelter causes rapid urbanization, land use and land cover (LULC) are changing, natural surfaces are turning into impermeable surfaces with high thermal capacity. Remote Sensing is an effective tool with the capacity for large-scale area observation to monitor urban area expansion and the stability of structures in these urbanization sites in addition to observing natural changes on Earth. With high-frequency data acquisition and large-scale monitoring, changes on the Earth’s surfaces can be monitored and be detected via active and passive satellite sensors. The Landsat mission provides optical (since 1972) and thermal (since 1982) satellite imagery that enables change detection to be monitored. In this study, the change of LULC in 35 years, in five-year periods in the selected area in Buyukcekmece district of Istanbul Megacity was investigated. Rapid urbanization in the region, a green area in 1985, has destroyed especially forest and vegetation areas. Such land cover change is the leading cause of urban heat islands and adversely affects human health and living standards [2, 5]. Besides, land surface temperature (LST) was calculated to show the urban heat intensity’s effect using Landsat thermal bands. LST is directly related to the land use and land cover classes and shows low temperature in green spaces, unlike urban areas [5, 6]. This study examined the relationship between the LULC change in the region and the LST change caused by urbanization. Urban ground subsidence is common phenomenon in cities with dense urbanization, and increased urban expansion triggers land subsidence [7]. The stability of the structures can be affected by various factors like earthquakes [8], construction sites around the buildings [9], and geological conditions of the ground [10] in dense urbanization sites. Synthetic Aperture Radar Interferometry (InSAR) is an ideal solution for observing surface deformation, and it is possible to monitor spatial extension and temporal evaluation of subsidence, especially in the reclamation and rapid urbanization areas. InSAR has been widely used for deformation monitoring of rapid urbanization in the studies [11–13]. Compared with traditional methods such as leveling and global positioning systems (GPS), InSAR has advantages with short revisit times and large-scale monitoring, and these types of passive systems are not affected by weather conditions. Furthermore, it measures surface deformation at millimeter accuracy at both regional and local scales [11]. Especially time series InSAR methods such as Persistent Scatter Interferometry (PSI) and Small

Assessment of Rapid Urbanization Effects …

573

Baseline Subsets (SBAS), which are commonly used, are preferred to detect slow movement. PSI identifies the movement of pointwise targets called “persistent scatterers” like buildings, whereas the SBAS method is more suitable for rural areas [12]. In the study area, the PSI time series InSAR method was selected for observing surface deformation due to its ability to detect urban subsidence, and results were evaluated. In summary, this study investigated two issues: (i) the effects of rapid urbanization by evaluating temporal change analysis of LULC and LST, (ii) stability of the structures in rapid urbanization sites with time series InSAR analysis. Optical bands of Landsat were used for temporal change analysis, and thermal bands were used for LST analysis. The normalized difference vegetation index (NDVI), the most widely used plant index, was used to determine the temporal change. Sentinel-1 SAR images were used for InSAR analysis. All the obtained results were evaluated together.

2 Related Works With classification methods and spectral indices, optical remote sensing data has been widely used to detect urban extension. Besides that, many studies have presented urban growth effects on the LULC change using thermal bands. Bhatta et al. (2010) used three Landsat temporal images at 15-year intervals (1975, 1990, and 2005) to determine the urban extent and growth of Kolkata-Howrah [14]. Xiong et al. (2012) investigated the spatiotemporal variations in the LST and LULC types over different urban/rural zones to determine the impacts of rapid urbanization using quantitative thermal remote sensing and spatial statistics methods. The normalized difference built-up index (NDBI) and NDVI were applied to determine the LULC variation in the study, in which Landsat images from four different dates were used. It was seen that while LST was related positively with NDBI and negatively with NDVI [15]. Villa (2012) used Landsat images to determine urban growth in terms of impervious surface expansion of Milan city over 20 years (1984–2003). In the study, soil and vegetation index (SVI), a spectral index aiming to distinguish between urban and non-urban land cover, was applied [16]. Dadras et al. (2015), using aerial photographs and satellite images from 5 different periods between 1956 and 2012, investigated the expansion of the boundaries of the city of Bandar Abbas. In the study, in which they determined the extent of expansion towards 32 different geographical directions, Landsat satellite images were used, and the classification method was applied to the images [17]. Singh et al. (2017) investigated the negative impact of urbanization and its effect on the increasing trends of temperature and degradation of urban ecology by using the Landsat optic and thermal images of Lucknow city, India [18]. Bala et al. (2021) analyzed LST variation with land cover changes in Varanasi city of India from 1989 to 2018 using Landsat satellite images. They applied the random forest classification algorithm for the classification of optical images [19]. Wang et al.

574

N. Yagmur et al.

(2021) investigated the spatial distribution and influencing factors on LST in twelve cities of Chine between 2000 and 2017 using temperature and spectral vegetation, built-up, and water indices [20]. In their study, Lei et al. (2018) have aimed to extract all possible information from SAR-InSAR data of Hong Kong city and assess to which extent this information can be exploited for change detection purposes. High-resolution optical images were used for comparison and comprehension. PSI is used to monitor subsidence and uplift phenomena in several cities around the world [21]. Aslan et al. (2018) investigated the spatial extent and rate of ground deformation in the Istanbul megacity by combining InSAR datasets obtained from multiple satellites (291 images in total) and analyzing permanent scatterers. They created maps of the average velocity of ground surface displacement in selected regions [22]. In a study conducted in Italy, the aim was to exemplify a typical subsidence issue related to external loads (urbanization) and a local geological and geotechnical setting (i.e., a subsoil composed of alluvial and coastal compressible deposits). The study evaluates the quantitative contribution of InSAR over a broader perspective of land-use planning and risk management [23]. In their study, Delgado Blasco et al. (2019) conducted urban deformation analysis of the Rome metropolitan area using Sentinel-1 SAR data and the PSI method [24].

3 Materials and Methods 3.1

Study Area

Cities with more than 10 million are called megacities, and Istanbul is one of the 33 megacities in the world with a population of 15.5 million, according to 2020 population data. Its annual growth rate is about twice the general average of Turkey due to internal migration, making it one of the fastest-growing cities in Europe [25]. Due to migration from other cities, urbanization has rapidly increased from the past to the present. According to CORINE land cover inventory, artificial surfaces in Istanbul have increased approximately 72% between 1990 and 2018 [26]. Especially study site (Fig. 1), located in Buyukcekmece district, is under the dense urbanization process. Buyukcekmece is one of the 39 districts in Istanbul and is a developing district in terms of urbanization and population. In 2020, Buyukcekmece had a 12.7% population growth [27]. Rapid growth leads to rapid urbanization. According to the temperature parameter provided by the Turkish State Meteorological Service, in the last five years temperature of this district has risen by approximately 2 ºC, which is shown in Fig. 2.

Assessment of Rapid Urbanization Effects …

575

Fig. 1 Study area on Google Earth satellite image

Fig. 2 Monthly Average Temperature of Buyukcekmece Meteorological Station between 2016 and 2020

3.2

Materials

Landsat Level 1 reflectance images which are radiometrically and geometrically corrected, were used to show the urban extension of the study site. Temporal analysis was made with five years period using Landsat 5 TM (Thematic Mapper) and Landsat 8 OLI (Operational Land Imager) satellite images. Optical satellite images are affected by the weather condition, mainly cloudy weather. That is why satellite images were selected in July, considering cloud coverage of scenes except

576

N. Yagmur et al.

Table 1 Characteristics of the optical satellite images Resolution/ Satellite image

Landsat 5 TM

Landsat 8 OLI

Spectral res. (µm)

Blue: 0.45–0.52 Green: 0.52–0.60 Red: 0.63–0.69 Near Infrared: 0.76–0.90 Shortwave Infrared1: 1.55–1.75 Shortwave Infrared2: 2.8–2.35 Thermal: 10.40–12.50

Spatial res.

Blue, Green, Red, Near Infrared, Shortwave Infrared1-2: 30 m Thermal: 120* (30) m

Radiometric res. Temporal res. (day) Date

8 bits

Coastal aerosol: 0.43–0.45 Blue: 0.45–0.51 Green: 0.53–0.59 Red: 0.64–0.67 Near Infrared: 0.85–0.88 Shortwave Infrared1: 1.57–1.65 Shortwave Infrared2: 2.11–2.29 Panchromatic: 0.50–0.68 Cirrus: 1.36–1.38 Thermal1: 10.60–11.19 Thermal2: 11.50–12.51 Panchromatic: 15 m Costal, Blue, Green, Red, Near Infrared, Shortwave Infrared1-2: 30 m Thermal1-2: 100 m 12 bits

16

16

July 1, 1985 July 10, 2000 July 31, 1990 July 24, 2005 July 13, 1995 May 3, 2010

July 20, 2015 July 1, 2020

2010. Because of cloudy weather, a satellite image in May was used for 2010. The properties of the Landsat satellite images used are given in Table 1. The optical satellite images were analyzed using Google Earth Engine cloud-based platform, a free platform that enables geospatial analysis using satellite images [28]. In addition, Sentinel-1 satellite images were used for the time series deformation analysis of the urban area. 145 ascending SLC scenes were processed with only VV polarization. Temporal resolution of Sentinel-1 is 6 days in Europe. Properties of Sentinel-1 are given in Table 2.

Table 2 Properties of SAR images Satellite

Polarization

Orbit direction

Time period

Number of images

Temporal res.

Sentinel-1 C band

VV

Ascending

November 05, 2018 March 24, 2021

145

6 days

Assessment of Rapid Urbanization Effects …

577

Fig. 3 Flowchart of the Land Surface Temperature (LST) algorithm

3.3

Method

Normalized Difference Vegetation Index (NDVI). NDVI, which is used to detect vegetation cover area on the Earth’s surface, is calculated using the near-infrared (NIR) and red bands of the optical satellite images (Eq. 1). NDVI values vary between −1 to +1. Poor vegetation areas like built-up and barren land values are around zero, and healthy vegetation values are close to +1. Water bodies have negative values around −1. NDVI ¼ ðNIR  RedÞ=ðNIR þ RedÞ

ð1Þ

Land Surface Temperature (LST). LST extraction with satellite images is possible using many algorithms [29]. Some of these used algorithms are SplitWindow Algorithm (SWA), Single Channel Algorithm (SCA), Mono-Window Algorithm (MWA), and Radiative Transfer Equation (RTE) [30]. In this study, MWA was used to generate LST maps [31, 32]. The process flow of the methodology is given in Fig. 3. For this algorithm, firstly, the thermal band formulated from the USGS official site is entered. Then Digital numbers (DN) refer to reflectance values are converted to radiance values. Spectral radiation is converted to brightness temperature (BT) using thermal constants provided in the metadata file of the thermal band data [33]. The Kelvin conversion is performed to get the results in Celsius (about −273.15).

Fig. 4 Flowchart of the PSI method using Sentinel-1

578

N. Yagmur et al.

The calculated NDVI was used both in the step of calculating the emission in the LST and in the step of comparing the existing vegetation areas between the LST [34]. The land surface emission (e) needs to be determined to produce the LST map,. Surface emissivity is defined as the relative ability of the surface of a material to emit energy by radiation. The emissivity is also expressed as the ratio of the energy emitted by a given material to the energy emitted by a black body at the same temperature. In short, it is the efficiency of transmitting thermal energy from the surface to the atmosphere [35]. Equation 2 and Eq. 3 are applied to obtain the land surface temperature map [33]. In the equations, r is Boltzmann constant (1.38  10 −23 J/K), ℎ is Planck’s constant (6.626  10−34 Js), and c is the velocity of light (2.998  108 m/s). LST ¼ BT = f1 þ ½ðkBT=qÞ  lnðeÞg

ð2Þ

q ¼ h  ðc=rÞ ¼ 1:438  102 m K

ð3Þ

Persistent Scatter Interferometry (PSI). PSI method was developed twenty years ago for time series deformation analysis of persistent scatters (PS) [36, 37]. This method uses a stack of co-registered SAR images and finds PS points using amplitude and phase variations from the images. Time-series deformation analysis is based on the phase information of the PS points. They are stable artificial objects like buildings and monuments or natural objects and could not be found in rural areas [12]. The steps of the PSI time series InSAR method are given in Fig. 4. SAR satellite images are split according to the study area before interferometric analysis. Orbital correction is applied using orbit files, and slave images are co-registered to the selected master image. After the interferograms are created for each pair, the topographic phase is removed from the interferograms. According to the calculated phase noise estimation value for each candidate pixel, PS pixels are selected from amplitude dispersion [37]. Amplitude dispersion is taken as between 0.4–0.42, which are reasonable values. After the initial selection step, noisy PS pixels are discarded. The wrapped phase is corrected for spatially uncorrelated look angle error and is unwrapped. Atmospheric filtering is applied using the Delaunay triangulator after look angle error estimation. Obtained deformation results are exported to geographical information systems for the overall assessment of rapid urbanization.

4 Results and Discussion 4.1

Detection of Urban Area Extension

To show changes in the vegetation coverage and extension of urbanization, NDVI was calculated using Landsat 5 TM and Landsat 8 OLI satellite images between

Assessment of Rapid Urbanization Effects …

579

Fig. 5 NDVI density map of study site between 1985–2020 with 5 years period

1985 and 2020 with five years period. With this purpose, NDVI density was determined in five categories as water body (NDVI < 0), built-up (0 < NDVI < 0.15), low vegetation (0.15 < NDVI < 0.4), medium vegetation (0.4 < NDVI < 0.6), and high vegetation (0.6 < NDVI). Produced NDVI density maps for the aforementioned images are shown in Fig. 5. It is clearly observed that the built-up area has covered a large part of the study site. The areal extension of five NDVI density classes was calculated as a time series analysis in Fig. 6. The built-up area has extended from 57.1 ha to 781.4 ha in the study area. The right part of the urbanization was clearly detected with NDVI

Fig. 6 The areal distribution of NDVI density values between 1985–2020

580

N. Yagmur et al.

results; however, the detection of the left part has not been successful because of having green spaces between buildings and the medium spatial resolution of the Landsat (30 m). Therefore, the area of the built-up class is more than detected 781.4 ha. Whereas the low and medium vegetation area has decreased, high vegetation has increased in this time interval. In 2010, the change in high vegetation class was due to the use of satellite imagery acquired in May having dense vegetation. Additionally, the coastline of the water body has changed, and the surface area of the water has increased from 94.07 ha to 154.8 ha.

4.2

Land Surface Temperature (LST)

LST was applied to selected images, and results are given in Fig. 7. It is clearly observed that the temperature on the urbanization site increased to 45 ºC. The average temperature of 1985 is 26.1 ºC and 2020 31.5 ºC. Average temperature has increased 5.4 ºC because of converting natural LULC to artificial surfaces. In the study area, there are two types of regions that can be divided as having green spaces or not. These regions are shown with red circles in Fig. 8a for 2020. Urban expansion in Region 2 can be clearly extracted with the NDVI map (Fig. 8b), and LST has reached 45 ºC (Fig. 8c). Unlike Region 2, Region 1 entered the mostly medium vegetation class in the NDVI map (Fig. 8b). The reason for that is to cover empty spaces with green spaces. Additionally, because of the spatial

Fig. 7 The Distribution of LST between 1985–2020

Assessment of Rapid Urbanization Effects …

581

Fig. 8 For 2020, study region’s; a Google Earth image, b NDVI classified map, c LST map

resolution (30 m) of Landsat satellite images, these sites were detected as medium vegetation class. Therefore, because of the above-stated reason, LST values of Region 1 have not increased as much as Region 2. This situation also reveals the importance of landscaping.

4.3

Time Series Deformation Analysis

A deformation velocity map of the study site was obtained using the PSI multi-temporal InSAR method. The SLC image acquired on November 30, 2019, was chosen as the master image, and the other 144 images acquired between November 2018 and March 2021 were selected as the slave images. The produced velocity map is shown in Fig. 9a. While blue points show uplift area, red points show the subsidence in line of sight (LOS) direction. A close view of the recent urbanization site is given in Fig. 9b. The time series of PS points showing uplift and subsidence movement in the white circle is given in Fig. 9c with the same color in the velocity map. The red points in Fig. 9b show that the subsidence of the buildings and construction sites in the middle of buildings can be one of the main reasons for the subsidence. Another reason is the ground stability of the region. Istanbul Metropolitan Municipality (IMM) carried out a landslide awareness project and prepared reports for all districts in Istanbul. In the created report for the Buyukcekmece district, a landslide activity status map was generated considering geological conditions, soil properties, and risk factors [38]. According to generated landslide activity status map, and study area covers passive landslide region which has the potential for a landslide. Additionally, approximately 30% of the landslides

582

N. Yagmur et al.

Fig. 9 a PSI deformation analysis results over the region, b a close view of the new urbanization site, c time series of the red and blue points in the white circle

were activated by human impact events. These effects are mainly caused by filling the crown of the landslides, constructing high-rise buildings on the crown of the landslides, and digging the heel of the landslide [39]. Bayik et al. (2021) also evaluated Buyukcekmece – Esenyurt region with the PSI method using ALOS-2 and Sentinel-1 SAR images between 2015–2020 and related with GNSS measurements and landslide inventory map to explain movements characteristics [40]. The region showed as a white circle in Fig. 9b was also evaluated in their study, and subsidence has been detected up to 10 mm in the last 5 years. In this study, however, PS points in the same region were detected as green and yellow color points, which are more stable than the others. It means that as long as the ground settlements are not exposed to other external factors and are not built in areas that will trigger landslides, they become stable over time.

5 Conclusion Human activities like urbanization can disrupt the Earth’s natural balance if it is not controlled. In the study, the urbanization effects were evaluated in two aspects. The first aspect is the urbanization effects on the Earth’s surfaces and natural balance;

Assessment of Rapid Urbanization Effects …

583

the second is the effects of the ground movements and construction sites on structural stability in the rapidly urbanized region. Effects on the Earth’s surfaces were determined with NDVI indices to determine LULC change and LST to explain urbanization’s urban heat island effect. Both of the methods showed successful results. LULC has significantly changed, and LST has increased 5.4 ºC in the study site. Furthermore, according to results, it is obtained that landscaping converting empty spaces with green spaces decreases the effects of the urban heat islands. In the second aspect, it is observed that construction sites can cause the surrounding buildings to collapse. Additionally, urbanization in the risky sites, sensitive to landslides, comes up against ground movement and deformation. Multi-temporal InSAR is a powerful tool to present the actual condition of the building, and the PSI method showed the deformations caused by construction sites and ground stability. Therefore, it is important to monitor risky sites, and this will prevent disasters that may occur. For more detailed analysis, high-resolution optical and SAR satellite images can be used; but these free satellite images are cost-effective and more useful for a continuous monitoring system.

References 1. N.B. Grimm et al., Global change and the ecology of cities. Science 319(5864), 756–760 (2008) 2. H. Luo, J. Wu, Effects of urban growth on the land surface temperature: a case study in Taiyuan, China. Environ. Dev. Sustain. 23, 10787–10813 (2021) 3. L.W. Lai, W.L. Cheng, Air quality influenced by urban heat island coupled with synoptic weather patterns. Sci. Total Environ. 407(8), 2724–2733 (2009) 4. R. Yao et al., Long-term trends of surface and canopy layer urban heat island intensity in 272 cities in the mainland of China. Sci. Total Environ. 772, 145607 (2021) 5. B. Halder, J. Bandyopadhyay, P. Banik, Monitoring the effect of urban development on urban heat island based on remote sensing and geo-spatial approach in Kolkata and adjacent areas, India. Sustain. Cities Society 74, 103186 (2021) 6. R. Amiri, Q. Weng, A. Alimohammadi, S.K. Alavipanah, Spatial–temporal dynamics of land surface temperature in relation to fractional vegetation cover and land use/cover in the Tabriz urban area, Iran. Remote Sens. Environ. 113(12), 2606–2617 (2009) 7. S. Guo, et al., Deformation velocity monitoring in kunming city using ascending and descending sentinel-1A data with SBAS-InSAR technique, in IGARSS 2020–2020 IEEE International Geoscience and Remote Sensing Symposium (IEEE 2020), pp. 1993–1996 8. G.X. Liu, X.L. Ding, Z.L. Li, Z.W. Li, Y.Q. Chen, S.B. Yu, Pre-and co-seismic ground deformations of the 1999 Chi-Chi, Taiwan earthquake, measured with SAR interferometry. Comput. Geosci. 30(4), 333–343 (2004) 9. N. Yagmur, E. Erten, N. Musaoglu, How to start gentrification process using interferometric stack of SENTINEL-1. ISPRS-Int. Arch. Photogram. Remote Sens. Spat. Inf. Sci. 43, 183– 188 (2021) 10. S. Stramondo, F. Bozzano, F. Marra, U. Wegmuller, F.R. Cinti, M. Moro, M. Saroli, Subsidence induced by urbanisation in the city of Rome detected by advanced InSAR technique and geotechnical investigations. Remote Sens. Environ. 112(6), 3160–3172 (2008)

584

N. Yagmur et al.

11. L. Solari, A. Ciampalini, F. Raspini, S. Bianchini, S. Moretti, PSInSAR analysis in the Pisa urban area (Italy): a case study of subsidence related to stratigraphical factors and urbanization. Remote Sens. 8(2), 120 (2016) 12. G. Chen, Y. Zhang, R. Zeng, Z. Yang, X. Chen, F. Zhao, X. Meng, Detection of land subsidence associated with land creation and rapid urbanization in the chinese loess plateau using time series insar: a case study of Lanzhou new district. Remote Sens. 10(2), 270 (2018) 13. S. Jiao, J. Yu, Y. Wang, L. Zhu, Q. Zhou, Estimating the impact of urban expansion on land subsidence using time series of dmsp night-time light satellite imagery. Int. Arch. Photogram. Remote Sens. Spatial Inf. Sci. 42(3), 691–698 (2018) 14. B. Bhatta, S. Saraswati, D. Bandyopadhyay, Quantifying the degree-of-freedom, degree-of-sprawl, and degree-of-goodness of urban growth from remote sensing data. Appl. Geogr. 30(1), 96–111 (2010) 15. Y. Xiong, S. Huang, F. Chen, H. Ye, C. Wang, C. Zhu, The impacts of rapid urbanization on the thermal environment: a remote sensing study of Guangzhou, South China. Remote Sens. 4 (7), 2033–2056 (2012) 16. P. Villa, Mapping urban growth using soil and vegetation index and landsat data: the Milan (Italy) city area case study. Landsc. Urban Plan. 107(3), 245–254 (2012) 17. M. Dadras, H.Z. Shafri, N. Ahmad, B. Pradhan, S. Safarpour, Spatio-temporal analysis of urban growth from remote sensing data in Bandar Abbas city, Iran. Egypt. J. Remote Sens. Space Sci. 18(1), 35–52 (2015) 18. P. Singh, N. Kikon, P. Verma, Impact of land use change and urbanization on urban heat island in Lucknow city, Central India. A remote sensing based estimate. Sustain. Cities Society 32, 100–114 (2017) 19. R. Bala, R. Prasad, V.P. Yadav, Quantification of urban heat intensity with land use/land cover changes using Landsat satellite data over urban landscapes. Theoret. Appl. Climatol. 145, 1–12 (2021) 20. Y. Wang et al., Spatial distribution and influencing factors on urban land surface temperature of twelve megacities in China from 2000 to 2017. Ecol. Indicators 125, 107533 (2021) 21. L. Lei, D. Perissin, Y. Qin, Change detection with spaceborne InSAR technique in Hong Kong, in 2013 IEEE International Geoscience and Remote Sensing Symposium-IGARSS (IEEE 2013), pp. 338–341 22. G. Aslan, Z. Cakır, S. Ergintav, C. Lasserre, F. Renard, Analysis of secular ground motions in Istanbul from a long-term InSAR time-series (1992–2017). Remote Sens. 10(3), 408 (2018) 23. F. Bozzano, C. Esposito, P. Mazzanti, M. Patti, S. Scancella, Imaging multi-age construction settlement behaviour by advanced SAR interferometry. Remote Sens. 10(7), 1137 (2018) 24. J.M. Delgado Blasco, M. Foumelis, C. Stewart, A. Hooper, Measuring urban subsidence in the Rome metropolitan area (Italy) with Sentinel-1 SNAP-StaMPS persistent scatterer interferometry. Remote Sens. 11(2), 129 (2019) 25. UN:https://www.un.org/en/events/citiesday/assets/pdf/the_worlds_cities_in_2018_data_ booklet.pdf. (2018). 26. CORINE (2020). https://land.copernicus.eu/ 27. TUIK (2020). https://www.tuik.gov.tr/ 28. N. Gorelick, M. Hancher, M. Dixon, S. Ilyushchenko, D. Thau, R. Moore, Google Earth Engine: planetary-scale geospatial analysis for everyone. Remote Sens. Environ. 202, 18–27 (2017) 29. B. Yamak, Z. Yagcı, B.B. Bilgilioglu, R. Comert, Investigation of the effect of urbanization on land surface temperature example of Bursa. Int. J. Eng. Geosci. 6(1), 1–8 (2021) 30. M. IsayaNdossi, U. Avdan, Application of open source coding technologies in the production of land surface temperature (LST) maps from Landsat: aPyQGIS plugin. Remote Sens. 8(5), 413 (2016) 31. Z. Qin, A. Karnieli, P. Berliner, A mono-window algorithm for retrieving land surface temperature from Landsat TM data and its application to the Israel-Egypt border region. Int. J. Remote Sens. 22(18), 3719–3746 (2001)

Assessment of Rapid Urbanization Effects …

585

32. F. Wang, Z. Qin, C. Song, L. Tu, A. Karnieli, S. Zhao, An improved mono-window algorithm for land surface temperature retrieval from Landsat 8 thermal infrared sensor data. Remote Sens. 7(4), 4268–4289 (2015) 33. S.K. Akher, S. Chattopadhyay, Impact of urbanization on land surface temperature-a case study of Kolkata New Town. Int. J. Eng. Sci. 6(1), 71–81 (2017) 34. U. Avdan, G. Jovanovska, Algorithm for automated mapping of land surface temperature using LANDSAT 8 satellite data. J. Sensors 2016, 1–8 (2016). https://doi.org/10.1155/2016/ 1480307 35. J.C. Jimenez-Munoz, J.A. Sobrino, A. Gillespie, D. Sabol, W.T. Gustafson, Improved land surface emissivities over agricultural areas using ASTER NDVI. Remote Sens. Environ. 103 (4), 474–487 (2006) 36. A. Ferretti, C. Prati, F. Rocca, Nonlinear subsidence rate estimation using permanent scatterers in differential SAR interferometry. IEEE Trans. Geosci. Remote Sens. 38(5), 2202– 2212 (2000) 37. A. Ferretti, C. Prati, F. Rocca, Permanent scatterers in SAR interferometry. IEEE Trans. Geosci. Remote Sens. 39(1), 8–20 (2001) 38. IMM (Istanbul Metropolitan Municipality): Landslide Awareness Booklet of Beylikduzu-Buyukcekmece-Esenyurt Districts (2020). (in turkish) 39. S. Dalgic, M. Turgut, I. Kusku, C. Coskun, T. Cosgun, Impact of soil and rock conditions on building foundations on the european side of Istanbul. Uygulamalı Yerbilimleri Dergisi 8(2), 47–70 (2009). ((in turkish)) 40. C. Bayik, S. Abdikan, A. Ozdemir, M. Arıkan, F.B. Sanli, U. Dogan, Investigation of the landslides in Beylikdüzü-Esenyurt districts of Istanbul from InSAR and GNSS observations. Nat. Haz. 109, 1–20 (2021)

Indexing Approach for the Evaluation of Heavy Metals in Drinking Water Produced by a Moroccan Water Treatment Plant Abderrahman Achhar, Mohamed Najy, Driss Belghyti, and Almehdi Alibrahimi Abstract In the present work, the assessment of drinking water quality was carried out through a monitoring of heavy metals in the treated and consumed waters in the city of Fes (Morocco). Monthly sampling was conducted for a period of 24 months between January 2016 and December 2019. Nine parameters were evaluated: pH, T (°C), Turbidity (NTU), Al, Fe, Cu, Mn, Al2(SO4)3 and CaO. Indexing approaches have been applied by calculating the Heavy Metal Pollution Index (HPI) and Metal Index (MI) for the assessment of influence of heavy metals on the overall quality of water. The obtained results for heavy metals are in good agreement with World Health Organization (WHO) standards. Though the aluminum concentration remains in the limits set by WHO, yet it shows a major contribution in the indices. This has been verified by the statistical analysis which demonstrates fair correlations between aluminum, HPI (r = 0.9) and MI (r = 0.77). Aluminum showed the important influence of seasonal change in the year as well as the doses of reagents injected during the treatment process on the concentration of aluminum is detailed. Keywords Heavy metals


HPI
Morocco
MI
Water treatment

1 Introduction Monitoring and study of heavy metals in drinking water present an immense importance for the protection of public health [1–4]. According to the United Nations [5], 3.5 million people die each year due to poor water supply conditions. Indeed, Africa is the driest continent after Australia. In an African country like Morocco where water is increasingly scarce, monitoring and control of pollution A. Achhar (&)
A. Alibrahimi Laboratory Renewable Energies and the Environment, Faculty of Sciences, University Ibn Tofail, 14 000, P.O. Box: 133, Kenitra, Morocco M. Najy
D. Belghyti Laboratory of Agro-Physiology, Biotechnology, Environment and Quality, Water Team, Wastewater, Health, Faculty of Science, Ibn Tofail University, Kenitra, Morocco © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_47

587

588

A. Achhar et al.

remains a major challenge for the protection of public health. About 2150 morocco which 1700 children under 5 years die each year due to the consumption of the water with poor quality. Contamination of drinking water by heavy metals has always generated a grave consequences on the health of populations and their socio-economic life [8]. The metal trace elements play an essential role for human health; Copper, Iron and Zinc are necessary for life when they do not exceed the limits; On the other hand, the Lead and the Mercury are very dangerous for the living beings. In addition to their natural origins in water, heavy metals come mainly from anthropogenic activities such as, industrial, agricultural and domestic discharges. Different methods like adsorption, chemical precipitation, physical separation, ion exchange, membrane filtration, distillation and hybrid methods are applied for the removal of heavy metals. The efficiency of the methods adopted during treatment plays a predominant role in reducing the level of these elements in order to achieve tolerable values in agreement with national/international standards, while also avoiding the production of corrosive water in order to eliminate the risk of corrosion of pipes which represent a potential source of heavy metals in drinking water.

2 Materials and Methods Heavy Metal Pollution Index (HPI): The HPI is a metal evaluation method and an effective tool which makes it possible to demonstrate the influence of heavy metals on the overall water quality. This technique is based on the evaluation of the unit weight assigned to each selected parameter. The HPI is usually calculated according to the equation proposed by [16]. We considered the limit values set by [1] for calculating HPI. HPI ¼

Pn ðw þ Qi Þ i¼1 Pn i i¼D wi

ð1Þ

Qi: Sub index of the ith parameter; Wi: The unit weight of the ith parameter and n is the number of parameters considered. The sub index Qi is given by the following formula: Qi ¼

Xn Mi  R  100 i¼1 Si  R

ð2Þ

Mi: Monitored value of heavy metal of the ith parameter; Si: Standerd value of ith parameter; Ii : Ideal value of the ith parameter. The unit Weight (Wi) is obtained by Eq. (3):

Indexing Approach for the Evaluation of Heavy Metals….

589

1 Si

wi ¼

ð3Þ

The critical value of the metal pollution index is 100.metal Index (MI): In order to calculate this index, the model proposed by [20] is used. The MI is an assessment method that provides an overview of the overall water quality based on the additive effect of heavy metals. It is given by the following equation MI ¼

Xn i¼1

ð

Ci Þ ðMACÞi

ð4Þ

Ci: is the concentration of each metal; (MAC)i: is Maximum Allowable Concentration.

3 Results and Discussion The results obtained on the physical parameters are presented in Table 1. These results show that the treated water is slightly basic with annual average pH values for 2016 and 2019 respectively 7.79 ± 0.177 and 7.83 ± 0.10. Average water temperatures range between 24.90 ± 2.85 and 25.68 ± 2.48 °C. Turbidity of the water vary within the ranges 0.30–1.25 and 0.25–0.59 NTU. During the period of this study, the physical parameters demonstrated the results in accordance with the standards set by the [1]. These results are in good agreement with those obtained by [30]. The overall assessment of heavy metals is presented in Table 1. The lowest concentrations were recorded for manganese (Mn) with mean values of 0.006 ± Table 1 Descriptive statistics of the physical parameters and metal concentrations Parameter

Unit

Si

2017

2019

Min

Average

Max

SD

Min

Average

Max

SD

pH

–

6– 8.5

7.54

7.79

8.12

0.177

7.69

7.83

7.99

0.1

Temperature

°C

25

20.57

25.68

29

2.482

20.7

24.9

28.32

2.85

Turbidity

NTU

5

0.3

0.68

1.25

0.354

0.25

0.44

0.59

0.09

Al

mg
L−1

0,2

0.04

0.06

0.08

0.014

0.055

0.07

0.1

0.02

Fe

mg
L−1

0.3

0

0.023

0.056

0.016

0

0.017

0.04

0.01

Mn

mg
L−1

0,4

0.0008

0.006

0.014

0.004

0

0.002

0.006

0.002

Cu

mg
L−1

2

0

0.028

0.06

0.017

0

0.021

0.05

0.01

590

A. Achhar et al.

0.004 and 0.002 ± 0.002 mg
L-1. These values correspond to the WHO standard (0.4 mg
L-1). Mean iron (Fe) concentrations are lower than the WHO standard (0.3 mg
L-1), oscillating between 0.023 ± 0.016 and 0.017 ± 0.01 mg
L-1. The copper (Cu) level very low compared to the WHO standard (1 mg
L-1) with mean values of 0.028 ± 0.017 and 0.021 ± 0.01 mg
L-1. The obtained results for manganese, iron and copper are generally low compared to those obtained by, for the assessment of heavy metals in raw water before treatments. Coagulation, flocculation and sedimentation allows the elimination of heavy metals. These results show the efficiency of the processes applied in treatment plant to reduce the level of metallic trace elements which, according to, are highly correlated with the suspended matter. Aluminum (Al) is the most predominant element compared to other heavy metals, with concentrations ranging from 0.06 ± 0.014 to 0.07 ± 0.02 mg
L-1, with maximum values of 0.08 mg
L-1 in 2016 and 0.10 mg
L-1 in 2019. The contents of aluminum are below the standard set by the WHO (0.2 mg
L-1). Aluminum sulphate (Al2(SO4)3) added during coagulation represents the main source of increase in the level of aluminum in drinking water [34, 35]. The data obtained for the heavy metals (Mn, Fe, Cu and Al) allowed the estimation of the Heavy Metal Pollution Index (HPI) based on Eq. (1), proposed by [16], and the standards of [1]. Annual changes in HPI are presented in Table 2. Indeed, the obtained values are below the critical threshold ( Fe > Cu > Mn. The evaluation of the influence of heavy metals on the overall quality of water and their additive effect were carried out using an indexing approach. HPI and MI indices remain below their critical thresholds. Aluminum demonstrates an important contribution in the indices. Aluminum concentration has been found to strongly relate the seasonal changes and is also influenced by the doses of chemical reagents injected during

594

A. Achhar et al.

treatment. In the prospects, it would be particularly interesting to study the behavior of aluminum during the transport of water in the distribution network, with possibilities of decreasing the concentration due to precipitation of Al(OH)3 in the pipes and reservoir.

References 1. World Health Organization: Guidelines for drinking-water quality, 4th edition, Geneva (2011) 2. M. Soylak, F.A. Aydin, S. Saracoglu, L. Elci, M. Dogan, Chemical analysis of drinking water samples from Yozgat Turkey. Polish J. Environ. Stud. 11(2), 151–156 (2002) 3. M.R. Abraham, T.B. Susan, Water contamination with heavy metals and trace elements from Kilembe copper mine and tailing sites in Western Uganda; implications for domestic water quality. Chemosphere 169, 281–287 (2017) 4. V. Sheykhi, F. Moore, Geochemical characterization of Kor River water quality Fars Province Southwest Iran. Water Qual. Exposure Health 4(1), 25–38 (2016) 5. United Nations: International Decade for Action Water for life, 2005–2016. http://www.un. org/waterforlifedecade/africa.shtml 6. M.H. Ibraheem, K. Izawa, On the morphology of Lamproglena monodi Capart, a parasitic copepod on the gills of Tilapia in Egypt. J. Zool. Middle East 21(1), 103–108 (2000) 7. D. Belghyti, M. Mouhssin, N. Mokhtar, et al., Systèmatique de deux nouveaux copépodes parasites de la sardine (Sardina pilchardusWalbaum 1792) de l’atlantique marocain (Kénitra-Mehdia). Actes Inst. Agron. Vet. (Maroc) 10. 17(3), 173–180 (1997) 8. A. Loukili, L. Mrabet, D. Belghyti, La diversité des ectoparasites de l’anguille européenne dans deux hydrosystèmes différents. Afrique Sci. 12(1), 183–190 (2016) 9. M. Boungou, Y. Sinaré, K. Mano et al., Parasitic copepods (arthropoda, crustacea, copepoda) from fishes in Burkina Faso, Africa. Int. J. Fish. Aquatic Sci. 2(3), 58–64 (2013) 10. I. Paperna, Parasites, infections et maladies du poisson en Afrique. CPCA, Doc. Tech. (7), 202 (1982) 11. R. Abdel-gaber, N. El Deeb, S. Maher et al., Diversity and host distribution of the external gill parasite Lamproglena monodi (Copepoda: Lernaeidae) among Tilapia species in Egypt: light and scanning electron microscopic studies. Egypt. J. Exp. Biol. (Zool.) 13(1), 23–30 (2017) 12. R.K. Azevedo, V.D. Abdallah, R.J. Silva et al., Expanded description of Lamproglena monodi (Copepoda: Lernaeidae), parasitizing native and introduced fishes in Brazil. Rev. Bras. Parasit. Vet. 21(3), 263–269 (2012) 13. J.P. Lazo, D.A. Davis, C.R. Arnold, The effects of dietary protein level on growth, feed efficiency and survival of juvenile Florida pompano (Trachinotus carolinus). Aquaculture 169(3– 4), 225–232 (1998) 14. M.J. Heilman, R.E. Spieler, The daily feeding rhythm to demand feeders and the effects of timed meal-fe eding on the growth of juvenile Florida pompano Trachinotus carolinus. Aquaculture 180, 53–64 (1999) 15. T. Pero, G.C. Nik_a, K. Valter, S. Bo_ko, G. Branko, Preliminary information on feeding and growth of pompano, Trachi-notus ovatus (Linnaeus 1758) (Pisces; Carangidae) in captivity. Aquaculture Int. 12, 387–393 (2004) 16. D. Belghyti, O. Berrada-Rkhami, V. Boy, P. Aguesse, C. Gabrion, Population biology of two helminth parasits of flatfishes from the atlantic coast of morocco. J. Fish Biol. 44, 1005–1021 (1994) 17. M.H. Barahona-Fernandez, L’élevage inten-sif des larves et des juvéniles du Bar Dicentrachus labrax, données biologiques, zoologiques et pathologiques. Thèse Dr. Etat Univ. AixMarseille (1978), p. 280 18. R.E.J. Maloine, Pathologie, p. 303. du poisson 1 vol. (1979)

Indexing Approach for the Evaluation of Heavy Metals….

595

19. L. Euzet, M.H. Ktari, Pyragraphorus Hol-lisae Sp. Nov (Monogenea) parasite de Lichia Glauca, en Méditerranée (L., 1758) 20. C.R. Kennedy, The Regulation of Fish Para-site Populations. (Acad. Press, San Francisco, New York, London, 1977), p. 253 21. P. Silan, Biologie comparée des popula-tions de Diplectanum aequans et D. laubieri Monogènes branchiaux de Dientrachus labrax. Thèse de 3ème cycle. U.S.T.L. Montpel- lier (1984), p. 275 22. J. Frnandez-Bargiela, Los parasitos de la Liza Mugil cephalus L. en Chile: sistema-tica y aspetos poblacionales (Mugilidae). Gayana, Zool 51(1–4), 3–58 (1987) 23. D. Noisy, C. Maillard, Microhabitat bran-chial préferentiel de Microcotyle chryso-phri Van Beneden et Hesse, 1863 (Mono-genea, Microcotylidae) parasite de la Daurade (Spa- rus aurata L, 1758). Annls. Parasit. Hum. Comp 55(1), 33–40 (1980) 24. K. Buchmann, Spatial distribution of Pseudo-dactylogyrus anguillae and P. bini on the gills of the European eel, Anguilla anguilla. J. Fish. Biol. 32, 801–802 (1988) 25. T. Morato, P. Afonso, P. Lourinho, J.P. Barreiros, R.S. Santos, R.D.M. Nash, Length- weight relationships for 21 coastal fish species of the Azores, north-eastern Atlantic. Fish. Res. (Amsterdam) [Fish. Res.] 50, 297–302 (2001) 26. A.J. Edwards, C.W. Glass, The fishes of Saint Helena Island, South Atlantic Ocean The pelagic fishes. J. Nat. Hist 21, 1367–1394 (1987) 27. M. Koskivaara, Environnemental factors affecting monogeneans parasitic on fresh-water fishes. Parasitol. Today 8(10), 339–342 (1992) 28. D. Belghyti, Les interactions hôte parasite et leurs expressions sur la biologie des pois- sons plats de la côte atlantique de Casablanca Mehdia. Thèse Doc. d’Etat, Fac. Sc. Kéni- tra (1996), p. 342 29. B.E. Bychowsky, Monogenetic Tremato-des, their classification and phylogeny. Academy of Sciences, USSR, ABHS. Agence du bassin hydraulique de Sebou Fes; Rapport (Presentation des bassins hydrauliques du Maroc (2007), p. 53 30. A. Alemad, M. Nagi, A. Ibeda, R. Naser, Y. Alwathaf, O. Elrhaouat, K. Elkharrim, D. Babaqi Aet Belghyti, The impact of Sana’a solid waste on the quality of groundwater in Yemen WIT Trans. Ecol. Environ. (2014). https://doi.org/10.2495/WS130151.Vol178:171-183 31. H. Alvares-Vazquez, B. Jefferson, S.J. Judd, Membrane bioreactors vs. conven- tional biological. Chem. Techn. Biotechn. 79, 1043–1049 (2004) 32. H. Beatriz, P. Rafael, V. Marisol, B. Enrique, M. Jose, F. Luis, Temporal evolu- tion of groundwater composition in an alluvial aquifer (Pisuerga River, Spain) by principal component analysis. Wat. Res. 34(3), 807–816 (2000) 33. O. El Rhaouat, I. El Kherrati, F. El khayyat, H. Chiguer, K. Ezziani, A. Ibeda, M. Fareh, Y. Saidi, K. El Kharrim, D. Belghyti, Physic-chemical evaluation of urban wastewater of the Town of Sidi Kacem. Comput Water Energy Environ. Eng. 3, 30–35 (2014). https://earth. google.com. (2015) 34. A. Imane Kherrati, M. Alemad, H. Sibbari, K. Ettayea, Y. Ezziani, M. Saidi1, S. Benchikh, H. Alzwi, Z. Chiguer, A. Zgourdah, H. Bourass, O. Daifi, K. Elrhouat, K. Elkharrim, D. Belghyti, Health risk of maamora’s groundwater pollution in Morocco. 6, 290–30 (2014) 35. ISO 5667/3: Qualite de l’eau - echantillonnage - guide pour la conservation et la manipulation des echantillons (1994) 36. J.E. Jackson, A User’s Guide to Principal Components (Wiley, New York, 1991) 37. R. Meglener, Examining large databases: a chemo-metric approach using principal component analysis. Mar. Chem. 39, 217–237 (1992) 38. Ministere de l’Environnement du Maroc. Normes marocaines, Bulletin officiel du Maroc, No. 5062 du 30 ramadan 1423, Rabat (2002) 39. Norme Marocain: Qualite des eaux d’alimentation humaine. Eduction et diffusion par le Service de Normalisation Industrielle Marocaine (SNIMA), no. 03.7.001, p. 14 (2011) 40. ONEP: Approche de la typologie des eaux usees urbaines au Maroc. ONEP et GTZ (1998) 41. J. Rodier, L’analyse de l’eau naturelle, eaux residuaires, eau de mer, 8eme edn. (Denod, Paris, 1996), p. 1383

Envirolarm: A Mobile App to Manage Natural Hazards – Scenarios for a Small Island States Rikeelesh Kumar Ramjattun, Mainkah Shicksha Rampersad, and Roopesh Kevin Sungkur

Abstract It is clear that by no means, natural or manmade disasters can be fully prevented. Disaster strikes countries causing tremendous destruction, impacting every aspect of the countries. Mauritius being a small island is not spared of natural hazards which can be in different forms. However, the negative effects can be mitigated and ICT can be used in all the phases of the Disaster Management process. In a current situation characterized by the attempt of the world to combat the invisible enemy which is that of COVID-19, it is obvious that preparedness and risk reduction are key concepts. The same concepts are primordial in disaster risk reduction. This research advocates the use of a mobile App, Envirolarm, to help the citizens of Mauritius to act in circumstances of disasters by providing to them useful and timely information. This research also highlights some scenarios about how ICT can be used to better prepare the Island of Mauritius against natural hazards. The findings show that Mauritius is not currently using technology to its fullest in order to deal with the consequences of various disasters. It can also be said that advancement in ICT in the form of Internet, remote sensing, and satellite communication can help a lot in the management of disasters.










Keywords Disaster management Google API Mobile app Push notifications Marker




1 Introduction During the past years, disasters have become more frequent and destructive towards social constructions and economic development worldwide. According to the World Economic Forum, extreme weather events and natural calamities are probably the two most possible global risks with the highest impact. Out of all the R. K. Ramjattun
M. S. Rampersad
R. K. Sungkur (&) Department of Software and Information Systems, Faculty of Information, Communication and Digital Technologies, University of Mauritius, Reduit, Mauritius e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_48

597

598

R. K. Ramjattun et al.

natural disasters, floods have occurred more frequently from 1995–2019, and droughts affected most people globally. Disasters can be classified as natural and human-made (man-made). Natural disasters are a major adverse event resulting from natural processes of the Earth; examples are floods, hurricanes, tornadoes, volcanic eruptions, earthquakes, tsunamis, and other geologic processes (Dunant et al. 2021). Human-made disasters are the results of technological or human hazards; for example, fires, oil spills and nuclear explosions/nuclear radiation (Tatlow 2016). How countries are prepared for such unexpected situations is important to develop resilient systems. Those systems in place might help reduce the number of casualties associated with disasters in the future. As far as risks are concerned when a disaster occurs, it can only be mitigated by means of some kind of procedure but cannot be controlled. Two cases of risk reduction can be considered; firstly, the risk of being associated before a certain disaster and secondly, post disaster procedures. According to studies done, three main disasters, namely cyclones, floods, and landslides are common in the local context. The early warning of those disasters is sent to the public via radio and social media reaching about 99% of the population (Kasenally 2014; Azmeri and Satria 2021). To minimize casualties and loss of lives, shelters are available for those whose houses are not able to withstand the disaster. The main problem with the shelters is that they tend to get full easily and the excess people coming are not notified. Moreover, if they did not get a message via radio about the shelters, they lose the information. During those harsh times, some shopkeepers have a hard keeping their inventory up to the point that they can satisfy the locality with emergency supplies as the early warning prediction is not within a range for such actions to be taken. From the above analysis we can point out that the precautions are not accessible to the public as provided by other application around the world. Disasters risks being on the rise, the population expects more measures from the authorities to mitigate them. The mitigation plan is usually precautionary measures that are transmitted to the public. The information is transmitted by radio and television mostly in our local context. According to UNDRR Mauritius invests 2% GDP in resilience, and has been a champion of the Sendai Framework for Disaster Risk Reduction. It can also be noted that the past events insist on re-adapting the current disaster risk reduction and management strategies. (UNDRR 2016) The information pertaining to disasters are delivered to the population in real time. This is one of the reasons that make the inhabitants prone to damages that can be mitigated further. Disaster Management is a never-ending process by which governments, civil, law enforcement officers and the society itself plan for and reduces the impact of disasters, react during and immediately after a disaster, and take steps towards recovery (Shidiq et al. 2021). The cycle can be in 4 phases which are shown in Fig. 1 below.

Envirolarm...

599

Fig. 1 Disaster management cycle

2 Literature Review 2.1

Disasters Around the Globe and in Mauritius

The major categories of disasters can be classified according to the table below (Table 1).

Table 1 Categories of disaster

Disaster groups

Disasters

Geophysical

Earthquake Landslides Tsunami Tropical Storms Thunderstorms Flash floods Floods Droughts Viral disease infections Epidemics

Meteorological Hydrological Climatological Biological

From the data of the Emergency Events Database, EM-DAT (2020), it can be deduced that the most frequent natural disaster that affects the world are floods and storms which are present every year from the year 2000–2019. Mauritius being a small island is frequently affected by those two disasters and landslides. If the total number of people affected by floods and storm worldwide is analysed, it can be concluded that the number of persons affected is high.

600

2.2

R. K. Ramjattun et al.

Appraisal of Existing Disaster Risk Reduction Systems

A. Humanitarian Kiosk This app has been designed and created by the United Nation. It is designed to provide a selection of up to the minute humanitarian updates on the emergencies happening around the globe. The Humanitarian Kiosk was established by United Nations Office for the Coordination of Humanitarian Affairs (OCHA) to meet the diverse information needs of humanitarian agencies and staff. This app enables an office to share important, up-to-date humanitarian content which can be readily accessed from a mobile device. Users may be on-the-ground humanitarian staff, or senior UN administrators attending high-level meetings. It provides easy access to maps, info graphics and information about narratives. Users can easily exchange information as it comes in a variety of formats accessible via email, SMS, QRCode and social media sites like Twitter and Facebook (UnitedNations 2018). B. FEMA (Federal Emergency Management Agency) App The application uses the Disaster management cycle identified above (Mitigation, preparation, Recovery, Response). The features provided by the application enforces the cycle of disaster management. The users can learn emergency tips covering over 20 disasters. The latter enforces the mitigation and preparation part of the cycle. The real time alerts support the response part of the cycle as it allows the users to act quickly. The application allows users to locate shelters and disaster recovery centers which tackle the recovery part partly (HomelandSecurity 2021). C. MyRadar Weather Radar MyRadar is an easy to use weather app that shows weather radar around your current location, allowing the user to see the forthcoming weather. On starting the app, the location of the user pops up with animated live radar, with radar loop lengths of up to two hours. The interface provides the best way to get a quick glimpse of the on-the-go weather and that is what made MyRadar so popular over the years. The app can send weather and environmental alerts including alerts from the National Weather Centre, such as Tornado and Severe Weather alerts. Another important feature of MyRadar is its ability to provide advanced rain alerts. The app alerts the user up to an hour before it will start raining at the user’s location. The application only covers the preparation and mitigation cycles out of the disaster management cycle (Bouchra.Azza 2021). D. My.t Weather From a local perspective, this app has been created by the Mauritius telecom. It allows authorities to send news an updates about calamities to the users. Using Windy.com the users can keep track of cyclones that are near the island and worldwide (Fig. 2). It has the necessary precautions to take during a disaster along with all the services that is available. The services are namely the police, health services and shelters. The application caters for the mitigation and preparation through the

Envirolarm...

601

Fig. 2 My.t weather screenshot

precautions and alerts. The feature that allows rescuers to be called is categorized in the response part of the disaster management cycle. The shelter location can be categorized in the recovery part partly (My.t Weather 2021).

3 Proposed Solution This research proposes a mobile app which will allow medical history; important information and information on pets will all be in one place at a touch of a finger. By allowing this information to be compiled during a non-emergent time, information will be more complete and organized then if it was gathered during the event itself. Emergency contacts included in the application will allow first responders to know who to contact if the need should arise, saving valuable time. The table below highlights the main functional requirements of the system (Table 2). Table 2 Functional requirements FR 1.0

The system shall have a profile for the user

FR FR FR FR FR FR

The system must record details of the family members The system shall be able to track the location of the user The system shall be able to display the nearby shelters The system shall be able to show the directions to the closest shelter The system shall allow users to contact the rescuers during emergencies The system shall allow the user to view the precautions to be taken in case of disasters The system must allow the user to get a list of essentials that they must have at home in case of a disaster The system shall allow users to make their own checklist The system shall be able to update users about disasters

2.0 3.0 4.0 5.0 6.0 7.0

FR 8.0 FR 9.0 FR 10.0

602

3.1

R. K. Ramjattun et al.

Use Case for Shelter Identification

The figure below shows the use case for the process behind the shelter mapping (Fig. 3).

Fig. 3 Use case for shelter identification

The location and details of the shelters are stored in a database. The application fetches the data and shows it on the application for the user to see. When selecting a shelter, the application makes a call to Google API. The latter pass the location of the user and that of the shelter as parameter. A Json is returned as a response and the data is shown on the application after processing.

3.2

Architecture Diagram

The architecture diagram shows an overview of the system. It also illustrates the structure of the application along with the names of the development tools (Fig. 4).

Envirolarm...

603

Fig. 4 Architecture diagram

3.3

Platform Chosen

Envirolarm was developed with the help of Visual studio code and React Native, which is a framework for hybrid mobile development. It supports a terminal for node server which is a must need for React native development. The expo framework has been used in addition to ease the development process. The application is compiled with android API 28 and is compatible with all previous versions (Table 3).

4 Results and Interpretation Figure 5(a) below shows the menu for an admin user. The current menu highlights in orange. The user story says the user must be able to view a list of emergency contacts and be able to make a call. The following picture shows the app displaying Table 3 Platform chosen

System software

Windows 10  64

Editor Database Server Smartphone

Visual Studio Code Firebase Node.js v12.13.0 Samsung Galaxy J1 2016

604

Fig. 5 (a) and 5(b). Main menu and Emergency contacts

Fig. 6 (a) and 6(b). Shelter location

R. K. Ramjattun et al.

Envirolarm...

605

(a)

(b)

Fig. 7 (a) and 7(b). Adding disaster types and Push notifications

a list of emergency contact. When the user clicks on the call he/she is redirected to the dialer. This is shown in Fig. 5(b). The system also includes features where the user can view a list of all precautions to be taken during the natural hazard. Envirolarm provides an Emergency Checklist so as to ensure that the user does not miss any important item during the preparation of the natural hazard. The features provided by the system includes the ability to view shelters around the Island and to find the shortest route to that specific shelter. The app shows the user location. When the user clicks on a marker the app renders the route from the user location to the shelter location. The markers show the location of the shelters. This is shown in Fig. 6(a) and (b) above. The system also allows the user to input and edit family member details. Figure 6(b) on the left shows the list of members after the user adds it. The picture on the rights shows the details of the member after the user clicks on the name of the member. This feature is helpful in case family members goes missing and this can help in search and rescue. The system also allows the admin to send an alert to users and to create push notification. The screenshots below show the alerts list. The second picture in the right shows the add page. The admin inserts the Disaster type and its description. The date and time is automatically added to the database. When the admin clicks on submit a push notification is generated and sent to all users. The push notification created is shown in Fig. 7(b) below. The performance of the application has been put to test after its implementation. The sign up process takes a couple of seconds to get the user to the login page. Additionally, the app takes around 1–2 s to retrieve the data from the database and display it. The real time

606

R. K. Ramjattun et al.

database updates the values in a negligible time frame. The navigation API takes about 3 s before returning a route and renders it. The average ram usage is 2.5 MB while the maximum is 133 MB.

5 Discussions and Conclusion In a nutshell it can be said that to mitigate risks more data needs to be made accessible. Other countries have checklists setup and plans for all disasters. The plans are accessible to the public at all times. There are also plans for kids. Education about Disaster management should start from kindergarten in order to make sure that everyone knows about the risks. Knowing about the risks is not enough and IT can help provide ways to mitigate it. This research depicts the design and development of a mobile app, Envirolarm, that helps the population at large in case of natural hazards. Mauritius being an island, is subject to many different types of hazards and being able to respond promptly in such situations can prevent the loss of lives and ensure the security of citizens. In such situations, ICT can definitely be used to come up with solutions that can help mitigate the negative effects of natural hazards.

References A. Azmeri, I. Satria, IOP Conf. Ser.: Earth Environ. Sci. 630, 012009 (2021) Bouchra Azza: Weather Radar (2021), https://play.google.com/store/apps/details?id=com.boshra. weather.widget.radar&hl=en. Accessed 28 Mar 2021 A. Dunant, M. Bebbington, T. Davies, P. Horton, Multihazards Scenario Generator: A Network‐ Based Simulation of Natural Disasters. Risk Analysis. Wiley Online Library (2021). https:// doi.org/10.1111/risa.13723 EM-DAT (2020), https://www.emdat.be/. Accessed 28 Mar 2021 HomelandSecurity: Mobile App (2021), https://www.fema.gov/mobile-app. Accessed 28 Mar 2021 A.T. Kasenally, Mauritius Country Report. Ministry of Land and Housing (2014) My.t Weather App (2021), https://www.myt.mu/mobile/mytweather-app. Accessed 28 Mar 2021 F. Shidiq, M.B.F. Bisri, N.R. Hanifa, R. Dwiyani, I. Rafliana, A. Kodijat, in Leveraging youth engagement in disaster risk reduction through science, engineering, technology, and innovation in Indonesia, ed. by R. Djalante, M.B.F. Bisri, R. Shaw. Integrated Research on Disaster Risks. Disaster Risk Reduction (Methods, Approaches and Practices). (Springer, Cham, 2021). https:// doi.org/10.1007/978-3-030-55563-4_14 K.D. Tatlow, Don’t Call It ‘Smog’ in Beijing, Call It a ‘Meteorological Disaster’ (2016), https:// www.nytimes.com/2016/12/15/world/asia/beijing-smog-pollution.html. Accessed 28 Mar 2021 UNDRR: Mauritius invests 2% GDP in resilience (2016), https://www.undrr.org/news/mauritiusinvests-2-gdp-resilience. Accessed 28 Mar 2021 UnitedNations: Humanitarian Kiosk (2018), https://play.google.com/store/apps/details?id=org. unocha&hl=en. Accessed 28 Mar 2021

Mechanical Characterization of a Geoconcrete Composite: Laterite with Addition of Peanut Shell Amadou Warore, Biram Dieng, Seydou Nourou Diop, and Senghane Mbodj

Abstract This paper presents a study on the mechanical characterization of laterite used as a building material. In Senegal, laterite has been used in construction on a semi-industrial scale since the 1990s through projects aimed at promoting local materials that can contribute to the energy efficiency of buildings. The work carried out, limited to the compression strength test, showed the possibility of using laterite and a composite (laterite + peanut shell) with satisfaction of established normative requirements governing the use of raw land in construction. The formulation BTS10 -12, 5–4 gives a compressive strength of 2,05 MPa for a minimum requirement of 2 MPa according to ARS 674:1996 and ARS 675:1996. The results obtained already indicate that the use of the peanut shell, from an environmental point of view, is of interest in managing bio-sourced waste by using it as a building material. The second interest of this work was to improve the thermal insulation of the building with this material; to be confirmed by studies of thermal characterization of the composite (laterite + peanut shell) that we have in perspective. Keywords Geoconcrete Thermal insulation


Laterite
Peanut shell
Compression strength


1 Introduction The thermal insulation character of a building envelope is directly based on the nature of the material or materials used in its construction. In the current context of energy scarcity and high energy prices, energy efficiency in buildings, particularly A. Warore (&)
B. Dieng
S. N. Diop
S. Mbodj Renewable Energy Research Team, Materials and Laser Department of Physic, UFR SATIC, University Alioune DIOP of Bambey (UADB), BP30, 21400 Bambey, Senegal e-mail: [email protected] S. Mbodj e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_49

607

608

A. Warore et al.

housing, is an imperative. However, the materials used must withstand mechanical stresses and the effects of water on the walls to ensure occupant safety. Hence the problem of the choice of materials on the basis of their mechanical characteristics. This work is part of a mechanical characterization study of a geoconcrete composite with the addition of peanut shells. The objective is to determine, through laboratory experiments, the compressive strength of the brick-made material in order to decide whether or not to use it in the construction of the building envelope. With a view to analysing the thermal performance of geoconcrete, which could be the subject of further work, the influence of granularity and peanut shell content on the mechanical properties of geoconcrete bricks will already be determined in this article.

2 Material and Method Geoconcrete bricks are bricks made with stabilized and compressed earth (laterite). Stabilization is the set of physical, chemical or physico-chemical treatment processes used to improve the mechanical characteristics of the earth. For stabilization we used cement as a stabilizer with a mass content of 10% laterite. Indeed, a stabilizer is a material that can eliminate the shrinkage effects of bricks with geoconcrete but also the protection against the effects of water on walls.

2.1

Material

Laterite Laterite or lateritic soil is soil that forms in humid tropical regions and is the result of a particular process of alteration. Laterite is a bright red or brown red soil, very rich in iron oxide and alumina formed in a warm climate [1]. The laterite that we will use for the production of the test tubes is exploited in the Mont Rolland quarry on the road to Mont Rolland Thies, an area where the soil is very lateritic. The Mont-Rolland quarry is located north of the town of Thies (Fig. 1). It developed on the soils of the Thies plateau after a chemical and mechanical alteration of sediments (clay and marl-limestone) eocenes [2]. Analysis of the particle size curve shows that the sample taken is essentially made up of aggregate with a grain size between 2 mm and 12.5 mm. Indeed, in this range, we have (93.5%–21.8%) = 71.7% in mass of the sample analyzed. Cement Cement is a widely used construction material on the planet and is its main application in concrete. Cement is a hydraulic binder, a material that allows aggregates and other materials to be joined together. Cement has several strength levels: 32, 5–42, 5–52, 5 (expressed as MPa). The higher the index, the stronger the

Mechanical Characterization of a Geoconcrete Composite…

609

Fig. 1 Location of the Mont-Rolland quarry

cement). Depending on its class of compressive strength, cement will be used in the manufacture of mortar or concrete for common purposes, structural work, or high-performance work. In our work, class 32,5 cement was used as a binder or stabilizer for sample formulation. Peanut Shell The peanut shell is a waste obtained during the shelling of the peanut pods for the recovery of the seeds. Knowing the total production, the amount of peanut shell produced per year can be calculated from the Residue-to-Product Ratios (RPR). The RPR calculation formula is as follows: RPR ¼

quantity of peanut shell quantity of seed

ð1Þ

For the peanut shell we have RPR = 0,58 [3]. An important fact to note and therefore to take into account is the share of the export of groundnut production to Senegal which is growing more and more in recent years. The exploitable potential of peanut shells will then be estimated by applying the RPR to the quantity of production processed at the local level. Sieves The sieves used comply with ISO 565, which prescribes the nominal dimensions of the openings of perforated metal sheet and electroformed sheets used as sieve bottoms in the test sieves. Scales The scales used are of electronic type and high precision. For the measurement of masses greater than 1 kg, a balance with a precision of 5 g was used.

610

A. Warore et al.

Mould For the preparation of the test pieces, we have chosen a cylindrical mould with the following dimensions: – – – –

Diameter 160 mm Height 320 mm Orthogonal Section 200 cm2 Sheet metal thickness: 8 mm The mould used is manufactured according to NF P18-400.

Press The press used is FORDIA, Model UTC-4712.FPR, manufactured for compression testing in accordance with ASTM and AASHTO standards. This machine also meets the requirements of the CE standards for operator health and safety. The press is equipped with a U-Touch PRO Control Unit system designed to automatically perform compression, bending and breaking strength tests on building materials such as concrete, Cement mortar, masonry blocks by controlling automatic compression/bending testing machines (Fig. 2).

2.2

Sampling Device

The laterite used is taken from a quarry on the road to Mont Rolland de Thiès, where the soil is very lateritic. The experimental work is carried out in the laboratories of the company called Tout Faire Géotechnique (TFG) based in Thies located 70 km from Dakar the capital of Senegal and Diamniadio which is a commune of the Dakar region. The treatment of the basic material (laterite) is done in the laboratory in order to obtain several samples on the basis of which the different specimens to be characterized will be manufactured.

Fig. 2 Press for compressive strength tests_ FORDIA Model: UTC-4712.FPR

Mechanical Characterization of a Geoconcrete Composite…

611

After processing the material (crushing, drying and sieving), two different batches of test pieces will be prepared. This will be a batch of specimens prepared with different granularities; a second batch consisting of composite material, consisting of 12,5 mm of laterite mixed with the sifted peanut shell with a sieve of 5 mm and with different peanut shell contents.

3 Effects of Particle Size on Geoconcrete Mechanical Properties The basic material used is laterite previously studied, with a different granularity for the different samples.

3.1

Preparation of Material and Preparation of Test Specimens

The preparation and preservation of the test pieces is done according to the standard NF P 18–404. To obtain different grain sizes, a treatment is carried out by sieving after natural drying of the laterite by spreading with a thin layer for more than 7 days in the open air and in the sun. It should be understood here that the consideration we have taken to speak of different particle size corresponds to the use of passers-by after sieving laterite with three different sieve sizes. Indeed, we used loops of mesh sieves 5 mm, 12.5 mm and 20 mm. The amount of spoilage water was taken equal to 10.2% representing Wopt. found per Proctor test. We therefore took a mass of water equal to 10.2% of the laterite mass. Cement is used as a binder with a quantity equal to 10% by mass of laterite (10% of the mass of laterite is replaced with cement for the same quantity). Table 1 below gives a summary of the composition of the samples of the material used in the preparation of the specimens.

Table 1 Mass composition of samples Designation of test pieces

Latérite (kg)

Cement (kg)

water (liter)

BTS10P5 BTS10P12,5 BTS10P20

11,25 11,25 11,25

1,25 1,25 1,25

1,3 1,3 1,3

612

A. Warore et al.

Fig. 3 Geoconcrete test piece with passing sieves of different diameters

Compression energy is evaluated as follows: ðnumber of strokes per layer Þ
ðnumber of layerÞ
ðtamper mass (in g)Þ Compression energy ðJ=m3 Þ ¼


ðfall height of the tamperÞ ð2Þ useful volume of feed

The application of a compaction energy of 2435 kJ.m−3 resulted in the following Fig. 3:

3.2

Mechanical Test Results

The tests are carried out on 28-day-old specimens after drying with cure. The purpose of curing the specimens is to preserve the properties of durability and consists of protecting the specimens from early drying for a sufficient period, for example by placing them under a tarpaulin, protected from direct solar radiation. Figure 4 below shows the installation for the determination of density and the performance of the compressive strength tests. The value of Rc (Compressive strength) is obtained by applying the following formula: Fmax S Fmax max imum force applied to the surface of the specimen

Rc ¼

S : surface of the specimen The results of the compression tests are given in Table 2 below.

ð3Þ

Mechanical Characterization of a Geoconcrete Composite…

613

Fig. 4 Measurement of the mass of the specimen (a) and Compressive strength test (b)

Table 2 Mechanical characteristics of geoconcrete specimens Designation of test pieces

Compaction energy (kJ. m-3)

Density of dried test piece (kg. m-3)

Compressive breaking force (Fc en kN)

Compressive strength (Rc en MPa)

BTS10P5 BTS10P12,5 BTS10P20

2435 2435 2435

1884 2073 2194

40,7 44,6 27,08

2 2,22 1,35

From the results presented in Table 2, the diagrams in Fig. 5 and 6 below have been plotted and analysed. As already noted in the analysis of the diagram in Fig. 5, the greatest compressive strength is obtained on the BTS10P12.5 specimens However, for BTS10P20 specimens, we have a lower compressive strength than for BTS10P12.5 specimens although the density always keeps the proportionality for an increase in the granularity of the material. Compaction is easier when the mixing water content is optimal. We can then note that cement has much more possibility to bind the

compressive strength (MPa) 2.5 2 1.5

2

2.22 1.35

1 0.5 0

Résistance en compression (Rc en MPa)

BTS10 P5

BTS10 P12,5

BTS10 P20

2

2.22

1.35

Fig. 5 Diagram of compressive strength for specimens with different granularity

614

A. Warore et al.

Compressive strength (MPa)

20, 2194

12.5, 2.22

2.5 2 5, 2

1.5

12.5, 2073

1

0

20, 1.35

5, 1884

0.5

5

12.5

20

Masse volumique de l’éprouvee séchée (kg.m-3)

1884

2073

2194

Résistance en compression (Rc en MPa)

2

2.22

1.35

Fig. 6 Effect of particle size on compressive strength and density

4 Effects of Peanut Shell Content on Geoconcrete Mechanical Properties The basic materials are the laterite and the peanut shell.

2250 2200 2150 2100 2050 2000 1950 1900 1850 1800 1750 1700

Density (kg,m-3

elements together and thus to increase the compressive strength, when the laterite used is made up of gravel (grain size between 2 mm and 12,5 mm). Analysis of the graphs in Fig. 5 shows that the greatest compressive strength is obtained with BTS10P12,5. The results obtained confirm the choice in previous works [4, 5], of the laterite of particle size less than 12,5 mm (sieve passes 12,5 mm). The results of the tests carried out are supplemented by the analysis of the graphs in Fig. 6 below. Analysis of the graphs in Fig. 6 shows an increase in density for an increase in grain size. This physical phenomenon is due to the increase in the intra-granular porosity of the compacting material when the size of the particles (grains) increases. The air contained in the porous space will be replaced by the mixing water, thus increasing. Indeed, a compromise must be found for the choice of grain size. Too thin, the laterite used requires, to have a good resistance in compression, much more binder (cement) and a greater quantity of water mixing to well bind the elements between them; too large, the elements that make up the earth to be compressed are less consistent, which leads to a decrease in compressive strength.

Mechanical Characterization of a Geoconcrete Composite…

4.1

615

Preparation of Material and Preparation of Test Specimens

The preparation and preservation of the test pieces is done according to the standard NF P 18–404. The peanut shell is dried in the open air and in the sun for more than 7 days. This ensures that its natural water content is almost negligible. Then it is crushed and sieved with a 5 mm sieve. In the series of specimens, laterite passing from sieve 12.5 mm is mixed with peanut shell at different percentages (4%, 6% and 8%). The amount of spoilage water was taken equal to 10.2% of the laterite mass. Cement is used as a binder with a quantity equal to 10% by mass of laterite (10% of the mass of laterite is replaced with cement for the same quantity). Table 3 below gives a summary of the mass composition of the samples of the material used in the preparation of the specimens (Fig. 7).

Table 3 Mass composition of composite samples with pass-through laterite 12.5 mm BTS10 -12,5 - 4 BTS10 -12,5 - 6 BTS10 -12,5 - 8

Laterite (kg)

Peanut shell (kg)

Cement (kg)

Water (liter)

10,75 10,50 10,25

0.5 0.75 1

1,25 1,25 1,25

1,3 1,3 1,3

Fig. 7 Composite specimens (laterite + peanut shell)

4.2

Mechanical Test Results

The results of the mechanical compression tests are summarized in Table 4 below. From the results obtained and given in Table 4, we draw the following graphs: Fig. 8: Influence of the peanut shell content on the density for different granularities and Fig. 9: Effect of peanut shell content on compressive strength. Through the analysis of the graph (Fig. 8), we find that the density of the test piece decreases for an increase in the peanut shell content.

616

A. Warore et al.

Table 4 Mechanical characteristics of ground concrete specimens with added Peanut Shell – Laterite passing 12.5 mm Designation

Compaction energy (kJ. m-3)

Density of dried test piece (kg.m-3)

Compressive strength (Rc en MPa)

mean

Compressive breaking force (Fc en kN) mean

BTS10 -12,5 - 4 (éch.1) BTS10 -12,5 - 4 (éch.2) BTS10-12,5 - 6 (éch.1) BTS10-12,5 - 6 (éch.2) BTS10-12,5 - 8 (éch.1) BTS10-12,5 - 8 (éch.2)

2435

1823,75

1836,46

42,7

41,3

2,05

2435

1849,17

2435

1731,72

38,1

1,89

2435

1713,89

2435

1413,79

23,65

1,17

2435

1465,18

39,9 1722,80

36,8 39,4

1439,48

22,7 24,6

2000

6, 1722.8

1000 8, 1439.48 500

Peanut shell content (%) 6 8

4 Density of the dried specimen (kg.m-3)

1836.46

1722.8

0

Density (kg.m-3)

1500 4, 1836.46

1439.48

Compressive strength

Fig. 8 Effect of peanut shell content on density

2.5 2 1.5 1 0.5 0

Compressive strength

4, 2.05 6, 1.89 8, 1.17

1 2.05

Peanut shell content 2 1.89

Fig. 9 Effect of peanut shell content on compressive strength

3 1.17

Mechanical Characterization of a Geoconcrete Composite…

617

2.5

4, 1836.46

6, 1722.8

8, 1439.48

2 1.5

4, 2.05 6, 1.89

1 8, 1.17

0.5 0

Density of the dried specimen (kg.m-3) Compressive strength (MPa)

Peanut shell content 4

6

8

1836.46

1722.8

1439.48

2.05

1.89

1.17

2000 1800 1600 1400 1200 1000 800 600 400 200 0

Density of the dried specimen (kg.m-3)

Compressive strength (MPa)

Since the peanut shell is lighter than the laterite, increasing its content in the composite material results in a decrease in density. The change in density is rather large with the change in peanut shell content. We note that the BTS10-12.5 test piece measured at 4% peanut shell gives a mechanical strength higher than the minimum value required by ARS 674:1996 and ARS 675:1996 which is 2 MPa. On the other hand, the 6% and 8% peanut shell dosages give a compression resistance lower than the minimum required. From the graph in Fig. 9, we find that the mechanical compressive strength decreases for an increase in the peanut shell content. Thus the mechanical strength is inversely proportional to the peanut shell content. By going from 4 to 8% hull strength decreases by almost 1 MPa. To find an explanation of the behaviour of the composite material studied in the face of mechanical stresses, the combined graph of the influence of the peanut shell content on the density and mechanical strength is shown in the next Fig. 10.

Fig. 10 Effect of peanut shell content on compressive strength and density

Analysis of the graph in Fig. 10 shows that the density and compressive strength change in the same direction for a change in the peanut shell content. They decrease for an increase in the peanut shell content. They are therefore inversely proportional to the peanut shell content. Indeed, the decrease in density is due to the presence of more and more air-filled pores. When the peanut shell content is increased, the presence of pores in the material reduces its mechanical compressive strength.

618

A. Warore et al.

5 Conclusion The results of the compression tests confirm the possibility of using the 10 mm and 12.5 mm pass-through laterite taken from the Mont Rolland quarry for building envelope construction. Indeed, the compressive strength of this laterite exceeds the minimum required by ARS 674:1996 and ARS 675:1996. The mechanical tests carried out on the series of composite specimens give a compressive strength of the sample BTS10-12,5–4 meeting the requirements of ARS 674:1996 and ARS 675:1996. Thus, these materials could be used for the construction of a building envelope, provided that the results of the thermal tests we have in perspective are interesting. The results we obtained from the mechanical characterization have thus motivated our ambition to follow up our work for studies of thermal characterization of laterite composite with addition of peanut shell.

References 1. Z.P.B. Bohi, Caractérisation des sols latéritiques utilisés en construction routière : cas de la région de l’Agneby (Côté d’Ivoire). Mécanique des matériaux [physics.class-ph], Ecole des Ponts ParisTech,. Français. NNT : 2008ENPC0834. pastel-00503010 (2008) 2. Strasbourg: Institut de Géologie – Université Louis-Pasteur, Genèse des phosphates alumineux du Sénégal occidental. Étapes et guides de l’altération. (Sciences Géologiques. Mémoire, 67), pp. 3–230 (1982). https://www.persee.fr/doc/sgeol_0302-2684_1982_mon_67_1 3. A. Diedhiou, Étude hydrodynamique et valorisation énergétique par transformation thermochimique de déchets de biomasse pour l’alimentation d’une briqueterie, Thèse présentée en cotutelle pour l’obtention du grade de Docteur de l’UTC, 28 April 2017 4. P. Meukam et al., Thermophysical and mechanical characterization of stabilized clay bricks for building thermal insulation. Matériaux et Constructions 36, 453–460 (2003) 5. H. Beshr, A.A. Almusallam, M. Maslehuddin, Effect of coarse aggregate quality on the mechanical properties of high strength concrete. Constr. Build. Mater. 17, 97–103 (2003)

Smart Water Management

Characteristics and Assessment of Heavy Metals in the Water of Lake Sidi Boughaba (Kenitra, Morocco) Mohamed Najy , Fatima Zahra Talbi , Hassan Ech-chafay, Omar Akkaoui, Nordine Nouayti, and Driss Belghyti

Abstract The goal of the presented research was to evaluate the heavy metal detection and potential ecological risks in Lake’s water. Here are the concentrations of heavy metals, which include Cu, Zn, Mn, Fe, Pb and As in the water which was studied. Samples for analysis were taken from six sites of the Sidi Boughaba Lake (Mehdia, Morocco). Our results showed that the mean metal concentrations in lake water were 0.472 for Cu, 0.657 for As, 1.02 for Fe, 0.28 for Mn, 0.076 for Pb and 0.091 for Zn. The existence of these metallic elements can be of natural origin as long as there are no direct pollutants discharged into this environment. Keywords Assessment


Heavy metals
Sidi Boughaba
Lake

1 Introduction Lakes are hollows which are filled, at least partially, with water [1], important environments, aims at the preservation of fresh water, supply and replenishment of groundwater. Indeed, changes are made between these environments and the surface water table, whether quantitative or qualitative, when the topographic surface is located at a lower coast than that of the piezometric surface of the water table. As much as, they play a very important ecological role, in adjusting the local climate and improving the environment [2]. M. Najy (&)
H. Ech-chafay
O. Akkaoui
D. Belghyti Natural Resources and Sustainable Development Laboratory, Faculty of Sciences, University Ibn Tofail, BP133, 14000 Kenitra, Morocco F. Z. Talbi Faculty of Sciences and Technologies, Laboratory of Biochemistry, Neurosciences, Natural Resources and Environment, Hassan First University of Settat, 577 Settat, Morocco N. Nouayti Laboratory of Applied Sciences, Research Team of Water and Environment Management (G2E), National School of Applied Sciences/Abdelmalek Essaadi University, Al Hoceima, Morocco © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_50

621

622

M. Najy et al.

The main anthropogenic sources are associated with mining activities, the metallurgical industry, fertilizers and pesticides used in agriculture and domestics effluents [3, 4]. Mobility and the degree to which a metal is absorbed or precipitated in water depends notably on its intrinsic properties, the physico-chemical parameters of the environment [5, 6]. The presence of metals at levels of abnormally high concentrations relative to the geochemical background may form harmful complex compounds and with potential impacts on the environment [7]. In this study, Sidi Boughaba Lake was selected to study the geochemical and polluant characteristics of heavy metals, including Mn, Cu, Zn, Pb, As and Fe. The possible geochemical sources of these metals were evaluated by correlation analyzes and major components.

2 Methodology 2.1

General Properties of the Lake

Sidi Boughaba Lake is located on the Atlantic coast of northwestern Morocco, oriented NNE - SSW [8] and is sited in an interdunal depression. It stretches for 5.5 km in length, and a variable width from 100 to 350 m, and a depth that varies between 0.5 and 2.50 m maximum. The existence of this body of water is due to the fact that the topographic surface is at a shallower coast than the piezometric surface of the coastal water table, rainwater and runoff [9].

2.2

Statistical Analysis

All multivariate statistical analyzes, including Principal Component Analysis (PCA), Correlation Matrix (CM) were performed using SPSS 20.0 and Origin 8 Trial Version. Principal component analysis (PCA), is part of multivariate descriptive analyzes usable as a tool to help interpret a data matrix. This analysis makes it possible to synthesize and classify a large amount of data in order to extract the main factors at the origin of the simultaneous evolution of variables or individuals and their relationships. The ACP makes it possible to highlight the similarities and the graphical position that present two or more variables during their evolution. Heavy metal concentrations were tested to establish correlations

Characteristics and Assessment of Heavy Metals…

623

between them using the Pearson correlation coefficient, assuming that the data was normally distributed. Differences between concentrations are considered significant if P < 0.05.

3 Results and Discussion 3.1

General Properties of the Lake

The physical and chemical characteristics of the water of Sidi Boughaba Lake, which were measured during the sampling, will be indicated in Table 1. It was observed that the average pH of the water sample ranged from 7.05 to 8.36 (Fig. 1), but did not fluctuate significantly between sampling sites, suggesting that it was slightly alkaline to moderately alkaline.

Table 1 Summary of heavy metal concentration’s statistics in surface water from the Sidi Boughaba Lake. All the concentrations are in mg/l

pH

Cu

As

Fe

Mn

Zn

Pb

Mean

7.60

0.047

0.657

1.02

0.286

0.091

0.076

Median

7.44

0.041

0.008

0.69

0.086

0.061

0.030

0.001

0.006

0.125

0.039

0.001

0.004 0.101

Mode SD

0.41

0.069

1.288

0.86

0.347

0.090

Variance

0.17

0,005

1,66

0.75

0.121

0.008

0.010

0.352

3.939

2.09

0.925

0.231

0.354

Range

Fig. 1 Spatial variation of PH in the water of Lake Sidi Boughaba during the study period

Min

7.05

0.001

0.002

0.125

0.039

0.001

0.004

Max

8.36

0.354

3.942

2.21

0.965

0.233

0.359

624

M. Najy et al.

Fig. 2 Boxplots of heavy metal concentrations in the water of Sidi Boughaba Lake

The results showed that the average concentrations of As, Fe, Mn and Pb were higher than the drinking water equivalent (DWEL) threshold, whereas they were lower for Cu and Zn. However, they were higher than the threshold concentration for aquatic life tolerance (TC) Fig. 2. Figure 2 shows the spatial variation of these metal concentrations in different lake sites. Generally, remarkable fluctuations were recorded in the same site during the study period. This can be explained by the variation of several factors such as tempe ature, pH and organic matter. The average concentrations of most metals were higher in the study area of the samples taken. The mean concentration of Fe, Cu, Mn, As and Pb in water exceeded the WHO safety limits, with the exception of Zn, and the maximum concentrations of these metals exceeded the WHO safety limits.

Characteristics and Assessment of Heavy Metals… Table 2 Acceptable levels of some heavy metals in natural waters according to WHO [2011] standards [16–18]

Limit

Cu

625 Pb

Zn

Fe

Mn

MCL 1.3 0 – (mg/l) MAC 1 0.015 3 0.1 0.05 (mg/l) TC (mg/ 0.004 0.007 0.05 0.3 0.05 l) DWEL 2 0.01 3 0.2 0.4 (mg/l) MAC: Maximum allowable concentrations. MCL: Maximum contaminant level TC: Threshold concentration for aquatic life tolerance. DWEL: Drinking water equivalent level. NG: No Guideline.

As

0.05 NG 0.05

Some of these metals considered as oleo-elements (Cu, Fe, Zn and Mn) are needed as trace nutrients for vital processes in plants and micro-organisms, but become toxic to plants higher concentrations [10]. heavy metals have been widely studied for their toxic effects [11, 12], accumulation in organisms [13, 14], and bio-accumulation in food chains [15]. Based on the standards set by WHO mentioned in (Table 2), the results noted that the levels of metallic trace elements in the waters of Lake Sidi Boughaba greatly exceed the maximum level of contamination (MCL), maximum admissible concentration (MAC) and above the threshold for the concentration to the tolerance to aquiatic life, with regard to the following elements (As, Pb, Fe and Mn), However, the content of Cu remains in conformity for MCL and MAC, which allows us to say that, the waters of the Lake are much polluted with the metallic point of life. In this regard, these elements in high concentration could prevent the existence of some aquatic animals such as benthic macroinvertebrates.

3.2 3.2.1

Multivariate Analysis Correlation Matrix (MC) and Component Principal Analysis (CPA)

Table 3 Showed that the majority of heavy metals are strongly positively correlated with each other due to their higher concentration in water. Positive correlation coefficients are observed among the metals Fe, Mn, Pb, Zn and other elements, such as Pb and Cu, which have very low correlation. High positive correlation coefficient between Fe and Zn (r = 0.96); Mn and Pb (r = 0.78).

626 Table 3 Correlation coefficients for the elements during the study period

M. Najy et al. Elements Cu

Cu

Zn

Pb

Fe

Mn

As

1.000

Zn

0.358

Pb

0.62

1.000 0.556

1.000

Fe

0.429

0.961

0.508

Mn

−0.016

0.397

0.787

0.373

1.000

As

−0.336

−0.504

−0.186

−0.517

0.011

1.000 1

According to the Table 3, lead is strongly positively correlated with magnesium (r = 0.78), suggesting that dissolved Pb could be derived from dolomite, moreover; in addition, Pb is moderately correlated with Fe (r = 0.5) indicating iron hydroxide as a possible source. Zinc is strongly correlated with Fe (0.65), suggesting a release of this metal under the reducing conditions of iron. Negative correlation for As with Fe (r = −0.51), Zn (r = −0.50) suggests the metals do not have common sources [19], and this might be due to the variation in the sources these metals are bound weakly to the hydrous clay minerals. In addition, a possible agricultural origin from fertilizers [20]. The projection of the variables on the factorial plane F1  F2 (76.98%) shows that the factor F1 is the most important. He alone controls about 52.37% of total inertia. The metallic elements of the waters of Lake Sidi Boughaba are grouped into three components; the first is controlled by the following elements: Zn, Cu and Fe. The results of the PCA that based on the rotation of the characteristic value of 1 showed that six heavy metals in the water of Sidi Boughaba Lake can be identified as three principal components, accounting for 88.3% of the total variance, include 52.37%, 24.61 and 11.32 of the total variance, respectively (Fig. 3). Concerning the first main component, the heavy-metal Zn, Cu and Cu have high loads. A second group, including Pb and Mn, could be ascribing to an anthropogenic factor. The spatial distribution characteristic of As is quite different from that of the remaining heavy metals, with a relatively higher concentration appearing in the S3, S4, S5 and S6 sites. Main natural sources of arsenic include basin-fill deposits [21, 22], geothermal and volcanic activities [23].

Characteristics and Assessment of Heavy Metals…

627

Fig. 3 Principal component analysis (PCA) for the different lakes and variables

4 Conclusion The results of the present study show that the levels of metallic trace elements in the waters of Lake Sidi Boughaba present a high risk for aquatic life. Based on the various recommendations for water quality, it is revealed that almost all of the elements studied exceed the standard values , which signify moderate to frequent threats.

References 1. D. Uhlmann, L. Paul, M. Hupfer, R. Fischer, 2.08 - Lakes and Reservoirs A2 - Wilderer, Peter. in Treatise on Water Science (Elsevier, Oxford, 2011), pp. 157–213. https://doi.org/10. 1016/B978-0-444-53199-5.00034-8 2. D. Hou et al., Distribution characteristics and potential ecological risk assessment of heavy metals (Cu, Pb, Zn, Cd) in water and sediments from Lake Dalinouer, China. Ecotoxicol. Environ. Saf. 93, 135–144 (2013). Consulté le: oct. 04, 2017. [En ligne]. Disponible sur: http://www.sciencedirect.com/science/article/pii/S0147651313001000 3. G. Zarazua, P. Ávila-Pérez, S. Tejeda, I. Barcelo-Quintal, T. Martínez, Analysis of total and dissolved heavy metals in surface water of a Mexican polluted river by total reflection X-ray fluorescence spectrometry. Spectrochim. Acta Part B At. Spectrosc. 61(10), 1180–1184 (2006). Consulté le: oct. 01, 2017. [En ligne]. Disponible sur: http://www.sciencedirect.com/ science/article/pii/S0584854706001789 4. N. Nehme et al., The distribution of heavy metals in the lower River Basin, Lebanon. In: 8th International Conference on Materials Science CSM8-ISM5, vol. 55, no Supplement C (January 2014), pp. 456–463. https://doi.org/10.1016/j.phpro.2014.07.066 5. K.M. Mohiuddin, K. Otomo, Y. Ogawa, N. Shikazono, Seasonal and spatial distribution of trace elements in the water and sediments of the Tsurumi River in Japan. Environ. Monit. Assess. 184(1), 265–279 (2012). Consulté le: oct. 03, 2017. [En ligne]. Disponible sur: http:// www.springerlink.com/index/X37G0K8514223843.pdf 6. M.S. Islam, M.K. Ahmed, M. Raknuzzaman, M. Habibullah-Al-Mamun, M.K. Islam, Heavy metal pollution in surface water and sediment: a preliminary assessment of an urban river in a developing country. Ecol. Indic. 48, 282–291 (2015). Consulté le: oct. 03, 2017. [En ligne]. Disponible sur: http://www.sciencedirect.com/science/article/pii/S1470160X14003719

628

M. Najy et al.

7. A. Rajbanshi, Study on heavy metal resistant bacteria in Guheswori sewage treatment plant. Our Nat. 6(1), 52–57 (2009). Consulté le: oct. 05, 2017. [En ligne]. Disponible sur: http:// www.nepjol.info/index.php/on/article/view/1655 8. M. Najy et al., Assessment of the water quality of Lake Sidi Boughaba (Ramsar Site 1980) Kenitra, Morocco. Invent. J. Res. Technol. Eng. Manag. 2(5), 52–60 (2018) 9. M. Thévenot, Les oiseaux de la réserve de Sidi-Bou-Rhaba, Bull. Inst. Sci. Rabat 1, 68–90 (1976). Consulté le: mai 02, 2017. [En ligne]. Disponible sur: https://www.israbat.ac.ma/wpcontent/uploads/2015/01/BIS_1_67-99.pdf 10. D. Paul, Research on heavy metal pollution of river Ganga: a review. Ann. Agrar. Sci. 15(2), 278–286 (2017). https://doi.org/10.1016/j.aasci.2017.04.001 11. A. Sáinz, J.A. Grande, M.L. De la Torre, Characterisation of heavy metal discharge into the Ria of Huelva. Environ. Int. 30(4), 557–566 (2004) 12. M. Olías, C.R. Cánovas, J.M. Nieto, A.M. Sarmiento, Evaluation of the dissolved contaminant load transported by the Tinto and Odiel rivers (South West Spain). Appl. Geochem. 21(10), 1733–1749 (2006) 13. W. Ashraf, Levels of selected heavy metals in tuna fish. Arab. J. Sci. Eng. 31(1A), 89 (2006) 14. Y. Tao, Z. Yuan, H. Xiaona, M. Wei, Distribution and bioaccumulation of heavy metals in aquatic organisms of different trophic levels and potential health risk assessment from Taihu Lake, China. Ecotoxicol. Environ. Saf. 81(Supplement C), 55–64 (2012). https://doi.org/10. 1016/j.ecoenv.2012.04.014 15. D.X. Soto, R. Roig, E. Gacia, J. Catalan, Differential accumulation of mercury and other trace metals in the food web components of a reservoir impacted by a chlor-alkali plant (Flix, Ebro River, Spain): implications for biomonitoring. Environ. Pollut. 159(6), 1481–1489 (2011) 16. D. Hou et al., Distribution characteristics and potential ecological risk assessment of heavy metals (Cu, Pb, Zn, Cd) in water and sediments from Lake Dalinouer, China, Ecotoxicol. Environ. Saf. 93(Supplement C), 135–144 (2013). https://doi.org/10.1016/j.ecoenv.2013.03. 012 17. M.E. Goher, A.M. Hassan, I.A. Abdel-Moniem, A.H. Fahmy, S.M. El-Sayed, Evaluation of surface water quality and heavy metal indices of Ismailia Canal, Nile River, Egypt. Egypt. J. Aquat. Res. 40(3), 225–233 (2014). https://doi.org/10.1016/j.ejar.2014.09.001 18. M. Chabukdhara, S.K. Gupta, Y. Kotecha, A.K. Nema, Groundwater quality in Ghaziabad district, Uttar Pradesh, India: Multivariate and health risk assessment. Chemosphere 179 (Supplement C), 167–178 (2017). https://doi.org/10.1016/j.chemosphere.2017.03.086 19. Y. Wang, L. Yang, L. Kong, E. Liu, L. Wang, J. Zhu, Spatial distribution, ecological risk assessment and source identification for heavy metals in surface sediments from Dongping Lake, Shandong, East China. CATENA 125(Supplement C), 200–205 (2015). https://doi.org/ 10.1016/j.catena.2014.10.023 20. N. Colombani, M. Mastrocicco, E. Dinelli, Trace elements mobility in a saline coastal aquifer of the Po river lowland (Italy). J. Geochem. Explor. 159(Supplement C), 317–328 (2015). https://doi.org/10.1016/j.gexplo.2015.10.009 21. J. Wang, G. Liu, H. Liu, P.K. Lam, Multivariate statistical evaluation of dissolved trace elements and a water quality assessment in the middle reaches of Huaihe River, Anhui, China. Sci. Total Environ. 583, 421–431 (2017) 22. S. Wang, K. Luo, R. Ni, Y. Tian, X. Gao, Assessment of elemental background values and their relation with lifespan indicators: a comparative study of Jining in Shandong Province and Guanzhong area in Shaanxi Province, northern China. Sci. Total Environ. 595, 315–324 (2017) 23. R. Bondu, V. Cloutier, E. Rosa, M. Benzaazoua, Mobility and speciation of geogenic arsenic in bedrock groundwater from the Canadian Shield in western Quebec, Canada. Sci. Total Environ. 574, 509–519 (2017)

Multiple Water Reservoirs in African Continent: Scarcity, Abundance and Distribution Ahmed El Bakouri , Mourad Bouita, Fouad Dimane, Mohamed Tayebi, and Driss Belghyti

Abstract The focus of this article is to give an overview of inland water bodies (lakes, dams and lagoons) and surface water bodies (rivers and wetlands), as well as the various groundwater reserves (water tables and aquifers). These natural water reservoirs, its distribution in Africa plays a fundamental role in the constraint of its geological evolution and habitability. The aquifers constitute good underground water reservoirs, from fissured or fractured rocks, allowing a water supply, which are less sensitive to climatic variations. On the other hand, surface waters are more sensitive to pollution and drought. The framework for sustaining and preserving these resources is good environmental management of the various watersheds and coastal zone planning. In addition, the control of groundwater pumping, to avoid a drop in piezometric levels. Keywords Inland water bodies Watersheds Africa





Surface water bodies
Rivers
Aquifers


The original version of this chapter was revised: The affiliation of Driss Belghyti has been updated correctly. The correction to this chapter is available at https://doi.org/10.1007/978-3-030-94191-8_94 A. El Bakouri (&)
D. Belghyti Laboratory of Natural Resources and Sustainable Development, Faculty of Sciences, Ibn Tofail University, P.O. Box: 242, 14000 Kenitra, Morocco e-mail: [email protected] M. Bouita Laboratory of Materials Physics and Subatomics, Faculty of Sciences, Ibn Tofail University, P.O. Box: 242, 14000 Kenitra, Morocco F. Dimane Laboratory of Applied Sciences, Department of Civil Engineering, Energy and Environment, National School of Applied Sciences of Al Hoceima, Abdelmalek Essaadi University, P.O Box: 03, 32000 Ajdir, Al Hoceima, Morocco M. Tayebi Laboratory of Geosciences, Faculty of Sciences, Ibn Tofail University, P.O. Box: 242, 14000 Kenitra, Morocco © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022, corrected publication 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_51

629

630

A. El Bakouri et al.

1 Introduction Water is a very important factor for socio-economic development. Africa has a considerable amount of water resources, but suffers from chronic deficits, due to the uneven distribution of rainfall and runoff in time and space. Groundwater reserves are linked to flows maintained by the water cycle and are therefore, to a large extent, renewable, by recharging the aquifers from rainfall infiltration [33]. A particular type of water reserves is represented by the so-called fossil aquifers, whose exploitation can be compared to that of a mining deposit [3]. The surface water systems of Africa are diverse, the continent has seventeen major rivers and a hundred lakes, however, this resource is poorly distributed between Africa of potential scarcity in the North, Africa of water shortage or Saharan and Sub-Saharan Africa and finally a third Africa, that of excess water, in the equatorial zone [5]. Africa, although it has a significant potential in water resources, it remains confronted with a number of situations, by the increasing pressure of water withdrawals, irrigation, water pollution by untreated wastewater discharges and climate change which constitute the greatest threat to water resources, by the drought that frequently affects many regions, which causes a reduction of wetlands in several African countries and a deficit of water supply.

2 Related Works In this section, we discuss existing models for modeling water resources, as well as the entire hydrological system whether at the watershed or continental scale. We give two models: – Water Resources System Model (WRSM): The model serves as a decision support system that evaluates the capacity of existing and proposed water resources systems by simulating physical, water quality, statistical and operational aspects. This required long-term time series of hydrological and climatic inputs for calibration and validation, as well as analysis of hydrogeological data, through the description of the basic state and management of water resources. At the scale of large watersheds, studies can be carried out using GIS tools (Arc Hydro). – Performance Based Water Resources Engineering: the performance based approach to water resources engineering, including risks such as flooding, drought, and to develop planning, design, and operation procedures in which the consequences of competing risks are properly balanced and investments in damage reduction and recovery can be made appropriately [35]. The computer program of the model is then operated with various input data, the detailed analysis of the output is the last step of the simulation process. Climate models,

Multiple Water Reservoirs in African Continent…

631

simulating global thermal changes on natural and anthropogenic origins on global warming and then their impact on water resources, generally concerning, precipitation, temperature and evapotranspiration.

3 Proposed Method The Water Budget Methods focuses on the different components of groundwater flow and storage variations, considering only the inputs and outputs of the aquifer. Water budgets are fundamental to the conceptualization of hydrologic systems at all scales. Initial analysis of a water balance can provide insight into the suitability of any recharge estimation technique. Refinement of the water balance and the conceptual model of the system throughout the life of a study can further guide study efforts [17]. The water balance equation for a watershed and the underlying unsaturated and saturated zones can be written: P þ Qon ¼ ET þ DS þ Qoff

ð1Þ

Where: P is Precipitation, Qon is surface and groundwater flow into the watershed, ET is Evapotranspiration, DS is stock change, and Qoff is surface and groundwater flow out of the watershed. Water flow in the watershed can be written as the sum of surface water flow gw (Qsw on ) and groundwater flow (Qon ): gw Qon ¼ Qsw on þ Qon

ð2Þ

Evapotranspiration can be divided according to the source of evaporated water: ET ¼ ET sw þ ET gw þ ET uz

ð3Þ

Where ETsw is the evaporation or sublimation of water stored on the soil surface, ETuz is the evaporation from bare soil and plant transpiration of water stored in the unsaturated zone, and ETgw is the evapotranspiration of water stored in the saturated zone. The challenge of water in the watershed, or at the continental scale, involves all components: the physical, climatic, variability of rainfall in time and space, intensity of evaporation and overexploitation of aquifer systems that causes the drying up and decrease in the flow of sources and lowering of the piezometric level, as well as several rivers can become temporary or dry (see Fig. 1).

632

A. El Bakouri et al.

Water Management Climate Regular and irregular precipitation in time and space Unequal temperatures in time and space Drought Water resources Groundwater (Overuse) Surface water (irregularity)

Field structure

Topography, Hydrogeology, Hydrology Surface water (irregularity)

Water challenges

Management of aquifer systems and watersheds Construction of dams for a good balance of rivers Managing water in a context of climate change

Fig. 1 Work flowchart applied to produce the water resources geodatabase for the Africa

4 Description of Study Area 4.1

Topography and Climate

Africa has many mountains, as all or almost all of the land in Lesotho, Rwanda, and Burundi is at or above 1500 m. Mountains that rise above 4500 m are concentrated in the northwestern, central and eastern regions. By satellite images we can identified several mountains which are mentioned in Table 1:

Multiple Water Reservoirs in African Continent… Table 1 Topographical characteristics of the main mountains of Africa

Mountain

633 Geographical coordinates

Elevation

Latitude

Longitude

Kilimandjaro

−3.06742470

37.35562730

5826

Mount Kenya

−0.15213840

37.30840790

5083

Rwenzori

0.38583300

29.87166700

4965

Mount Karisimbi

−1.50640560

29.45076780

4477

Thabana Ntlenyana

−29.46795800

29.26909090

3465

Mount Heha

−3.60305560

29.49944440

2656

– Thabana Ntlenyana, is a black mountain located in Lesotho at 3465 m altitude – Mount Karisimbi is a volcano located on the border separating Rwanda from the Democratic Republic of Congo, with an altitude of 4477 m – Mount Heha, located in the province of Bujumbura rural at 2656 m altitude – Kilimanjaro, located in northeastern Tanzania, consisting of three volcanoes: the Shira, the Mawenzi and the Kibo, with an altitude of 5826 m at Uhuru Peak in Kibo volcano – Mount Kenya is the highest point of Kenya and the second highest peak in Africa, behind Kilimanjaro at 5083 m altitude – Rewenzori, is a small mountain range in Central Africa, located on the border between Uganda and the Democratic Republic of Congo, culminating at 4965 m at Mount Stanley Peak Topographic elevations and proximity to the equator cause seasonal variations. These characteristics create climatic variability that is sometimes aggravated by cycles of flooding and drought. (see Fig. 2). Topographic features and temperature differences between the sea and land surface influence the climatic differences between the eastern and western parts of the continent. Groundwater distribution is based on topography as shallow aquifers are found near natural ponds or in the hollows of mountains [37]. In areas with extremely flat topography and especially near marine areas, the aquifers are particularly sensitive to saltwater intrusion [7]. The climate in Africa is characterized by random rainfall [34]. There are two levels of rainfall extremes, ranging from almost zero in dry regions such as the Sahara Desert, to very high rainfall in the Congo-Guinean rainforests [14, 37].

634

A. El Bakouri et al.

Fig. 2 Map of the slopes of Africa

4.2

Hydrogeology

The different geological reservoirs of Africa are located in a wide range of climatic zones: desert to semi-desert, arid, humid tropical, equatorial. The variability of aquifer recharge depends on climatic variability, as well as on geological formations, since a reservoir with matrix porosity is more productive than a fractured or fissured reservoir or an ancient consolidated sedimentary terrain [9, 21]. The origin of groundwater varies according to the climatic, geomorphological context and sedimentary processes, may originate from rainwater that is infiltrated and stored in pores or fissured of saturated rocks, or marine by a connection between aquifer water and sea water [12, 28] (See Fig. 3). The water balance equation can be written for a given period (see Eqs. (4) and (5)), according to Law of conservation of mass: Input þ Output ¼ Stock variation

ð4Þ

P  ET  R ¼ DS

ð5Þ

Water storage comes in different forms. We can distinguish three main types of reservoirs:

Multiple Water Reservoirs in African Continent…

635

Infiltration

R

Bedrock

Recharge

Fig. 3 Water balance. P: the water, due to condensation and aggregation processes inside the clouds, has become too heavy to remain suspended in the atmosphere, R: Runoff, the flow of water on the surface of the earth, especially the surface of the soil, ET: the amount of water transferred to the atmosphere by evaporation from sea, watercourse and from the interception according to which meteoric water is retained by tree leaves and plants, and by transpiration of water from plants

– Surface water stock, such as a natural lake or dams – Groundwater stock, a reservoir rock whose porosity allows the accumulation of water (water table, aquifer) – Water stock in solid form by snow and ice covers The groundwater recharge process depends in particular on highly concentrated rainfall events, accumulation of runoff water in depressions and streams and rapid percolation through fissures and fractures [8, 16]. In Table 2, we have the spatial distribution of the different aquifers, we got, Continuous media aquifers, composed of sedimentary formations of Mesozoic to Quaternary age, occupy 41.7% of the total area, then ancient sedimentary aquifers of Precambrian and Paleozoic ages, often assimilated to basement aquifers in fractured environments, share with the basement aquifers 41.5% of Africa’s surface, and at last complex aquifers of 16.8% of the surface, of which volcanic aquifers of 4% [4, 30]. Underground reservoirs have an underlay of geological formations with low storage capacity, are dependent on rainfall for recharge. Since is facilitated by the percolation of precipitation through the unsaturated zone leading to the replenishment of the saturated zone, so there is a significant variation in the African continent [29]. Figure 4 shows the spatial distribution according to the sub-region, as the majority of groundwater is concentrated in central Africa, including the southern Sahara, the eastern West African Shield and the western Great Rift Valley.

636

A. El Bakouri et al.

Table 2 Hydrogeological entities of Africa [4] Nature of formation

Type of environment

Quaternary sediment Paleogene-Neogene sedimentary Nubian sandstone formations Karoo type formation (Carboniferous to Jurassic)

Continuous media with matrix porosity or double porosity

Cretaceous carbonate formations Jurassic-Triassic formations Detrital/Carbonate to volcano-sedimentary formations (Neoproterozoic to Paleozoic and Precambrian) Sedimentary to volcano-sedimentary formations and associated volcano-plutonism (Precambrian) Plutonic and metamorphic complexes (Precambrian to Paleozoic) Plutonic massifs (Cambrian to Precambrian) Volcanic and volcano-plutonic massifs of the Phanerozoic

Complex structure with multiple superimposed reservoirs Complex structures locally karstic Fissured/fractured dominant

Fissured/fractured

Fissured/fractured dominant, but porous materials may be interspersed

Fig. 4 Groundwater resources in sub-regions of Africa (Data: [15])

5 Water Resources 5.1

Major River Basins

African rivers experience dramatic seasonal variability and interannual variations, which reflect the rainfall patterns in these basins. The Congo Basin is the second

Multiple Water Reservoirs in African Continent… 10°W

0°

10°E

20°E

30°E

637 40°E

50°E

30°N

30°N

N O RT H I N T E R I O R 20°N

20°N

NIGER

LAKE CHAD NILE

10°N

0°

10°N

0°

CONGO

10°S

10°S

ZAMBEZI 20°S

20°S

30°S

30°S

10°W

0°

10°E

1 000 500

0

20°E 1 000

30°E

40°E

2 000

3 000 Km

Legend Hydrological basins (%) 0.000- 0.005 0.005 - 0.010 0.010 - 0.020 0.020 - 0.030 0.030 - 0.040 0.040 - 0.050 0.050 - 0.070 0.070 - 0.080 0.080 - 0.090 0.090 - 0.100 0.010 - 0.110 0.110 - 0.120

50°E

Fig. 5 Major river basins of Africa

largest river basin in the world, after the Amazon (see Fig. 5). This is mainly due to the basin’s highly intense rainfall, with an mean flow of 1260  109 m3/s (see Fig. 6), the Congo River accounts for more than one-third of Africa’s freshwater resources and offers considerable potential [27, 37]. The Congo Basin covers about 1.2  106 km2 and is one of the largest intracratonic basins in the world. It is underlain by a thick lithosphere (200 ± 30 km) and coincides with a region of pronounced long-wave gravity anomaly [11, 19]. The Nile Basin is the third largest watershed in the world, after the Amazon and the Congo, and the second largest in Africa. The Nile Basin is the third largest river basin in the world, after the Amazon and Congo, and the second largest in Africa. It has two main tributaries: the Blue Nile, originating from Lake Tana in Ethiopia; and the White Nile, originating from Lake Victoria and the mountains of Burundi, Rwanda, and the Democratic Republic of Congo [38]. Large regional aquifer systems containing substantial amounts of groundwater underlie the Nile region. Some of the aquifers contain fossil water, but others are recharged by rainfall over the basin, or by irrigation areas and the Nile base flow. The Niger Basin is the fourth largest basin in Africa. Its river originates in well-watered regions before crossing the Sahelian zones. The right bank of the Ansongo-Niamey reach comprises three major sub-basins:

638

A. El Bakouri et al.

Fig. 6 Annual mean flow of major river systems in Africa (Data: [37])

– The Gorouol, the northern and largest, extends over Mali, Burkina Faso and Niger – The Dargol, shared between Niger and Burkina Faso – The Sirba, the southern one shared between Burkina Faso and Niger The left bank of the Niger River is characterized by smaller, largely endoreic watersheds with a limited contribution to the flow of the Niger. However, since the beginning of the century, outbursts of endoreism have been observed in the small basins of the left bank [13, 23, 25, 26]. The Zambezi is the fourth largest river in Africa and the largest river flowing into the Indian Ocean. It originates in the Kalene Hills. Tributaries flow along both banks, draining portions of eastern and southeastern Angola and northern Zambia into low-lying areas, which form the Barotseland floodplain. The Zambezi River is characterized by two large hydroelectric dams framing the large Kafue Flats floodplain: the Itezhi-Tezhi Dam upstream and the Upper Kafue Gorge Power Station downstream [40].

5.2

Inland Water Bodies

The amount of water supply depends on the availability and sustainability of the resource. Precipitation, surface runoff and groundwater recharge are intimately linked in the hydrological cycle. Therefore, a good knowledge of temperature and

Multiple Water Reservoirs in African Continent… 20°W

10°W

0°

10°E

20°E

639

30°E

40°E

50°E Baghdad IRAN

30°N

Cairo

MOROCCO ALGERIA

LIBYA

30°N

EGYPT

Duba

SAUDI ARABIA

N O R TH

EUROPE A SIA

A M E R IC A

20°N

20°N MALI

NIGER

SUDAN

A

CHAD

N IG E R I A

10°N

S OU TH

A FR IC A

A M E R IC A

A U S TR A LI

A N TA R C TIC A

10°N ETHIOPIA

Accra

0°

0° DR CONGO

Legend Inland water bodies

T A N Z A N IA

10°S

10°S ANGOLA ZA M B I A

20°S

20°S

NA MIBIA

Johannesburg

30°S

30°S

SOUTH AFRICA

Atlantic Ocean

20°W

10°W

0° 1 000 500

10°E 0

20°E 1 000

30°E 2 000

40°E

50°E

3 000 Km

Fig. 7 Inland waters of Africa

precipitation trends is very important for a better management of water resources in a basin, water demand and availability [2, 24]. Lakes and Dams. In Africa, the Great Lakes region is a system of lakes located in East Africa, bordered by four countries: Burundi, the Democratic Republic of Congo, Uganda and Rwanda (see Fig. 7). It is characterized by its ancient volcanic activity; this part of Africa is also one of the most fertile regions. Its altitude also gives it a rather temperate climate despite its equatorial location. In Central Africa, the Lake Chad Basin has seen an uninterrupted decline in its water level over the past few decades due to variables such as climate change and overexploitation, as well as some drought periods. Lake Chad is a large shallow lake in Africa with fresh water, which is rare for an endoreic lake. It is bounded by four countries: Chad, Cameroon, Niger and Nigeria [18]. South Africa and Zimbabwe have the largest dams, allowing rainwater harvesting. This is seen as a coping strategy in drought-prone environments [1]. The following dams can be mentioned: Akosombo Dam, Kariba Dam, Cahora Bassa Dam, Aswan High Dam, Al Wahda Dam, Katse Dam, Mohale Dam and the Renaissance Dam under construction. Lagoons. They are transitional water systems between land and ocean. Their ecosystem is particularly productive, they are shallow aquatic environments that generally present a wide variety of colonization by macrophytes [22].

640

A. El Bakouri et al.

In Africa, the coastal lagoons have a wide geographical distribution on the Mediterranean coast in the North, so in the West African coast, they are environments, where fresh water and salt water of the Atlantic Ocean mix. But the environments of the coastal zone can also be affected indirectly by human actions located inland [31].

5.3

Surface Water Bodies

Over the past decade, water supply and sanitation conditions have deteriorated in rural Africa and remained at a poor level in urban areas (see Fig. 8). Total water storage changes in basins that drain into the ocean closely follow the El Niño-Southern Oscillation cycle, and it is likely that some endoreic basins will be similarly affected [32]. Major Rivers. Africa is characterized by a diversified hydrographic network, we can quote the main rivers and we will start with the Congo River which crosses the countries of the Democratic Republic of Congo and the Republic of Congo, with a length of 4700 km, it is the eighth longest river in the world but the second after the Amazon for its mean flow of 41,800 m3/s at the mouth. The Niger is a West African river, with a flow of 6000 m3/s. However, a closer look at the time series reveals the presence of abrupt changes in these records, similar to those previously detected in some West African rivers, notably the Senegal and Niger Basins [36, 39].

Fig. 8 Total water resources in sub-regions of Africa (Data: [37])

Multiple Water Reservoirs in African Continent…

641

The Nile, with a length of 6650 km, by the origin of his waters, its source is the Blue Nile is located in Ethiopia, while another arm of water, the White Nile, finds its origin in Rwanda. In Egypt, the Nile where most of the water supply comes from the transboundary river flow, in Sudan clearly appears to be highly dependent on upstream flows [10, 41]. The Zambezi is a southern African river, originating in the Black Swamp in northwestern Zambia, because the amplitude of seasonal water level changes is strongly buffered and controlled by the Kariba Dam, the Zambezi downstream does not cause such strong seasonal backwater effects [40]. Wetlands. Wetlands exist in all climate zones, from the polar regions to the tropics. Africa’s wetland ecosystems are estimated at over 131 million hectares. They are found where rivers such as the Congo, Zambezi, Nile, Niger, Senegal flow into the ocean (see Fig. 9). They are generally a unique combination of physical characteristics associated with their shape, watershed, connection to the sea and tidal regime [6, 20]. We can cite several wetlands in Africa, we have: Inner Niger Delta, is the largest delta in West Africa, it is a natural region of Mali extending over an area of 40,000 km2, Okavango Delta, is the second largest inland delta in Africa, with an area of 18,000 km2, the Hadejia-Nguru wetlands, concern a part of the flood plain of the 20°W

10°W

0°

10°E

20°E

30°E

40°E

50°E

TURKEY

Baghdad IRAN

30°N

Cairo

MOROCCO ALGERIA

LIBYA

30°N

EGYPT

Dub

SAUDI ARABIA

N O R TH

E UROP E

A M E R IC A

20°N

20°N MALI

NIGER

A M E R IC A

SUDAN

A S IA

A FR IC A

S OU TH

AUS TRA L

A

CHAD

A N TA R C TIC A

N I GE R I A

10°N

10°N ETHIOPIA

Accra

0°

0° DR CONGO

10°S

10°S ANGOLA ZAMBIA

20°S

20°S

NA M IB IA

Johannesburg

30°S

30°S

SOUTH AFRICA

Atlantic Ocean

20°W

10°W

0° 1 000 500

10°E 0

Fig. 9 Surface waters of Africa

20°E 1 000

30°E 2 000

40°E 3 000 Km

Legend Surface water bodies

T A N ZA N I A

50°E

642

A. El Bakouri et al.

Komadougou-Yobe river basin in the Lake Chad basin in the North-East of Nigeria, we can also mention Sidi Boughaba, it is a freshwater swampy wetland of the North-West coast of Morocco.

6 Conclusion The African water reservoirs, whether underground or surface, constitute a resource to be protected and not wasted. With the exception of arid and semi-arid areas, Africa is very rich in water. The optimal development of this resource must be based on the management of watersheds. The first step is to gain a better understanding of groundwater and surface water resources, by having quantitative and qualitative data on these resources, in order to better use water to support socio-economic development. Indeed, by carrying out projects for drinking water supply, liquid sanitation, irrigation, as well as energy. The results of this study can inform that water resources must hold groundwater management, especially aquifers and fossil water deposits that have a high risk of overexploitation and consumption and require consideration. For an adequate supply of fresh water for the population, dams are considered an essential tool to supplement dry periods, capturing ephemeral river flows. Another challenge is to anticipate and preserve water resources and associated ecosystems, and to fight against the quantitative degradation of water resources resulting from various factors, such as changes in the hydrological regime or pollution. The participation of the populations through the establishment of participatory management mechanisms constitutes a better management and sustainability of water and the environment.

References 1. G.M. Bombelli et al., Impact of Prospective Climate Change Scenarios upon Hydropower Potential of Ethiopia in GERD and GIBE Dams. Water. 13(5), 716 (2021). https://doi.org/10. 3390/w13050716 2. A.V. Borges et al., Globally significant greenhouse-gas emissions from African inland waters. Nat. Geosci. 8(8), 637–642 (2015). https://doi.org/10.1038/ngeo2486 3. J. Brendryen et al., Eurasian Ice Sheet collapse was a major source of Meltwater Pulse 1A 14,600 years ago. Nat. Geosci. 13(5), 363–368 (2020). https://doi.org/10.1038/s41561-0200567-4 4. BRGM: Une carte hydrogéologique de l’Afrique | BRGM, https://www.brgm.fr/fr/referenceprojet-acheve/carte-hydrogeologique-afrique. Accessed 3 Aug 2021 5. S. Brunel, L’Afrique: un continent en réserve de développement. Editions Bréal (2003) 6. A.D. Canning et al., Financial incentives for large-scale wetland restoration: beyond markets to common asset trusts. One Earth. 4(7), 937–950 (2021). https://doi.org/10.1016/j.oneear. 2021.06.006

Multiple Water Reservoirs in African Continent…

643

7. B. Caputo et al., Anopheles gambiae complex along The Gambia river, with particular reference to the molecular forms of an. Malar. J. 7(1), 182 (2008). https://doi.org/10.1186/ 1475-2875-7-182, gambiae s.s. 8. A. Carletti et al., A combined methodology for estimating the potential natural aquifer recharge in an arid environment. Hydrol. Sci. J. 64(14), 1727–1745 (2019). https://doi.org/10. 1080/02626667.2019.1662422 9. J. Casanova et al., Recharge artificielle des eaux souterraines : état de l’art et perspectives. BRGM (2013) 10. D. Conway, From headwater tributaries to international river: observing and adapting to climate variability and change in the Nile basin. Glob. Environ. Chang. 15(2), 99–114 (2005). https://doi.org/10.1016/j.gloenvcha.2005.01.003 11. A.G. Crosby et al., Structure and evolution of the intracratonic Congo Basin. Geochem. Geophys. Geosystems. 11(6) (2010). https://doi.org/10.1029/2009GC003014. 12. D. Defrance et al., Consequences of rapid ice sheet melting on the Sahelian population vulnerability. PNAS 114(25), 6533–6538 (2017). https://doi.org/10.1073/pnas.1619358114 13. L. Descroix et al., Spatio-temporal variability of hydrological regimes around the boundaries between Sahelian and Sudanian areas of West Africa: a synthesis. J. Hydrol. 375(1), 90–102 (2009). https://doi.org/10.1016/j.jhydrol.2008.12.012 14. P.R. Elsen et al., Global patterns of protection of elevational gradients in mountain ranges. PNAS 115(23), 6004–6009 (2018). https://doi.org/10.1073/pnas.1720141115 15. FAO: The state of the world’s land and water resources for food and agriculture: Managing systems at risk. FAO, Earthscan from Routledge, London, England (2011) 16. C.M. Fraser et al., A methodology to identify vulnerable transboundary aquifer hotspots for multi-scale groundwater management. Water Int. 45(7–8), 865–883 (2020). https://doi.org/10. 1080/02508060.2020.1832747 17. R.W. Healy, Estimating Groundwater Recharge. Cambridge University Press, Cambridge (2010). https://doi.org/10.1017/CBO9780511780745 18. A. Hussaini et al., Change detection for the past three decades using geospatial approach in Lake Chad, Central Africa. IOP Conf. Ser. Earth Environ. Sci. 540, 012001 (2020). https:// doi.org/10.1088/1755-1315/540/1/012001 19. E. Kadima et al., Structure and geological history of the Congo Basin: an integrated interpretation of gravity, magnetic and reflection seismic data. Basin Res. 23(5), 499–527 (2011). https://doi.org/10.1111/j.1365-2117.2011.00500.x 20. A.-H.A. Khedr, Vegetation zonation and management in the Damietta estuary of the River Nile. J. Coast Conserv. 4(1), 79–86 (1998). https://doi.org/10.1007/BF02806493 21. M. Konaté et al., Structural evolution of the Téfidet trough (East Aïr, Niger) in relation with the West African Cretaceous and Paleogene rifting and compression episodes. C.R. Geosci. 351(5), 355–365 (2019). https://doi.org/10.1016/j.crte.2018.11.009 22. Y.J.-M. Koné et al., The role of macrophytes in biogenic silica storage in ivory coast lagoons1. Front. Earth Sci. 7, 248 (2019). https://doi.org/10.3389/feart.2019.00248 23. M.J. Leblanc et al., Land clearance and hydrological change in the Sahel: SW Niger. Glob. Planet. Chang. 61(3), 135–150 (2008). https://doi.org/10.1016/j.gloplacha.2007.08.011 24. R. Mahmood et al., Analysis of climate variability, trends, and prediction in the most active parts of the Lake Chad basin, Africa. Sci. Rep. 9(1), 6317 (2019). https://doi.org/10.1038/ s41598-019-42811-9 25. I. Mamadou et al., Exorheism growth as an explanation of increasing flooding in the Sahel. CATENA 131, 130–139 (2015). https://doi.org/10.1016/j.catena.2015.03.017 26. G. Massazza et al., Recent changes in hydroclimatic patterns over medium Niger River basins at the origin of the 2020 flood in Niamey (Niger). Water. 13(12), 1659 (2021). https://doi.org/ 10.3390/w13121659 27. B. Meinier, Transboundary Water Management in the Congo Basin. Creative Republic, Thomas Maxeiner Kommunikationsdesi, Deutsche Gesellschaft für Internationale Zusammenarbeit (2014)

644

A. El Bakouri et al.

28. A.-S. Mohamed et al., Impacts of climate change and anthropization on groundwater resources in the Nouakchott urban area (coastal Mauritania). C.R. Geosci. 349(6), 280–289 (2017). https://doi.org/10.1016/j.crte.2017.09.011 29. S. Oiro et al., Using stable water isotopes to identify spatio-temporal controls on groundwater recharge in two contrasting East African aquifer systems. Hydrol. Sci. J. 63(6), 862–877 (2018). https://doi.org/10.1080/02626667.2018.1459625 30. F.E. Oussou et al., Fracture aquifers identification in the Zou basin (West Africa) using remote sensing and GIS Geocarto Int. 1–24 (2020). https://doi.org/10.1080/10106049.2020.1852613 31. J. Patrício et al., DPSIR—two decades of trying to develop a unifying framework for marine environmental management? Front. Mar. Sci. 3, 177 (2016). https://doi.org/10.3389/fmars. 2016.00177 32. T.M. Pavelsky, World’s landlocked basins drying. Nat. Geosci. 11(12), 892–893 (2018). https://doi.org/10.1038/s41561-018-0269-3 33. L.J. Robbins et al., Hydrogeological constraints on the formation of Palaeoproterozoic banded iron formations. Nat. Geosci. 12(7), 558–563 (2019). https://doi.org/10.1038/s41561-0190372-0 34. S. Sarkar, R. Maity, Global climate shift in 1970s causes a significant worldwide increase in precipitation extremes. Sci. Rep. 11(1), 11574 (2021). https://doi.org/10.1038/s41598-02190854-8 35. S.P. Simonovic, Systems approach to management of water resources—toward performance based water resources engineering. Water. 12(4), 1208 (2020). https://doi.org/10.3390/ w12041208 36. Y. Tramblay et al., Observed changes in flood hazard in Africa. Environ. Res. Lett. 15(10), 1040b5 (2020). https://doi.org/10.1088/1748-9326/abb90b 37. UNEP: Africa: Mountains Atlas. UNEP, AMCEM, United Kingdom (2014) 38. K.G. Wheeler et al., Exploring cooperative transboundary river management strategies for the eastern Nile Basin. Water Resour. Res. 54(11), 9224–9254 (2018). https://doi.org/10.1029/ 2017WR022149 39. C. Wilcox et al., Trends in hydrological extremes in the Senegal and Niger Rivers. J. Hydrol. 566, 531–545 (2018). https://doi.org/10.1016/j.jhydrol.2018.07.063 40. R.S. Winton et al., Potential of aquatic weeds to improve water quality in natural waterways of the Zambezi catchment. Sci. Rep. 10(1), 15467 (2020). https://doi.org/10.1038/s41598020-72499-1 41. M. Zeitoun, N. Mirumachi, Transboundary water interaction i: reconsidering conflict and cooperation. Int. Environ. Agreements 8(4), 297 (2008). https://doi.org/10.1007/s10784-0089083-5

Seasonal Dynamics of Sandflies and Soil Texture of Breeding Sites, Aichoune Locality, Sefrou Province, Morocco Fatima Zahra Talbi, Mohamed Najy, Hajar El Omari, Abdelkarim Taam, and Abdelhakim El Ouali Lalami

Abstract In Morocco, leishmaniases constitute a real public health problem. Sandflies are the only known vectors of these parasitoses. In order to study the dynamics of sandflies using two different traps and to identify the nature of the soil of potential sandfly breeding sites at Aichoune, monitoring the activity of sandflies was carried out over a period from September 2013 to August 2014. A total of 4471 sandflies were identified. Phlebotomus sergenti, vector of human cutaneous leishmaniasis, is the most abundant followed by Larroussius perniciosus. The maximum species was harvested in the months of June and August. This study shows that the two high-risk Leishmanian months are in June and August, hence the need to strengthen efforts to fight against this disease during these periods. The results allowed us to identify a sandy nature of the larval sites for sandflies. The characterization of the substrate at the level of these larval sites has informed us about the ecological requirements of the larval development of the species, hence the need for a substrate rich in organic matter. These results could help the authorities to prevent from the risk of leishmaniasis. Indeed, medium-term climate forecasts are essential tools for developing a leishmaniasis warning system.

F. Z. Talbi (&) Hassan First University of Settat, Faculty of Sciences and Technologies, Laboratory of Biochemistry, Neurosciences, Natural Resources and Environment, 577, Settat, Morocco e-mail: [email protected] M. Najy Natural Resources and Sustainable Development Laboratory, Faculty of Sciences, Ibn Tofail University, BP133, 14000 Kenitra, Morocco H. E. Omari Natural Resources Management and Development Team, Laboratory of Health and Environment, Faculty of Sciences, Moulay Ismail University, Meknes, Morocco A. Taam Laboratory of Engineering Sciences, National School of Applied Sciences (ENSA), Ibn Tofail University, Kenitra, Morocco A. E. O. Lalami Higher Institute of Nursing Professions and Health Techniques of Fez, Regional Health Directorate Fez-Meknes, EL Ghassani Hospital, 30000 Fez, Morocco © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_52

645

646

F. Z. Talbi et al.
















Keywords Sand flies Texture soil Seasonality Dynamic Aichoune Sefrou Province Morocco




1 Introduction Sandflies are insects of medical interest that transmit the pathogens responsible for leishmaniasis through hematophagous females [1]. Phlebotomus species are known for their role as vectors of medically important pathogens, such as the parasitic protozoa of the genus Leishmania [2]. The most important of these diseases is leishmaniasis, both visceral leishmaniasis (VL) and cutaneous leishmaniasis (CL) constitute growing public health problems. Leishmaniasis has a wide worldwide distribution with more than a billion people at risk of infection in 98 countries [3]. Several risk factors influence the distribution of leishmaniasis cases [4]. Among more than 800 species of sandflies recorded, 98 are suspected vectors of human leishmaniasis; 42 species of Phlebotomus in the Old World and 56 species of Lutzomyia in the New World (Diptera: Psychodidae) [5]. In Morocco, 22 sandfly species [6], distributed in 13 species of the genus Phlebotomus and 09 species of Sergentomyia. Only species of the genus Phlebotomus are of medical importance, they ensure the transmission of leishmaniasis which is a real public health problem in this country [7] Sandflies are present in temperate regions during summer [8]. Adult sandflies of both sexes can be collected by several methods, during this research we used two capture methods the adhesive traps and the light traps. In order to study the seasonality of the different species of sandflies, we considered it interesting to study the seasonal distribution and the behavior of sandflies captured by the different capture methods cited as well as the relationship with the nature of the soil.

2 Materials and Methods 2.1

Study Area

The collections were carried out in Aichoune, a small town located in Sefrou Province, in the northwest of the Moroccan Middle Atlas (33°39’N, 04°38’W), Fez-Meknes region (Fig. 1).

Seasonal Dynamics of Sandflies and Soil Texture of Breeding Sites

647

Fig. 1 Presentation of Province Sefrou, Commune of Tazouta (Aichoune locality)

2.2

Specimen Collection

Collections were performed by using sticky papers (21  27, 3 cm) coated with castor oil in four stations. The sticky traps were used independently to monitor populations of sand flies. They were placed in different spots and covered an area of approximately 1 m2. Each trap was placed horizontally on a square metal frame support about 20 cm above the ground or it is shaped rolled up in the holes or between the crevices of the walls. Only the upper sides of the sticky traps were coated with castor oil because previous studies had shown that practically no fly adhered to the other side of the sticky traps [9]. The total of traps in each trapping campaign is 32 traps. The compound was sampled during two nights each month with 64 papers placed in four stations for a 12 h dusk to-dawn period. Combined morning sand fly counts for the period yielded a density estimate, the mean number of sand flies/paper per night. The species also coached by CDC miniature light-traps, set out from sunset (from 6 to 7 pm) to sunrise (from 6 to 8 am) in five stations. During each month of survey, ten CDC light traps, placed at low level (0.5–0.7 m), were operated for 12 h dusk-to-dawn periods (Fig. 2). Sample collection began in early September 2013 and continued until late August 2014. Trapped sand flies removed from stick papers with needles, washed with ethanol, and transferred in glass tubes containing a solution of 70° ethanol. After sex determination, all sand flies collected were identified by examining the morphology of the pharyngeal armature and spermathecae of female flies and the external genitalia of males using the morphological key [10–12]. Morphological differentiation of the two sympatric species Ph. longicuspis and Ph. perniciosus was made according to description of Berchi et al. [13].

648

F. Z. Talbi et al.

Fig. 2 Examples of sampling methods for phlebotomines sand flies (Sticky traps in the left and CDC traps in the right)

2.3

Processing of Soil Samples

Soil samples were collected from damp spots along the perimeters of cattle sheds of Aichoune locality in 4 stations each month. The soil samples were collected from the same stations containing the traps. the soil analysis was followed according to the proposed protocol of Talbi Fatima Zahra’s work [9] And Boussa’s work [14].

2.4

Statistical Analysis

The data were recorded in an Excel spreadsheet (Excel 2010) and then presented in graphical form. To study the results statistically, we chose to use The correlation coefficient gives us information on the existence of a linear relationship between the two quantities considered. We have calculated the correlation index by the following formula: P ðX  XÞ:ðY  YÞ r ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi P P 2 ðX  XÞ  ðY  YÞ2 A p-value less than 0.05 was considered significant.

Seasonal Dynamics of Sandflies and Soil Texture of Breeding Sites

649

3 Results 3.1

Seasonnal Fluctuation of Species Sand Flies with Both of Catch Method

The entomological survey that we have conducted in Aichounee locality allowed to capture 4471 sandflies divided into two genus: Phlebotomus with three subgenus (Phlebotomus, Paraphlebotomus and Larroussius) and Sergentomyia with one subgenus (Sergentomyia). In the locality of Aichoune, no sand fly was found in the houses during the winter season. Five of phlebotomines sandflies were continuously present during the season. Conversely, Ph. ariasi, S. minuta and S. fallalx were rare throughout the studied period. There was a difference in terms of density peaks between trap types. In other seasons, catches using sticky paper, 82% (=1508) of the flies caught were Ph. sergenti. They presented with a percentage of 70% of sand flies in May, June, July and August. While in September and October, Ph. sergenti were present by only 12.5% (Fig. 3a). Ph. perniciosus were collected with only 9.22% (=169) of the total species caught. Our data show that Ph. sergenti and Ph. perniciosus had a long activity season from May to October, with one population peak at the month of June. Catches using light traps, 58.29% (= 578) of the flies caught were Ph. perniciosus. This time, the maximum was collected in August with 14.96% followed by 11.28% in June (Fig. 3b).

3.2

Subtract Nature of Different Station’s Traps

The stations containing the adhesive traps are characterized by a high rate of moisture and organic matter (probably due to its high content of animal excrement and plant debris (nutrients)). The comparison of the granulometry of the two substrates shows that the soils are of sandy nature for the types of the stations. Comparing with the abundance of each species of sand flies, we have found that Ph. sergenti prefers a substrate rich in organic matter and of a sandy nature. For Ph. perniciosus, its high abundance in CDC light traps sites may be related, in addition to the sandy nature of the substrate, to the use of this type of traps (Table 1).

650

F. Z. Talbi et al.

(a) Number of sandflies captured by Sticky traps

600 500 400 300 200 100 0

(b) Number of sand flies captured of CDC light traps

Ph. sergenti Ph. perniciosus Ph. longicuspis Ph. papatasi Ph. ariasi S. minuta S. fallax

450 400 350 300

Ph. sergenti Ph. perniciosus Ph. longicuspis Ph. papatasi Ph. ariasi S. minuta S. fallax

250 200 150 100 50 0

Fig. 3 a. Seasonal fluctuation with Number of sand flies captured by Sticky traps between September 2013 to August 2014 b. Seasonal fluctuation with Number of sand flies captured by CDC light traps between September 2013 to August 2014

3.3

Statistical Analysis

Statistical analysis showed a significant correlation between the variation of the seasons and the number of sandflies captured with p less than 0.05. In fact the number of sandflies captured during the dry season is greater than that captured in

Seasonal Dynamics of Sandflies and Soil Texture of Breeding Sites

651

Table 1. Relation between characterization of the substrates and the use of two types of traps in different stations Stations with

Total Sand flies

Moisture (%)

Organic mater (%)

Soil texture Clay Loam (%) (%) Gross Fin Loam Loam

Sand (%) Gross Sable Sand fin

Sticky traps Total CDC light traps Total

1831

28.7

13.17

18.64

8.1

11.4

10.01

54.63

2640

6.4

6.6

16.49

19.5 9.23

12.15

64.64 23.68

46.32

21.38

70

the wet season. Statistical analysis showed a significant correlation between the type of trap used and the capture season (p. value  0.05).

4 Discussion For this study, two different types of traps were used because they provide additional information on sandfly populations. In the present study, five Phlebotomus and two Sergentomyia species were identified in the Aichoune locality. The genus Phlebotomus which groups together the mammalian species vectors of Leishmania, [15] and the genus Sergentomyia which feeds on reptiles, amphibians, and birds and rarely bites the man [16, 17]. These results are consistent with the research of Lahouiti, in the region of Moulay Yakoub in Fez (Morocco) [18] and Hajar El Omari’s Work [19] which showed the dominance of the genus Phlebotomus by comparison to the genus Sergentomyia. A total of seven species have been identified. Five species are of medical interest, and implicated in the transmission of leishmaniases in Morocco [20]. These species are Ph. sergenti which is widely distributed in Aichoune foci, Ph. perniciosus, Ph. longicuspis, Ph. papatasi, Ph. ariasi, Sergentomyia minuta, and Sergentomyia fallax. These species are considered the most abundant among the 23 species recorded in Morocco [10]. The study of the activity of sand flies in study area was followed from September 2013 to August 2014. According to the capture sessions conducted in Aichoune, we found a monthly evolution that varies from one species to another, biphasic in some species and monophasic in others. Throughout the study period, the activity of sand flies was marked by the dominance of Ph. sergenti, Ph. perniciosus and Ph. longicuspis with two peaks in June and August. Ph. sergenti is confirmed as a vector of L. tropica anthroponotic leishmaniasis in Northern Africa, the Middle East, and Central Asia [21, 22]. The variation in density of phlebotomine fauna during the study period showed that the period of activity of sandflies extends between the

652

F. Z. Talbi et al.

months of May and October which corresponds to the dry season where there is no rain. These results confirm the data obtained by Galvez in Spain [23]. The high densities between the month of June and the month of August. That could be explained by the weather disturbances in the area. Ph. sergenti was the only species present in all the studied stations with important values; this could be explained by the best adaptation of this species to the ecological and climatic conditions of study area. This vector prefers a semiarid bioclimate, even though it was captured in all bioclimatic zones and was widespread throughout Morocco [24]. These results are in agreement with those obtained by Guessouss-Idrissi et al. [25] in the Province of Taza, and El Miri et al. [26] in the Province of Sidi Kacem in northern Morocco. The evolution of the monthly specific density of the insects collected showed that Ph. sergenti, vector of L. tropica cutaneous leishmaniasis in Morocco [27], was the most dominant in our study. The period of its activity extended from May to October with two peaks. Our result confirms the work of El Aasri in Sidi Yahya [28] who showed that the annual cycle of Ph. sergenti is of biphasic type, while the work of Boussa in the Marrakesh region [29] has revealed that the cycle of this insect is monophasic. Ph. perniciosus, one of the most competent L. infantum vectors in the Mediterranean foci [30]. Ph. perniciosus is considered one of the most important species. The analysis of the seasonal dynamics of populations of Ph. perniciosus revealed that the period of activity extended from May to October with a biphasic evolution with two peaks. These results confirm those found in the Chichaoua region [31]. Ph. longicuspis was collected in Aichoune locality. The highest density of this species was observed in August. This result is in accordance with Guernaoui et al. [32] in Chichaoua Province where this species showed a monophasic cycle, with one density peak in August. The other species that remain represent a minimal percentage. The seasonal dynamics of sandflies is due to several factors, not only climatic but also environmental [33] which favor the multiplication of insects. Many authors have emphasized the role of soil conditions in the spatial distribution of sandflies and also for characterization of sandfly reproduction sites [34, 35]. The study in Israel by Schlein et al. [36] revealed that the scarcity of water was the limiting factor in the abundance of sandflies and the spread of leishmaniasis. However, soil moisture remains important for maintaining sandfly populations in all habitats. According to Boussa and Boumezzough in Morocco, Ph. sergenti requires a high rate of organic matter for larval development [14]. In Southern Anatolia, the increased cases of anthroponotic cutaneous leishmaniasis in is due especially to the presence of organic matter of animal origin, as the dung of cows, offering the conditions necessary for the development of eggs of sand flies [37]. The rate of organic matter is high in the various stations of the locality of Aichoune, probably due to its richness in vegetation or animal waste. In 1935, Parrot reported the ease with which sandfly larvae can be reared by nourishing with plant debris [38]. The study of the particle size distribution of the substrate shows, in the soil of Aichoune, a richness in Sands (70%) with a 21.38% Loam. These preliminary results at the level of the larval sites, inform us about the ecological requirements of each of the phlebotomian species. As for soil texture, sandflies prefer sandy soil. This result is consistent with that of Israel where

Seasonal Dynamics of Sandflies and Soil Texture of Breeding Sites

653

analysis of the soil of a sandfly larval nest (Ph. sergenti) also showed a richness in organic matter with a sandy texture [35]. According to WHO (2002) [37], the increase in cases of anthroponotic cutaneous leishmaniasis in southern Anatolia is due mainly to the presence of organic matter of animal origin, such as cow dung, providing conditions necessary for the development of sandfly eggs.

5 Conclusion Seasonal fluctuations of sandflies in the Aichoune are characterized by a evolution for Ph. sergenti, Ph. perniciosus, Ph. longicuspis and Ph. papatasi. The first peak is often observed in Jun and the second is in August. These results reveal two high-risk leishmaniosis annual periods in Aichoune locality marked by a high total density of sandflies and coincides with the hot and dry period. The large distribution and the long activity period of Ph. sergenti and species of the subgenus Larroussius in Aichoune locality indicate the high potential risk of L. tropica and L. infantum transmission in this area. Acknowledgements We are grateful to National Institute of Hygiene. The authors thank the Health regional directorate of the Fes-Boulemane, provincial delegate of Ministry of Health, Sefrou province, and the staff of Aichoune locality for their cooperation, assistance, information, and help. Conflicts of Interest The authors declare that they have no conflicts of interest.

References 1. R. Killick-Kendrick, Phlebotomine vectors of the leishmaniases: a review. Med. Vet. Entomol. 4, 1–24 (1990) 2. A.V. Dolmatova, N.A. Demina, Les phlébotomes (Phlebotominae) et les maladies qu’ils transmettent, p. 18 (1966) 3. P. Desjeux, The increase in risk factors for leishmaniasis worldwide. Trans. R Soc. Trop. Med. Hyg. 95, 239–243 (2001) 4. F.Z. Talbi, M. Najy, N. Nouayti, H. En-nkhili, A. El Ouali Lalami, Cutaneous Leishmaniasis cases and risk factors in north central of Morocco, sefrou province: an impact study. E3S Web Conf. 234, 00027 (2021). https://doi.org/10.1051/e3sconf/202123400027 5. WHO, 2010: Control of the leishmaniasis: report of a meeting of the WHO Expert Committee on the Control of Leishmaniases. Geneva (2010) 6. H. Bailly-Chaumara, E. Abonnec, J. Pastre, Contribution à l’étude des phlébotomes du Maroc (Diptera: Psychodidae). Données faunistiques et écologiques. Cahier ORSTOM, série Entomologie Médicale Parasitol, IX 4, 431–460 (1971) 7. A. El Aasri et al., Sand flies of Morocco: biodiversity of the phlebotomian fauna of Had Kourt region (Province of Sidi Kacem, Morocco). J. Pharm. Chem. Biol. Sci. 3(2), 310–315 (2015) 8. E. Abonnenc, Les phlébotomes de la région éthiopienne (Diptera, Psychodidae). Paris ORSTOM (55), 289 (1972). multigr. (Mémoires ORSTOM ; 55), 1–290

654

F. Z. Talbi et al.

9. F.Z. Talbi, A. Janati Idrissi, M. Fadil, A. El Ouali Lalami, Soil analysis of potential breeding sites of sand flies (Diptera: Psychodidae) in Aichoune locality, central Morocco. Bull. Soc. Pathol. Exotique 1–8 (2020). https://doi.org/10.3166/bspe-202010. Ministry of Health (MS): Fight against Leishmaniasis, Activity Guide, Ministry of Health (MS), Rabat, Morocco (2010) 11. F.Z. Talbi, A. El Ouali Lalami, M. Fadil, M. Najy, H. Ech-Chafay, M. Lachhab, S. Lotfi, N. Nouayti, K. Lahouiti, C. Faraj, A. Janati Idrissi, Entomological investigations, seasonal fluctuations and impact of bioclimate factors of phlebotomines sand flies (diptera: psychodidae) of an emerging focus of cutaneous leishmaniasis in aichoune, Central Morocco. J. Parasitol. Res. 2020(6495108), 10 (2020). https://doi.org/10.1155/2020/ 6495108 12. F.Z. Talbi, C. Faraj, F. EL-Akhal, F. El Khayyat, D. Chenfour, A. Janati Idrissi, A. El Ouali Lalami, Diversity and dynamics of sand flies (diptera: psychodidae) of two cutaneous leishmaniasis foci in the fes-boulemane region of Northern Morocco. Int. J. Zool. 2015 (497169),6 (2015). https://doi.org/10.1155/2015/497169. 13. S. Berchi, A. Bounamous, K. Louadi, B. Pesson, Morphological distinction between two sympatric species: Phlebotomus pernlciosus and Phlebotomus longicuspis (Diptera: Psychodidae). Ann. Soc. Entomol. Fr. 43(2), 201–203 (2007) 14. S. Boussaa, A. Boumezzough, Identification et caractérisation des gîtes larvaires de phlébotomes (Diptera: Psychodidae) à Marrakech (Maroc). Entomol. Faunistique Faunist. Entomol. 67, 93–101 (2014) 15. E. Guilvard, J.A. Rioux, M. Gallego et al., Leishmania tropica in Morocco. III—the vector of Phlebotomus sergenti. Apropos of 89 isolates. Ann. Parasitol. Hum. Compare 66(3), 96–99 (1991) 16. C. Moulinier, Parasitologie et mycologie médicale. Éléments de morphologie et de biologie. Ed. Méd. Int. Lavoisier 796, 28 (2002) 17. P.D. Ready, Biology of phlebotomine sand flies as vectors of disease agents. Annu. Rev. Entomol. 58, 227–250 (2013). https://doi.org/10.1146/annurev-ento-120811-153557 18. K. Lahouiti, K. Bekhti, M. Fadil et al., Entomological investigations in moulay Yaacoub, leishmaniasis focus in the center of Morocco. Asian J. Pharm. Clin. Res. 9, 340–345 (2016). https://doi.org/10.22159/ajpcr.2016.v9i6.14783 19. H. El Omari, A. Chahlaoui, K. Ouarrak, C. Faraj, A. El Ouali Lalami, Surveillance of Leishmaniasis: Inventory and Seasonal Fluctuation of Phlebotomine Sandflies (Diptera: Psychodidae), at the Prefecture of Meknes (Center of Morocco). Bull. Soc. Pathol. Exot. 111, 309–315 (2018). https://doi.org/10.3166/bspe-2019-0061 20. J.A. Rioux, F. Petter, O. Akalay et al., Meriones shawi (Duvernoy, 1842) (Rodentia, Gerbillidae), reservoir of Leishmania major, Yakimoff and Schokhor, 1914 in South Morocco. Rendered Acc. Meet. Acad. Sci. Ser. III Life Sci. 294, 515–517 (1982) 21. Ministry of Health: Fight against Leishmaniasis, Activity Guide, Ministry of Health, Rabat, Morocco (2010) 22. K. Mondragon-Shem, W.S. Al-Salem, L. Kelly-Hope et al., Severity of old world cutaneous leishmaniasis is influenced by previous exposure to sandfly bites in Saudi Arabia. PLoS Negl. Trop. Dis. 9(2), 1–14 (2015) 23. A. Tabbabi, N. Bousslimi, A. Rhim, K. Aoun, A. Bouratbine, Short report: first report on natural infection of Phlebotomus sergenti with Leishmania promastigotes in the cutaneous leishmaniasis focus in southeastern Tunisia. Am. J. Trop. Med. Hyg. 85(4), 646–647 (2011) 24. R. Gálvez, M.A. Descalzo, G. Miró, et al., Seasonal trends and spatial relations between environmental/meteorological factors and leishmaniosis sand fly vector abundances in Central Spain. Acta Trop. 115, 95–102 (2010). https://doi.org/10.1016/j.actatropica.2010.02.009. Epub 2010 Feb 18 25. S. Boussaa, M. Neffa, B. Pesson, A. Boumezzough, Phlebotomine sand flies (Diptera: Psychodidae) of Southern Morocco: results of entomological surveys along the Marrakech-Ouarzazat and Marrakech-Azilal roads. Ann. Trop. Med. Parasitol. 104(2), 163– 170 (2010)

Seasonal Dynamics of Sandflies and Soil Texture of Breeding Sites

655

26. N. Guessouss-Idrissi, B. Berrag, M. Riyad, H. Sahibi, M. Bichichi, A. Rhalem, Leishmania tropica: etiologie agent of a case of visceralizing canine leishmaniasis in north Morocco. Am. J. Trop. Med. Hyg. 57(2), 172–173(1997) 27. H. El Miri, M. Rhajaoui, O. Himmi, S.Ouahabi, A. Benhoussa, C. Faraj, Entomological study of five cutaneous leishmaniasis foci in the Sidi Kacem province, north Morocco. Annales de la Société entomologique de France 49(2), 154–159(2013) 28. M. Ajaoud, N. Es-sette, S. Hamdi et al., Detection and molecular typing of Leishmania tropica from Phlebotomus sergenti and lesions of cutaneous leishmaniasis in an emerging focus of Morocco. Parasit. Vectors 6, 217 (2013). https://doi.org/10.1186/1756-3305-6-217 29. A. El Aasri, D. Belghyti, M. Hadji, K. EL kharrim, Étude de risque de la leishmaniose cutanée et viscérale dans la région de Sidi yahia du Gharb (province de Kenitra, Maroc). Sci. Lib Editions Mersenne 5, 13 (2013) 30. S. Boussaa, S. Guernaoui, B. Pesson, A. Boumezzough, Seasonal fluctuations of phlebotomine sand fly populations (Diptera: Psychodidae) in the urban area of Marrakech Morocco. Acta Trop. 95, 86–91 (2005) 31. R. Benikhlef, Z. Harrat, M. Toudjine, A. Djerbouh, S. Bendali Braham, M. Belkaid, Detection of Leishmania infantum MON-24 in the dog. Medecine Tropicale: Revue du Corps de Sante Colonial 64(4), 381–383 (2004) 32. S. Guernaoui, A. Boumezzough, A. Laamrani, Altitudinal structuring of sand flies (Diptera: Psychodidae) in the HighAtlas mountains (Morocco) and its relation to the risk of leishmaniasis transmission. Acta Trop. 97, 346–351 (2006) 33. S. Guernaoui, A. Boumezzough, B. Pesson, G. Pichon, Entomological investigations in Chichaoua: an emerging epidemic focus of cutaneous leishmaniasis in Morocco. J. Med. Entomol. 42(4), 697–701 (2005) 34. N. El Bada, M. Mountadar, Évaluation méso-économicoenvironnementale de la gestion des déchets solides de la ville d’Azemmour (Maroc). J. Mater. Environ. Sci. 3, 786–799 (2012) 35. S. Bettini, P. Melis, Leishmaniasis in sardinia. iii. soil analysis of a breeding site of three species of sandflies. Med. Vet. Entomol. 2(1), 67–71 (1988) 36. A. Moncaz, R. Faiman, O. Kirstein, A. Warburg, Breeding sites of Phlebotomus sergenti, the sand fly vector of cutaneous leishmaniasis in the Judean Desert. PLoS Negl. Trop. Dis. 6(7), e1725 (2012) 37. Y. Schlein, A. Warburg, L.F. Schnur, S.M. Le Blancq, A.E. Gunders, Leishmaniasis in Israel: reservoir hosts, sandfly vectors and leishmanial strains in the Negev, Central Arava and along the Dead Sea. Trans. R Soc. Trop. Med. Hyg. 78(4), 480–484 (1984) 38. WHO: Urbanisation: facteur de risque croissant pour la leishmaniose. Relevé épidémiologique hebdomadaire 7(44), 365–70 (2002)

Hydrogeochemical Study of the Hamma My Yacoube, Sidi Slimane – Morocco Salah Aitsi, Jalal Ettaki, Khalid Doumi, Ahmed Chabli, and Driss Belghyti

Abstract The source Ain Moulay Yacoub Hamma located upstream of Oued Hamma (Outita: pre-rifaine ride) 18 km southeast of the city Sidi Slimane, presents non-permanent flows during the three months February, March and April of the year 2017. The objective of this study is to evaluate the quality of the natural waters of Ain Moulay Yacoub Hamma, the hydro-chemical facies and the origin of their mineralization. Sampling is carried out at the source of these natural waters in order to carry out physico-chemical analyses in our specialised laboratory. As a result, the study showed that the waters of the Ain Moulay Yacoub Hamma spring are meso-thermal waters (42 °C), with a poor quality sulphurous odour loaded with mineral salts (chlorides, sodium and sulphates). In the results presented in the Piper, Schoeller and Stiff diagrams, show that these waters are characterized by a chloride sodium chemical facies.







Keywords Source Moulay Yacoub Hamma Oued Hamma Physico-chemistry Hydro-chemical facies Sidi Slimane Morocco










1 Introduction The river Hamma near the city Sidi Slimane, tributary of the river R'dom, is a non-permanent stream. It is fed by a set of freshwater sources upstream in the Outita pre-Riparian ride.

S. Aitsi
J. Ettaki
K. Doumi
D. Belghyti (&) Laboratory of Agro-Physiology, Biotechnology, Environment and Quality, Department of Biology, Faculty of Sciences, Ibn Tofail University, B.P:133, 14000 Kenitra, Morocco A. Chabli Regional Centre for Education and Training Professions, Department of Geology, Mohammed V University, Rabat, Morocco © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_53

657

658

S. Aitsi et al.

The discharge of thermal waters from the Ain Hamma Moulay Yacoub spring in the city of Sidi Slimane into the river could modify the quality of the latter's waters. The physico-chemical characteristics of a thermal water are linked to its underground course, its depth (temperature) and the mineral constitution of the rocks. At depth, water can also be enriched with gas (CO2, H2S) depending on the nature of the rock [1]. Thermal waters have been the subject of several scientific studies [2–7]. They have particular physico-chemical characteristics that may undoubtedly modify the quality of the receiving environment. In this context, the study of the quality of natural resources Ain Hamma Moulay Yacoub of Sidi Slimane is included. A follow-up of the physico-chemical parameters is done during the three months February, March and April in 2017. The main objective is on the one hand to study the quality of the spring water and to determine the hydro-chemical facies and the origin of the mineralization of these waters.

2 Materials and Methods 2.1

Presentation of the Study Area

The city of Sidi Slimane is located on the Gharb plain, in north-west Morocco, and belongs to the Rabat-Salé-Kénitra region. The thermal spring of El Hamma-Outita emerges on the outskirts of the pre-rifain mountain range, 12 km from the town of Sidi Slimane, on the road that leads to Meknes, and is very close to the marabout of Sidi Moulay Yacoub, at the bottom of the N-S cluse of Oued El Hamma, and its coordinates are Lambert, X = 459,60, Y = 392,7 and Z = 120 m, with a scale of 1/50000 from the map of El Kansera.

2.2 2.2.1

Geological Description Lithostratigraphic Section of the El-Hamma Wrinkle

The section of the El-Hamma wrinkle shows that the two valley wrinkles start with limestones listed as the first (S1) oldest sequence followed by other sequences of number five deposited one on top of the other in a successive manner: S2, S3, S4, S5 and S6 (with; - S2: marl-limestone; - S3: grey marls; - S4: marls and bioclastic limestone; - S5: calcareous marls and sandstones; - S6: molasses) (Fig. 2, 3). The six sequences are present on the left bank, but on the right bank there is an absence of S4 and S6.

Hydrogeochemical Study of the Hamma…

2.2.2

659

Interpretation of the Lithostratigraphic Section

The two banks of the Hamma wadi in the area of emergence of the My Yacoube Hamma spring, are formed of two lithostratigraphic formations, one limestone-limestone of Lower Domérian age and the other of marl-limestone of Upper Domérian age (FAUGERES 1978, BOUTAKIOUT 1990) (Fig. 6, 7). In the form of a small El Hamma wrinkle also called My Yacoube Hamma wrinkle. A -The Jurassic The Jurassic series composed of five formations from Domérien to Bajocien (FAUGERES 1978, BOUTAKIOUT 1990) can be seen from bottom to top: – – – – –

The limestones The marly limestones Grey marls Bioclastic marls and limestones Limestone marl and sandstone.

B - The Miocene The distinct Miocene formations are four in number, from bottom to top: lower molasses, red clays, upper molasses, and white marls. This Miocene series is transgressive on the upper Jurassic formations. – – – –

2.3 2.3.1

Lower molasses Brecciated red clays Upper molasses The White Marls

Sampling Et Analysis Sample Collection

The sampling campaign was carried out monthly over the period from February to May 2017, because of the flow of the source that is very important during this period on the other hand the other months of the year or this flow is very low to nil. Concerning the sampling method, plastic bottles were used which were rinsed several times with tap water beforehand, then allowed to dry, and before filling the bottles they were rinsed three times with water to be analysed, filled until overflowing to avoid any water–air reaction. Each bottle was labelled with the number, location and date of sampling.

660

S. Aitsi et al.

2.3.2

Physico-Chimical Analyses

The analyses concerned the major elements expressed in cations (Ca2+, Mg2+, Na+ and K+) and anions (Cl–, HCO3–, SO42–), overall hardness, full alkali content, dry residue and heavy metals (Pb, Cd, and Ni). The analyses were carried out in the laboratory of the Institut Nationale de Recherche Agronomique (Rabat). The pH and temperature measurements were made in situ by a portable instrument.

2.3.3

Evaluation of Analyses

The results of the physico-chemical analyses were used in the Avignon Hydrochemistry (L.H.A.) software version 4.2008, which enabled us to classify the waters into chemical facies and drinking and irrigation water classes, and in particular to construct the Piper, Schoeller-Berkaloff and Stiff diagrams [12].

3 Results and Discussion 3.1

Organoleptic Characteristics of the Waters of Hamma Outita Spring

– Colour: Water is clear, i.e. colourless. – Odour: Water has a sulphurous odour.

3.2

The Physico-Chemical Characteristics of the Water Studied

a -Physical parameters • Temperature: The water temperature studied varies between 41 °C and 43 °C, so we can say that this water is ortho-thermal. • Hydrogen potential (pH): The pH varies according to the concentration of HCO3− or H3O+ ions, the pH of water measured at Hamma Outita source is neutral, varies between 6.89 and 7.06.

Hydrogeochemical Study of the Hamma…

661

• Electrical conductivity: The electrical conductivity values are very high for the natural waters studied, varying between 8000 and 10,000 µS/cm. The latter result from the high mineralization of the waters of the Ain Moulay Yacoub Hamma spring. b - Chemical parameters b-1. The major elements

– The cations • The calcium ion Ca2+: Calcium is introduced into the water system by the weathering of rocks, particularly limestone rocks, and by leaching and runoff from the ground into infiltrating waters. The concentration of calcium in water depends on the residence time of the water in calcium-rich geological formations [8]. The waters of the Hamma Outita spring have an average calcium concentration of around 150 mg/l. We estimate that this is linked to the importance of the liasic formations (limestone and dolomite) of the main karstic aquifer in the supply of this spring. Moroccan standards recommend a concentration between 75 mg/l and 200 mg/l, which explains why these waters comply with national standards. • The Magnesium ion Mg2+: The magnesium content of the water is extremely varied and mainly related to the nature of the terrain crossed. In areas rich in magnesic rocks, water can contain concentrations of 10 to 50 mg/l of this element [4, 8]. The water from the Hamma Outita spring has a magnesium concentration significantly lower than that of calcium, with an average of 49.34 mg/l. Therefore these waters are of good quality in comparison with the Moroccan standard requiring a maximum concentration of 150 mg/l in Mg2+, The hardness of the water is due to dissolved polyvalent metal ions. Mainly calcium and magnesium ions. In the water samples from the Hamma Outita spring the hardness reaches a value of 57.05 °F, this would be related to the lithological nature of the aquifer formation and in particular to its composition in calcium and magnesium. Good quality drinking water has a hydrotimetric degree of less than 15 °F, they are acceptable up to 50 °F, but if they exceed 60 °F their use will cause problems either with consumption or with certain domestic uses according to the W.H.O [13]. • The sodium ion Na+: The sodium is a so-called conservative element because once in solution, no reaction makes it possible to extract it from groundwater. Precipitation brings a minimal amount of sodium to groundwater, abnormally high levels can come from

662

S. Aitsi et al.

salt leaching, or percolation through saline soils or brackish water infiltration [4, 8]. In unpolluted groundwater without contact with evaporates, the sodium content is between 1 and 20 mg/l [3]. Analysis of the data showed that the average sodium value of 713.02 mg/l, in the waters of the Hamma Outita spring, is very high, exceeding the Moroccan standard of 200 mg/l. – The potassium ion K+: Potassium is generally the least abundant major element in water after sodium, calcium and magnesium; it only exceptionally ranks third among cations [7]. Potassium occurs as double chlorides in many minerals such as corrollite and sylvinite. It is also found in plant ashes as carbonate. Potassium is an essential element for life and especially for plant growth. Potassium content is almost constant in natural waters. This usually does not exceed 10 to 15 mg/l [3]. Its concentration in the Hamma Outita spring is high and reaches an average value of around 39.31 mg/l. – The anions: • Chlorides Cl–: Chlorides are widely distributed in nature, usually in the form of calcium and potassium salts; they represent about 0.05 of the lithosphere, they are naturally present in groundwater due to weathering and leaching of sedimentary rocks and soils, and dissolution of salt deposits [7]. The average chloride value for the water source studied (Hamma Outita) is very high, reaching 1120.23 mg/l. This high value could be due to water contact with triasic saline deposits. The admissible concentration set by the Moroccan standard is 200 mg/l. • Sulphates SO42−: Under natural conditions, sulphates, the most responsive form of dissolved sulphur in natural waters, have essentially two origins: geochemical and atmospheric [1]. Due to the high solubility of sulphates, groundwater under normal conditions can contain up to 1.5 g/l [3]. Oxidation of sulphides and degradation of biomass in soil are other possible sources. Many human and natural activities can generate sulphate inputs to groundwater: application of sulphate fertilisers, precipitation loaded with sulphur dioxide, etc. The average value of sulphates in the waters studied (Hamma Outita) is about 655.88 mg/l. This high content seems to be linked to the triasic saline formation brought into contact with the aquifer reservoir through major faults that dominate the Outita wrinkle structure (Fig. 1), [2]. The results obtained exceed the Moroccan standard of between 200 mg/l and 400 mg/l in sulphates.

Hydrogeochemical Study of the Hamma…

Fig. 1 Geographic location of the study area

Fig. 2 The area of emergence of the Moulay Yacoube Hamma spring

663

664

S. Aitsi et al.

• Complete alkalimetric title: The complete alkalimetric titre (TAC) in the water samples analysed is mainly due to the presence of bicarbonates (HCO3-) is the Water Buffer Index, it is closely related to hardness, although many species of solutes can contribute to it. Alkalinity is expressed in equivalent amount of carbonate or in French degrees °F. The bicarbonate content of groundwater not subject to human influences varies between 50 and 400 mg/l [3]. The median values of bicarbonate contents are around 302 mg/l in the usual range of unpolluted groundwater [3]. The high average bicarbonate value of the waters studied is of the order of 467.8 mg/l, seems to be due to the circulation of these waters in the aquifer of calcaro-dolomitic nature. b-2. Trace elements The presence of minor elements or trace elements in thermo-mineral waters is of great interest, because it makes it possible to specify the characteristics of the waters and the mineral deposits through which they pass. For the thermo-mineral waters of our study site, we performed atomic absorption analyses to determine the concentrations in (mg/l) of certain trace elements which are: Lead (Pb), Cadmium (Cd) and Nickel (Ni). – Lead (Pb) According to the results obtained, during this campaign and by comparison with the maximum admissible values (MAV) (25 lg/l) for drinking water according to the Moroccan standard [9]. The results obtained do not exceed the MAV in the waters of the Hamma spring, which is in the order of 0.0015 mg/l. – Cadmium (Cd) The water from the source analysed has an average value of about 0.015 mg/l, higher than the MAV (3 lg/l) according to the standard for drinking water supply [9]. – Nickel (Ni) The average Nickel value of 0.02 mg/l recorded in natural waters from the Hamma source actually exceeds the MAV of drinking water, which is in the order of 500 lg/l [9].

3.3 3.3.1

Study of Relative Values Basic Exchange Index (B.E.I.)

During its underground journey, water comes into contact with different substances that have the property of exchanging their ions against those contained in the water,

Hydrogeochemical Study of the Hamma…

665

among these substances, we have clay minerals: ferric hydroxide and organic substances [11]. Schoeller in 1934 specified that the basic exchange index (B.E.I.) as being the ratio between the exchanged ions and ions of the same nature primitively existing, when there is exchange of Na+ and K+ of water, against the then calcarino-earthly clays: – If I.E.B < 0: The water is of crystalline origin (Ca2+ and Mg2+ are exchanged by Na + and K + ); – If I.E.B > 0: The water is of sedimentary origin (Na+ and K+ are exchanged by Ca2+ and Mg2+); – If I.E.B = 0: No ionic exchange between the water and the surrounding soil. With; I:E:B ¼ r Cl rðNa þ þ K þ Þ=r Cl In our case, this index is in the order of –0.005, so we can see that the Ca2+ and Mg2+ ions of the water are exchanged slightly against the K+ and Na+ ions of the surrounding formations.

3.3.2

Characteristic Reports

The characteristic ratio is defined as the ratio of certain chemical elements expressed in milliequivalents per litre (meq/l). The reports studied are: r Mg2þ =r Ca2þ ;

rðNaþ ; Kþ Þ=r Cl;

r SO42=r Cl:

The study of variations in these ratios provides valuable information on groundwater recharge and circulation and sometimes allows the detection of other non outcropping deep formations. a. Report r Mg2+/r Ca2+ When this ratio is higher than 1, it reflects the predominance of Magnesium, and when it is lower than 1 Calcium predominates and this is the case of our study (the ratio Mg2+/r Ca2+ is equal to 0.54), this can be explained by the solubility of limestones richer in Calcium than in Magnesium [10]. b. Report r SO42–/r Cl– – When this ratio is greater than 1, Sulphates predominate, which are essentially linked to leaching of gypsum soils and oxidation of sulphides (pyrites) [10]. – When the ratio r SO4 2–/r Cl– is less than 1 there is predominance of Chlorides, this is the case of our study (r SO4 2–/r Cl– equal to 0.43).

666

S. Aitsi et al.

C. Report r (Na + K+)/ r Cl–+ – When the ratio r(Na+ + K+)/r Cl– is less than 1, it reflects the predominance of Chlorides which are linked to saline soils. – When this ratio is greater than 1, it reflects the predominance of Sodium [11]. In our case, this ratio is equal to 1, which explains the existence of a balance between chlorides, potassium and sodium.

3.4

Chemical Classification of Spring Water

The values of the chemical analyses can be plotted on diagrams to classify the waters into chemical families in order to determine the facies of these waters from the Hamma spring. To determine the different facies crossed by water (source Hamma Outita) and classify them into chemical families, we used the Piper, Stiff and Schoeller Berkaloff diagrams that will be reviewed. – Piper Diagram The results were represented on the Piper diagram, with the position of the representative points of the anions and cations that characterize the chemical composition of the water from the Hamma Outita thermal spring. According to the Piper diagram (Fig. 3), the hydro-chemical facies of these waters is: Chloro-sodium.

Fig. 3 Geological profile of the My Yacoube Hamma spring emergence zone

Hydrogeochemical Study of the Hamma…

667

Fig. 4 Stratigraphic log of EL Hamma Outita (Faugères 1978, modified)

– Schoeller Diagram The graphical appearance obtained (Fig. 4) shows the facies of the mineral water concerned. The analysis of Schoeller Berkaloff's diagram allows us to conclude that the waters of the spring Hamma Outita rich in chlorides, sodium and sulphates. They are due to the existence of a source of saline sediments. The important values of chlorides, sodium and Sulphates indicate a relationship of evaporite rocks formed mainly by chlorinated and sulphated minerals. – Stiff Diagram The representation of the analysis results on the Stiff diagram (Fig. 5), clearly shows the hydro-chemical facies mentioned above which is: Chloride-sodium, with the following chemical formula: r Cl- > rSO42- > rHCO3- ≈ r(Na+ + K+) > rCa2+ > rMg2+

668

S. Aitsi et al.

Fig. 5 Piper diagram

Fig. 6 Schoeller diagram

Fig. 7 Stiff diagram

4 Conclusions The study of the chemistry of the water of the Hamma Moulay Yacoube thermal spring using hydro-chemical and hydrogeochemical tools made it possible to determine its typological characteristics: – The waters of the spring are classified in ortho-thermal family because they have a temperature of order 42 °C (37 °C < T = 42 °C < 45 °C). – The chemical facies is chloride-sodium, which shows that the waters of the source having circulated through saliferous formations of the Triassic.

Hydrogeochemical Study of the Hamma…

669

– A study of the relative values of the characteristic ratios of the mineral elements in these waters shows that: • The Ca2+ and Mg2+ ions are changed against the Na+ and K+ ions of the surrounding formations; • The predominance of calcium in relation to magnesium, this can be explained by the solubility of limestones richer in calcium than in magnesium. • The predominance of chloride in relation to sulphate may be due to the solubility of evaporites rich in chloride rather than sulphates. • The existence of a balance between chloride, potassium and sodium. – The physico-chemical quality of the source is poor, because its chemical composition in different elements does not respect the maximum admissible values for the Moroccan standards of drinking water.

References 1. A. Ezzaïdi, M. Khaloufi, M.A. Bouagou, M. El Youssi, A health tourism site to promote: abaynou resort (Guelmim province). In: The First International Workshop on Geotourism and Ecotourism in the Souss-Massa-Draâ and Guelmim-EsSmara regions. Mirleft-Guelmim and Assa. (2006) 2. O.K. Hakam, A. Choukri, J.L. Reyss, M. Lferde, Comparison of uranium and radium isotopes activities in some well and thermal springs samples in Morocco. Rev. Sci. 13/2, 185–192 (2000) 3. A. Duriez, Origin and mineralization process of thermal waters in the Mediterranean continental environment: case of the Thermopylae geothermal system (Greece). Thesis, N° d'ordre: 85 47 Université Paris Sud 11, Faculté des Sciences d’Orsay, p. 292 (2006) 4. A. Lakhdar, A. Ntarmouchant, M.L. Ribeiro, M. Beqqali, K. El Ouadeihe, L. Benaabidate, M. Dahire, Y. Driouche, A. Benslimane. A., Nouvelle approche géologique et géochimique du complexe hydrothermal de Moulay Yacoub (frontière nord du sillon du Rifain Sud). Comunicacoes Geologicas 93, pp 185–204 (2006) 5. A. Lakhdar, A. Ntarmouchant, M.L. Ribeiro, M. Beqqali, K. El Ouadeihe, L. Benaabidate, M. Dahire, Y. Driouche, A. Benslimane, Determination of the origin of the mineralization of the thermal waters of Moulay Yacoub by geological and geochemical approaches. Revue des Énergies Renouvelables. CER’07, Oujda, pp. 81–84 (2007) 6. Y. Zarhloule, A. Rimi, M. Boughriba, M. Verdoya, A. Correia, J. Carneiro, A. Lahrach, The geothermal province of North Eastern Morocco. Revue des Énergies Renouvelables. CER’07, Oujda, pp. 89–94 (2007) 7. A. Fekraoui, Geochemical characteristics of the geothermal waters of the Oran region. Revue des Énergies Renouvelables CER’07, Oujda, pp. 75–80 (2007) 8. J. Rodier, B. Legube, N. Merlrt, Water Analysis. 9th edn. (Dunod, Paris, 2009), p. 1579 9. MOROCCAN STANDARD (2006): Moroccan standard approved by joint order of the Minister of Industry, Trade and Economic Upgrading and the Minister of Equipment and Transport and the Minister of Health No. 221–06 of 2 February 2006, published in the Official Gazette. No. 5404, of 16 March 2006.

670

S. Aitsi et al.

10. M. Benhamza, Contribution of geophysics to the hydrogeological study of the Mercurielle Nord Numidique zone (Azzaba). North-East Algeria, Consequences of the exploitation of mercury deposits on the environment, PhD thesis, UBM Annaba, p. 147 (2007) 11. H. Dib, The thermalism of the far east of Algeria: Guelma, Souk Ahras, Skikda and Tarf. Supervision of a thesis in Hydrogeology. Faculty of Earth Sciences and Town and Country Planning. Constantine University (2004) 12. MIOURIGH: Evaluation of the hydrochemical quality of groundwater in the M'Zab valley: Case of Oued N'Tissa. Thesis of State Engineer in Agronomy, option Agricultural Hydraulics. 13. O.M.S: Drinking Water Quality Guidelines; Hygiene Criteria and Supporting Documentation WHO, Geneva, 2nd edn, vol. 2 (2002), p. 1050

Hydrogen Production via Wastewater Electrolysis—An Integrated Approach Review M. Cartaxo , J. Fernandes , M. Gomes , H. Pinho , V. Nunes , and P. Coelho

Abstract Human activities generate enormous amounts of wastewater. The hydrogen production from this new resource has gained attention as an emergent technology. Incorporating photovoltaic energy production with different electrolysis systems which can treat wastewaters and produce hydrogen simultaneously will lead to an environmentally-friendly and sustainable hydrogen production. Keywords Wastewater


Water
Electrolysis
Hydrogen
Renewable energy

1 Introduction Environmental concerns regarding the high demand for energy are one major issue, as human civilization is accompanied by an ever-growing need for energy. There is a global effort to reduce greenhouse gas (GHG) emissions associated with fossil fuels, as their use has a tremendous destructive environmental impact, such as climate change, ozone layer depletion, acidification, and air pollution [1, 2]. M. Cartaxo (&)
M. Gomes
H. Pinho
V. Nunes
P. Coelho Smart Cities Research Center – Ci2 – Instituto Politécnico de Tomar, Quinta do Contador, Est. da Serra, Tomar, Portugal e-mail: [email protected] M. Gomes e-mail: [email protected] H. Pinho e-mail: [email protected] V. Nunes e-mail: [email protected] P. Coelho e-mail: [email protected] J. Fernandes Instituto Politécnico de Tomar, Quinta do Contador, Est. da Serra, Tomar, Portugal e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_54

671

672

M. Cartaxo et al.

In the near future, most of the energy required for human activities should be provided by incident solar radiation instead of the air combustion of fossil hydrocarbons or nuclear reactions. This would prevent the deleterious impact of GHG emissions on the environment and would also bring some additional benefits such as the promotion of the energy independency of nations, a reduction in the competition for access to oil resources and could also contribute to solving the problems associated with the treatment and long-term storage of nuclear wastes, promoting an energy transition [3]. Implementing such an energy transition on a large scale and turning sunlight into renewable electricity needs an easy and affordable way to transport and store it. This energy transition depends on the development of technical solutions for energy storage at different time and size scales. A universal chemical energy carrier which could be used for storage, transportation and distribution of renewable energy is required [3]. Gaseous hydrogen of electrolytic grade has been considered for many decades to be such a game-changing energy carrier. This is a by-product of brine electrolysis, but it can also be obtained in an efficient and affordable way through water electrolysis, while the Earth's atmosphere can be used as an oxygen reservoir of infinite capacity. Commercial water electrolysers are found in quite varied fields in industry, but so far the technology has been mainly used for stationary production in view of chemical applications. Regarding energy applications, existing plants of limited size are available only on a demonstration scale. Future water electrolysers will have to be customized to allow non-stationary modes of operation. Therefore, efforts are currently being made to develop hundred-MW scale systems [3]. In recent years, hydrogen has increasingly become an alternative energy carrier in the energy and automotive sectors, due to its renewable, transportable and emission-free properties [1, 4]. Moreover, by using fuel cells, hydrogen can be converted into electricity and operated continuously under hydrogen and oxygen. Unlike fossil fuels, hydrogen is not available freely in nature as a primary fuel. This means that it must first be produced from its sources such as natural gas or water, and then used as an energy source. Currently, reforming fossil fuels is the most common process used for commercial production of hydrogen, but this process has associated emissions of carbon dioxide (CO2) and low efficiency. Therefore, there is great interest in developing alternative hydrogen production processes based on water-splitting (e.g. thermolysis, thermochemical, electrochemical, photochemical, photocatalytic and photoelectrochemical processes) which are environmentally friendlier and more sustainable. Taking the current environmental and energy issues into account, water electrolysis can be an efficient, clean and promising alternative technology for hydrogen production. This technique could be integrated well with the available renewable sources (such as solar, wind or hydropower). The major limitation of large-scale water electrolysis for mass production of hydrogen is the rather high overpotential required for the reactions. This overpotential plus the ohmic losses in the electrolysis process make the actual potential to exceed the standard 1.23 V potential of water electrolysis and reach values of around 1.8– 2.0 V [4].

Hydrogen Production via Wastewater Electrolysis…

673

Most human and industrial activities lead to enormous wastewater generation. The majority of wastewater is routed for disposal in the environment without treatment, due to a lack of strict regulations in many developing countries. It causes a detrimental impact on human health, economic productivity, freshwater resource quality and the ecosystem. Sources of wastewater generated include domestic residence, commercial properties, industrial operations and agriculture. Hence, management of this wastewater is necessary with safe and effective collection methods. Treatment of wastewater and returning treated water back into the environment could have the least impact on the environment and ecosystem. Furthermore, wastewater is a valuable resource, which has more utility in generating cost-effective and valuable sustainable sources of energy, nutrients, organic matter, and other useful by-products.

2 Hydrogen Production Due to its highest gravimetric energy density, hydrogen has been regarded as the preferred clean-energy carrier, with potentially environmentally-friendly production through the solar-assisted splitting of water [5, 6]. Due to fluctuations in renewable energy production and consumption rates, “buffers” for energy storage, such as electrochemical energy conversion and storage systems (EECSs), are required. The most promising EECSs are redox flow cells, regenerative fuel cells, electrochemical conversion of CO2 (and water or H2) into fossil fuels (e.g. formic acid) and electrochemical capacitors. The annual consumption of hydrogen increases by 6% per year. Today, hydrogen is produced mainly from fossil fuels, mostly natural gas reforming and coal gasification [7], so these methods result in high CO2 emissions, and efficiency is approximately 70– 75% and 45–65% respectively. A small percentage of hydrogen (about 4%) is produced by water electrolysis, which consumes substantial energy (6–7 kWh per m3 of hydrogen) [8]. Large amounts of pure hydrogen (99.999 vol%) can be produced using water electrolysis without emissions of gaseous pollutants. For these reasons, water electrolysis is considered today to be the most important hydrogen-production technology. The main research on hydrogen energy now focuses on the development of energy-efficient water electrolysis systems (electrode materials, electrolytes, separators, etc.) and effective H2 storage methods required due to the low volumetric energy density of H2 [7]. The water-splitting reactions (Eq. (1)) are endothermic and therefore require energy, which can be provided by an electric current through a suitable electrochemical cell.

674

M. Cartaxo et al.

H2 O + energy ! H2 þ 1=2 O2

ð1Þ

Although it seems simple, the electrochemical splitting of water requires a detailed understanding of the anodic and cathodic mechanisms to achieve high rates and minimum electrical energy input needs. The hydrogen evolution reaction (HER) exchange current of Pt is at least two orders of magnitude higher in acid than in alkaline electrolytes. On the other hand, the oxygen evolution reaction (OER), may occur through various pathways and adsorbed intermediates. RuO2 and IrO2, alone or in combination, are often considered the reference OER catalysts, but neither is “ideal”. The reverse of the OER reaction is the oxygen reduction reaction (ORR), where molecular oxygen is reduced to water. ORR involves the same intermediates as the OER and is most conveniently carried out in an alkaline environment where catalysts which are more active and stable than in acid are available. Therefore, it remains a challenge to generate catalysts (beyond Pt, Ir and Ru) for both the HER and OER which can exhibit selectivity, remain stable over time, and can be produced using low-cost methods in large quantities. To reduce the cost of water electrolysis, substantial effort has been made to develop catalysts such as non-noble metals and non-metallic materials. Because high-activity Pt-based materials’ low element abundance hinders their large-scale application, more cost-effective, earth-abundant catalysts with long-term stability are needed for the HER. To meet the demand, a variety of transition metal-based catalysts, including their corresponding single atomics, alloys, nitrides, phosphides, carbides, selenides, sulphides, and tellurides, have been proposed. Examples are ternary metal phosphides (in particular, of Co and Mo) [5, 6, 9–13]. Water electrolysis could be carried out in different types of electrolytic systems such as alkaline electrolysers, PEM electrolysers, and the solid-oxide electrolysers (SOEs). However, using expensive materials (precious metals) in the fabrication of PEM or SOE electrolysers, along with their shorter lifetime, make them relatively more expensive compared to the alkaline cells for small-scale H2 production systems. The current commercial alkaline water electrolysers mainly use aqueous solutions of sodium or potassium hydroxide as the state-of-the-art electrolyte. Although alkaline water electrolysis possesses an inherent low-cost characteristic, the development of large-scale hydrogen production systems based on this technology still needs potential improvements in the overall cell efficiency through reduction of the energy consumption and increase of the electrolyte ionic conductivity [4, 14, 15]. The “conventional” alkaline electrolysers typically operate with cell potentials near 2 V with a system efficiency near 60% and demonstrated lifetimes of more than 10 years. Commercial PEM electrolysers are available in smaller sizes than alkaline electrolysers with a lifetime of 5 or more years, where lifetime certainly would be a function of the operating temperature, resulting in increased efficiency but decreased durability when temperature rises. Furthermore, increased catalyst loading would cause the cost of the device to rise. Not only the price of the noble

Hydrogen Production via Wastewater Electrolysis…

675

metals is high, but rather also system complexity, difficulties in scaling up, and the high cost of the membrane are the main disadvantages of PEM water electrolysers [9]. There is an incentive today to improve alkaline electrolysis over system scales larger than PEM electrolysis. This can be achieved in several ways: select better HER-OER catalysts; use system components and architectures which allow operation at high temperatures and reduce system complexity, capital, and operating costs; use an anion exchange membrane (AEM) in place of the electrolyte, resulting in a compact system, such as a PEM device and free of bubble effects, with a large choice of non-PGM catalysts, with low resistance and high stability [9].

3 Renewable Energy Sources Contributions of fossil fuels (oil, coal and gas), hydropower, nuclear power and renewable resources (e.g. solar and wind) to global energy demand are approximately 86%, 6.8%, 4.43% and 2.77% respectively. Today, more than 70% of electrical energy is generated from fossil fuels. In fact, the burning of fossil fuels in recent decades has caused serious environmental problems with a huge impact on the climate change and global warming. Concerns about global CO2 emissions and increasing costs of fossil fuels are motivating the development of technologies for energy production from renewable resources such as wind and solar energy [7]. Despite most Renewable Energy Systems (RESs) being known as “good” solutions to produce “green” energy, they require different structures which provide different energy costs and CO2 emissions [16]. Therefore, it is necessary to identify the appropriate structure for RESs to achieve energy sustainability. In the last decade, distributed energy resources (DERs) gained more attention and interest from governments because they are more “flexible” and have a low CO2 emission. Photovoltaic (PV) systems are one of the most representative DERs, and are widely used in power generation [17, 18]. PV is expected to grow significantly as global energy transitions towards renewable energy, and by mid-century, PV systems will supply a large portion of the world’s energy needs beyond the power sector as the result of new electricity-to-fuel technologies. Most of these technologies involve the production of renewable hydrogen from intermittent renewable electricity sources [19]. The production of H2 based on an energy mix through a system using a mini-hydro plant (MHP) and a PV system was evaluated. The purpose of an MHP complementary to the PV system is to supply electrical energy to a water electrolyser, as the PV system is an intermittent power source and only works with sunlight, and the study has shown its technical and economic viability [19]. In addition, combined photovoltaic/thermal (PV/T) systems which simultaneously produce electricity and hot water can be considered in H2 production, leading to higher electrical efficiency and an extended service life [20].

676

M. Cartaxo et al.

4 Wastewater Electrolysis Industrialization demands high water and energy consumption and generates large amounts of wastewater. The treatment of wastewater along with the generation of renewable energy resources is an ideal way of tackling these problems [1, 3]. Recently hydrogen production from wastewater has gained attention as an emergent technology. Hydrogen generation can be achieved by various methods, such as dark fermentation, photo fermentation, bio photolysis and microbial electrolysis. Using raw industrial wastewater for hydrogen production can be a high energy-intensive and low-yield process. This is because raw industrial wastewater consists of complex and recalcitrant biodegradable compounds which hinder the hydrogen yield and specific hydrogen production rate. Hence, to enhance the process’s feasibility and sustainability, pre-treatment processes such as physical, chemical, biological and mechanical ones can be incorporated as an additional step. The pre-treatment process accelerates the hydrogen generation capacity of wastewater and enriches the hydrogen yield [1, 2]. Electrochemical methods are used to remove various organic and inorganic species of pollutants from wastewater. Using electrochemical methods for the treatment of organic effluents has some advantages compared with biological or chemical methods. In this context, water-treatment technologies based on electrochemical advanced oxidation processes (EAOPs) are more effective. This is due to the action of highly reactive oxygen species (ROS) such as hydroxyl radical *OH in the reaction mechanism, usually based on anodic oxidation (AO) or Fenton's reaction. Anodic oxidation combines high efficacy and simplicity, favouring the oxidative destruction of organics in waters through the direct action of *OH formed at the anode surface by oxidation of water via (Eq. (2)). M þ H2 O ! Mð HOÞ þ H þ þ e

ð2Þ

In the anodic oxidation, the oxidation ability depends on the selected anode, and a wide variety of electrode materials have been investigated [3]. For example, dairy wastewater treatment was performed by electro-Fenton process [21]. Electrocoagulation was investigated for the treatment of agro-industry wastewaters [22], as well as swine manure effluents [23]. Electrochemical reactions, which can produce molecular hydrogen via water splitting, can degrade organic contaminants in water simultaneously, and the presence of wastewater organic matter can enhance hydrogen generation [24]. The electrooxidation of selected organic compounds was investigated to determine optimal process conditions towards the production of hydrogen from potential wastewater sources [25]. The photodegradation of organic wastewater and simultaneous hydrogen evolution was studied, using the photo-Fenton reaction [26, 27]. The HER in black liquor electrolysis was performed and it was concluded that it is kinetically more favourable than alkaline water electrolysis [28]. MoSx was used as an electrocatalyst for HER in real acidic wastewaters [29]. Waste generated by the

Hydrogen Production via Wastewater Electrolysis…

677

wine industry was also studied in a low-temperature PEM electrolysis cell [30]. Electro hydrolysis of landfill leachates was proven to be an effective method for hydrogen gas production with simultaneous COD removal [31], as well as olive mill wastewaters [32, 33], vinegar wastewaters [34], waste anaerobic sludges [35], domestic wastewaters [36–38], swine wastewaters [39], chicken industry wastewaters [40] and cheese whey wastewaters [27]. One different approach is to directly electrolyse biomass wastes [41, 42]. Hydrogen gas can also be produced from metal-plating wastewater by electro hydrolysis (or electro-coagulation process) [2].

5 Integration and Related Works Integrating wastewater treatment and hydrogen production powered by RESs can be a way to improve the economy of green hydrogen. This approach is shown in Fig. 1. There have been a few attempts to follow this integration path. At a lab scale, the direct treatment of wastewater generated in a textile plant by electrolysis powered by a photovoltaic system was evaluated. The results were promising for that kind of wastewater [43]. Similar experiments were carried out by evaluating the potential to produce H2 through electrolysis of wastewater from ornamental-stone industries. The prototype of an electrolyser powered by a photovoltaic system was developed, which was shown to be effective in simultaneously producing H2 and treat wastewater [44]. In another small-scale setup, experiments to produce H2 by electrolysis powered by photovoltaic cells and using low-strength industrial wastewater were carried out [45, 46]. The “Greenlysis” project aimed to demonstrate the potential of producing H2 via electrolysis of wastewater, powered by renewable electricity. In this approach, a PEM electrolyser and the treatment units of a wastewater treatment plant are powered by PV panels and wind turbines [47]. Experiments at a sub-pilot-scale for H2 production from urban wastewater using electrolysis powered by a PV system were also carried out [48]. There are few examples in the literature close to the proposed integration approach, and the idea of using hydrogen to power the electrical grid is a new Fig. 1 Integration of hydrogen production with wastewater treatment, renewable energy sources and the electricity grid

678

M. Cartaxo et al.

aspect in this area. This work is in its early stages. The authors are committed to pursuing it in the near future, using the PV system infrastructure already in place at the Polytechnic Institute of Tomar campus, aiming to study and test the viability of producing hydrogen from wastewater, by electrolysis powered by PV systems and using this hydrogen as fuel for different applications.

6 Conclusions The integrated approach of hydrogen production through electrolysis of wastewater powered by renewable sources of electricity seems to have the potential to boost the hydrogen economy. The coupling of hydrogen production with wastewater treatment can contribute to simultaneously reducing the costs of wastewater treatment and the environmental impact of hydrogen production. Powering the electrolysis of wastewater with renewable sources of electricity can contribute to balancing the electricity grid and to the use of non-renewable sources, towards a full greener production of hydrogen. Acknowledgements This research was funded by the Portuguese Foundation for Science and Technology (FCT), grant number UIDP/05567/2020. Special thanks to Trina Cairns (native English speaker and professional proofreader), who reviewed the text of this work.

References 1. V.G. Sharmila, J.R. Banu, S.-H. Kim, G. Kumar, A review on evaluation of applied pre-treatment methods of wastewater towards sustainable H2 generation energy efficiency analysis. Int. J. Hydrog. Energy 45(15), 8329–8345 (2020) 2. E. Çokay, Y. Gürler, Effects of metals in wastewater on hydrogen gas production using electrohydrolysis. Int. J. Hydrog. Energy 45, 3407–3413 (2020) 3. N. Klidi, D. Clematis, M. Delucchi, A. Gadri, S. Ammar, M. Panizza, Applicability of electrochemical methods to paper mill wastewater for reuse. Anodic oxidation with BDD and TiRuSnO2 anodes. J. Electroanal. Chem. 815, 16–23 (2018) 4. A.H. Bahman, M. Choolaei, A. Chaudhry, H. Qaalib, A highly-efficient hydrogen generation electrolysis system using alkaline zinc hydroxide solution. Int. J. Hydrog. Energy 44(1), 72– 81 (2019) 5. M. Liu, F. Hof, M. Moro, G. Valenti, F. Paolucci, A. Pénicaud, Carbon supported noble metal nanoparticles as efficient catalysts for electrochemical water-splitting. Nanoscale 12, 20165– 20170 (2020) 6. L. Gong, K. Lan, X. Wang, X. Huang, P. Jiang, K. Wang, M. Yang, L. Ma, R. Li, Carbon-coated Co-Mo-P nanosheets supported on carbon cloth as efficient electrocatalyst for Hydrogen Evolution Reaction. Int. J. Hydrog. Energy 45(1), 544–552 (2020) 7. G. Amikam, P. Nativ, Y. Gendel, Chlorine-free alkaline seawater electrolysis for hydrogen production. Int. J. Hydrog. Energy 43(13), 6504–6514 (2018) 8. C. Chen, Q. Bai, J. Liu, Z. Wang, K. Cen, Characteristics and anode reaction of organic wastewater-assisted coal electrolysis for hydrogen production. Int. J. Hydrog. Energy 45(41), 20894–20903 (2020)

Hydrogen Production via Wastewater Electrolysis…

679

9. S. Marini, P. Salvi, P. Nelli, R. Pesenti, M. Villa, M. Berrettoni, G. Zangari, Y. Kiros, Advanced alkaline water electrolysis. Electrochim. Acta 82, 384–391 (2012) 10. X. Huang, X. Xu, X. Luan, D. Cheng, CoP nanowires coupled with CoMoP nanosheets as a highly efficient cooperative catalyst for hydrogen evolution reaction. Nano Energy 68, 104332 (2020) 11. M. Bhavanari, K.-R. Lee, C.-J. Tseng, I.-H. Tang, H.-H. Chen, CuFe electrocatalyst for hydrogen evolution reaction in alkaline electrolysis. Int. J. Hydrog. Energy (2021) 12. T. Zhao, Y. Wang, S. Karuturi, K. Catchpole, Q. Zhang, C. Zhao, Design and operando/in situ characterization of precious-metal-free electrocatalysts for alkaline water splitting. Carbon Energy 2, 582–613 (2020) 13. N. Yao, R. Meng, F. Wu, Z. Fan, G. Cheng, W. Luo, Oxygen-Vacancy-Induced CeO2/Co4N heterostructures toward enhanced pH-Universal hydrogen evolution reactions. Appl. Catal. B: Environ. 277, 119282 (2020) 14. A. Buttler, H. Spliethoff, Current status of water electrolysis for energy storage, grid balancing and sector coupling via power-to-gas and power-to-liquids: A review. Renew. Sustain. Energy Rev. 82, 2440–2454 (2018) 15. S.A. Grigoriev, V.N. Fateev, D.G. Bessarabov, P. Millet, Current status, research trends, and challenges in water electrolysis science and technology. Int. J. Hydrog. Energy 45, 26036– 26058 (2020) 16. D. Gielen, F. Boshell, D. Saygin, M.D. Bazilian, N. Wagner, R. Gorini, The role of renewable energy in the global energy transformation. Energ. Strat. Rev. 24, 38–50 (2019) 17. Y. Wu, H. Lin, Standards and guidelines for grid-connected photovoltaic generation systems: a review and comparison. IEEE Trans. Ind. Appl. 53, 3205–3216 (2017) 18. D. Xu, G. Wang, W. Yan, A novel adaptive command-filtered backstepping sliding mode control for PV grid-connected system with energy storage. Sol. Energy 178, 222–230 (2019) 19. C.A. Pereira, P.M. Coelho, J.F. Fernandes, M.H. Gomes, Study of an energy mix for the production of hydrogen. Int. J. Hydrog. Energy 42(2), 1375–1382 (2017) 20. Y. Cui, J. Zhu, S. Zoras, J. Zhang, Comprehensive review of the recent advances in PV/T system with loop-pipe configuration and nanofluid. Renew. Sustain. Energy Rev. 135, 110254 (2021) 21. R. Davarnejad, M. Nikseresht, Dairy wastewater treatment using an electrochemical method: Experimental and statistical study. J. Electroanal. Chem. 775, 364–373 (2016) 22. P. Drogui, M. Asselin, S.K. Brar, H. Benmoussa, J.-F. Blais, Electrochemical removal of pollutants from agro-industry wastewaters. Sep. Purif. Technol. 61, 301–310 (2008) 23. S. Rahman, M.S. Borhan, Electrolysis of swine manure effluents using three different electrodes Fe-Fe, Al-Al and Fe-Al. Am. J. Agric. Biol. Sci. 9(4), 490–502 (2014) 24. H. Park, K.-H. Choo, H.-S. Park, J. Choi, M.R. Hoffmann, Electrochemical oxidation and microfiltration of municipal wastewater with simultaneous hydrogen production: Influence of organic and particulate matter. Chem. Eng. J. 215–216, 802–810 (2013) 25. W. Cheng, N. Singh, J.A. Maciá-Agulló, G.D. Stucky, E.W. McFarland, J. Baltrusaitis, Optimal experimental conditions for hydrogen production using low voltage electrooxidation of organic wastewater feedstock. Int. J. Hydrog. Energy 37, 13304–13313 (2012) 26. Y.-T. Lin, Y.-H. Wang, J.C.S. Wu, X. Wang, Photo-Fenton enhanced twin-reactor for simultaneously hydrogen separation and organic wastewater degradation. Appl. Catal. B: Environ. 281, 119517 (2021) 27. F. Kargi, S. Uzunçar, Valorization of cheese whey by electrohydrolysis for hydrogen gas production and COD removal. Waste Biomass Valoriz. 4, 517–528 (2013) 28. H.R. Ghatak, S. Kumar, P.P. Kundu, Electrode processes in black liquor electrolysis and their significance for hydrogen production. Int. J. Hydrog. Energy 33, 2904–2911 (2008) 29. M. Kokko, F. Bayerköhler, J. Erben, R. Zengerle, P. Kurz, S. Kerzenmacher, Molybdenum sulphides on carbon supports as electrocatalysts for hydrogen evolution in acidic industrial wastewater. Appl. Energy 190, 1221–1233 (2017)

680

M. Cartaxo et al.

30. A. Caravaca, A. de Lucas-Consuegra, A.B. Calcerrada, J. Lobato, J.L. Valverde, F. Dorado, From biomass to pure hydrogen: electrochemical reforming of bio-ethanol in a PEM electrolyser. Appl. Catal. B 134–135, 302–309 (2013) 31. F. Kargi, E.C. Catalkaya, Electrohydrolysis of landfill leachate organics for hydrogen gas production and COD removal. Int. J. Hydrog. Energy 36, 8252–8260 (2011) 32. F. Kargi, E.C. Catalkaya, Hydrogen gas production from olive mill wastewater by electrohydrolysis with simultaneous COD removal. Int. J. Hydrog. Energy 36, 3457–3464 (2011) 33. N.A. Oz, A.C.U. Eker, Simultaneous hydrogen production and pollutant removal from olive mill wastewaters using electrohydrolysis process. Chemosphere 232, 296–303 (2019) 34. F. Kargi, S. Arikan, Hydrogen gas production from vinegar fermentation wastewater by electro-hydrolysis: effects of initial COD content. Int. J. Hydrog. Energy 38, 2701–2708 (2013) 35. F. Kargi, E.C. Catalkaya, S. Uzuncar, Hydrogen gas production from waste anaerobic sludge by electrohydrolysis: effects of applied DC voltage. Int. J. Hydrog. Energy 36, 2049–2056 (2011) 36. C.C. Yarımtepe, B. Türen, N.A. Oz, Hydrogen production from municipal wastewaters via electrohydrolysis process. Chemosphere 231, 168–172 (2019) 37. K. Cho, M.R. Hoffmann, Molecular hydrogen production from wastewater electrolysis cell with multi-junction BiOx/TiO2 anode and stainless steel cathode: current and energy efficiency. Appl. Catal. B 202, 671–682 (2017) 38. L. Sharma, S. Prabhakar, V. Tiwari, A. Dhar, A. Halder, Optimization of EC parameters using Fe and Al electrodes for hydrogen production and wastewater treatment. Environ. Adv. 3, 100029 (2021) 39. G. Lourinho, P.S.D. Brito, Electrolytic treatment of swine wastewater: recent progress and challenges. Waste Biomass Valoriz. 12, 553–576 (2021) 40. K. Thirugnanasambandham, V. Sivakumar, J.P. Maran, Optimization of process parameters in electrocoagulation treating chicken industry wastewater to recover hydrogen gas with pollutant reduction. Renew. Energy 80, 101–108 (2015) 41. T. Hibino, K. Kobayashi, M. Ito, M. Nagao, M. Fukui, S. Teranishi, Direct electrolysis of waste newspaper for sustainable hydrogen production: an oxygen-functionalized porous carbon anode. Appl. Catal. B 231, 191–199 (2018) 42. L. Yang, W. Liu, Z. Zhang, X. Du, J. Gong, L. Dong, Y. Denga, Hydrogen evolution from native biomass with Fe3+/Fe2+ redox couple catalyzed electrolysis. Electrochimica Acta 246, 1163–1173 (2017) 43. A.K. Pathak, R. Kothari, V.V. Tyagi, S. Anand, Integrated approach for textile industry wastewater for efficient hydrogen production and treatment through solar PV electrolysis. Int. J. Hydrog. Energy 45, 25768–25782 (2020) 44. F.C. Marques, J.C.M. Silva, C.P. Libardi, R.R. de Carvalho, G.F. Sequine, G.M. Valane, Hydrogen production by photovoltaic-electrolysis using aqueous waste from ornamental stones industries. Renew. Energy 152, 1266–1273 (2020) 45. F. Kargi, Comparison of different electrodes in hydrogen gas production from electrohydrolysis of wastewater organics using photovoltaic cells (PVC). Int. J. Hydrog. Energy 36, 3450–3456 (2011) 46. S. Eker, F. Kargi, Hydrogen gas production from electrohydrolysis of industrial wastewater organics by using photovoltaic cells (PVC). Int. J. Hydrog. Energy 35, 12761–12766 (2010) 47. E. Marzo, A. Gali, B. Lefevre, L. Bouchy, A. Vidal, J.L. Cortina, A. Fabre, Hydrogen and oxygen production using wastewater effluent treated with ultra-filtration and membrane distillation (Greenlysis). Procedia Eng. 44, 1744–1746 (2012) 48. J. Choi, H.-H. Ou, S.H. Soong, H. Park, K.-H. Choo, M.R. Hoffmann, Sub-pilot-scale hybrid electrochemical system for hydrogen production and wastewater treatment using solar panel. In: ACS National Meeting Book of Abstracts, 2011 241st ACS National Meeting and Exposition, Anaheim, CA, 27 March 2011–31 March 2011

Flood Aleas Diagnostic and Assessment Case of the Jebha Zone Mohammed Benessayyad, Soufiane Saber, Driss Belghytı, and Kacem Naımı

Abstract Among the areas threatened by the floods and which integrate into the sustainable development components, the area of Jebha, which is located on the Mediterranean facade, depends on the commune of M’tioua and falls under the province of Chefchaouen, the center of Jebha is located in a back-country formed by high reliefs with steep rock slopes, There are many talwegs and intermittent rivers around the site, such as Wadi Mesiaba, which has experienced untimely flooding adjacent to the town. Several dwellings too close to the river were invaded by a surge of water that fell violently for more than 24 h on this area. The water level has reached nearly 40 to 50 cm in some places, a large part of the road is flooded with mud. However, the only sustainable stream is the Ouringa Wadi which flows into the Mediterranean. The catchment area of the Jebha area is of the peri-urban type, which manifests itself and exhibits behaviors of natural and urban catchments, this basin has subjected to a strong pressure, linked to the development of the urban part, and It is also subject to natural factors and constraints. Keywords Floods


Development
Basin
Peri-urban

1 Introduction Flood protection is one of the major challenges in water resource management. Indeed, the high irregularity of hydrological regimes, the predominance of mountainous terrain and the nature of the land, and the vegetation cover is often impermeable so that runoff is important and that the Jebha rivers generate large and violent floods. These sometimes lead to floods which have often caused major M. Benessayyad
S. Saber
D. Belghytı (&) Faculty of Science, Laboratory of Natural Resources and Sustainable Development, University Ibn Tofail, Kenitra, Morocco K. Naımı Centre of Education Guidance and Planing, Rabat, Morocco © The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 M. Ben Ahmed et al. (eds.), Innovations in Smart Cities Applications Volume 5, Lecture Notes in Networks and Systems 393, https://doi.org/10.1007/978-3-030-94191-8_55

681

682

M. Benessayyad et al.

damage to public infrastructure, people’s property, agriculture, and also completely paralyzes the economic activities of the city, activities related to the port of Jebha. This phenomenon of flooding is not recent at the level of the city of Jebha, except that it is today felt more and more strongly because of the demographic, economic, urban and tourist development that the area is experiencing. Moreover, the current situation of the area of Jebha, has been clearly aggravated by the sometimes-anarchic urbanization that operates in the area and its peripheries. Indeed, from Oued Messiaba, Chaabas and their flood-prone areas, which were once protected areas and not aedificandi, have become areas increasingly coveted by developers. – What is the history of flooding in the region? – What types of floods in the Jebha area?

2 Area Flood History Referring to local memory, the area of Jebha experienced several decreed events, and according to the inhabitants of Jebha, this is not the first time this has happened. The city is flooded every three or four years. They occurred in 1957, 1982, 1992, 1995, 2008, 2011, 2015 and 2020. Despite the importance of these floods, no archived information on the description of these floods and the extent of the damage caused is available. The testimonies of the local population and representatives of the commune were collected. During our visits, we found that for the area of Jebha, Oued Messiaba drains the city with a strong potential for runoff. An urban development that increasingly strangles the bed of the wadi. This shrinkage, particularly in the lower part of the left bank, affects the natural areas of solid deposits and consequently its hydraulic flow.

Fig. 1 Downstream part of the Messiaba wadi (to the right of the RN16 crossing)

Flood Aleas Diagnostic and Assessment Case of the Jebha Zone

683

Fig. 2 Delimitation of the watersheds of the study area

Table 1 The characteristics of the Oued Mesiaba watershed BV

S (Km2)

Medium slope (m/m)

L (km)

Tc (min)

Lag (min)

Vp (m/s)

Cr adim

Messiaba

16.00

0.125

7.50

41.56

24.94

2.97

0.51

Solid transport increases downstream, which justifies the clogging of the structure that ensures the crossing of RN16 and overflows at this point. The figure taken from the SDAU of the city and the following photos, illustrate the problem posed from the downstream to the upstream of the wadi Messiaba, before the realization trapezoidal channel open air reinforced concrete (Fig. 1): During the years of 2015 and 2016, and in order to compensate for the flood problems posed, work was carried out, by the Loukkos Water Basin Agency, following the studies carried out at the enclosure of the area, they propose: – An open-pit trapezoidal channel of reinforced concrete of type TR10.0-D (b = 10 m and h = 2 m) on *1000 m allowing the capture of the waters occurred; – Bonded 400–500 mm  50 m seam transitions upstream and downstream of the projected TR10.0-D channel; – Trapezoidal channel projected successive falls of height of 1.0 m to limit the flow speed by respecting a maximum slope of 3.0%; – Screened walls upstream of the projected canal to reduce the volume of bodies loaded. Reconstruction of an art work that franchise oued, at the level of the RN16.

684

M. Benessayyad et al.

Table 2 Table of runoff coefficients by area and vegetation cover Nature of the plant cover

Small basin from 0 to 10 ha

Medium basins from 10 to 1500 ha

With a slope of: Less of 5%

from 5 to10%

from 10 to 30%

More than 30%

Less of 1.5%

From 1.5% to 5%

From 5 to 10%

From 10 to 30%

More than 30%

Flat - form and put on by roads

0.95

0.95

0.95

0.95

0.95

0.95

0.95

0.95

0.95

Bare ground or in vegetation couvrante not, Fields already tainted by erosion, cool Plowing

0.80

0.85

0.90

0.95

0.60

0.65

0.75

0.80

0.85

Cultures couvrantes high grain, Fields of scattered courses, small jungle

0.75

0.80

0.85

0.90

0.47

0.51

0.80

0.72

0.80

Meadows, dense Jungles, Savannah with brushwood

0.70

0.75

0.80

0.85

0.30

0.30

0.36

0.42

0.50

Ordinary forest in timber, brushwood thick

0.30

0.50

0.60

0.70

0.13

0.13

0.20

0.25

0.30

Big primary forest

0.20

0.25

0.30

0.40

0.15

0.15

0.18

0.18

0.25

(Hydraulic road, BCEOM page 115, changed by the addition of slopes