Internet of Things (IoT): Key Digital Trends Shaping the Future: Proceedings of 7th International Conference on Internet of Things and Connected ... (Lecture Notes in Networks and Systems, 616) 9811997187, 9789811997181

Citation preview

Lecture Notes in Networks and Systems 616

Rajiv Misra · Muttukrishnan Rajarajan · Bharadwaj Veeravalli · Nishtha Kesswani · Ashok Patel   Editors

Internet of Things (IoT): Key Digital Trends Shaping the Future Proceedings of 7th International Conference on Internet of Things and Connected Technologies (ICIoTCT 2022)

Lecture Notes in Networks and Systems Volume 616

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, Türkiye 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]).

Rajiv Misra · Muttukrishnan Rajarajan · Bharadwaj Veeravalli · Nishtha Kesswani · Ashok Patel Editors

Internet of Things (IoT): Key Digital Trends Shaping the Future Proceedings of 7th International Conference on Internet of Things and Connected Technologies (ICIoTCT 2022)

Editors Rajiv Misra Department of Computer Science and Engineering Indian Institute of Technology Patna Bihta, Bihar, India Computer Science, Central University of Rajasthan Kishangarh (Ajmer), India

Muttukrishnan Rajarajan City University of London London, UK Nishtha Kesswani Department of Computer Science Central University of Rajasthan Ajmer, Rajasthan, India

Bharadwaj Veeravalli Department of Electrical and Computer Engineering National University of Singapore Singapore, Singapore Ashok Patel Department of Computer Science Florida Polytechnic University Lakeland, FL, USA

ISSN 2367-3370 ISSN 2367-3389 (electronic) Lecture Notes in Networks and Systems ISBN 978-981-19-9718-1 ISBN 978-981-19-9719-8 (eBook) https://doi.org/10.1007/978-981-19-9719-8 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 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 Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Organization

Program Committee Chairs Misra, Rajiv (2022, ICCM), Central University of Rajasthan, Computer Science, Kishangarh (Ajmer), India

Program Committee Members Dr. E. Arul (2022, ICCM), Coimbatore Institute of Technology, Information Technology, Coimbatore, India; Central University of Rajasthan, Computer Science, Kishangarh (Ajmer), India Dr. A. Akila, Indian Institute of Technology Guwahati, Electronics and Electrical Engineering, Guwahati, India Shaymaa Amer Abdul Kareem, Indian Institute of Technology Guwahati, Electronics and Electrical Engineering, Guwahati, India Raktim Acharjee, Indian Institute of Technology Guwahati, Electronics and Electrical Engineering, Guwahati, India Vishali Aggarwal, Computer Science and Engineering, Indian Institute of Information Technology Bhagalpur, Bihar, India Ghazi Alkhatib, Computer Science and Engineering, Indian Institute of Information Technology Bhagalpur, Bihar, India B. R. Arunkumar, Computer Science and Engineering, Indian Institute of Information Technology Bhagalpur, Bihar, India C. Siva Balaramudu, Computer Science and Engineering, Indian Institute of Information Technology Bhagalpur, Bihar, India Deepshikha Bhatia, Computer Science and Engineering, Indian Institute of Information Technology Bhagalpur, Bihar, India Aninda Bose, Computer Science and Engineering, Indian Institute of Information Technology Bhagalpur, Bihar, India

v

vi

Organization

Anupama Chadha, Computer Science and Engineering, Indian Institute of Information Technology Bhagalpur, Bihar, India Dr. Dilip Kumar Choubey, Computer Science and Engineering, Indian Institute of Information Technology Bhagalpur, Bihar, India Bhushan Deore, Kalaignar Karunanidhi Institute of Technology, Aeronautical Engineering, Coimbatore, India Mohammad Faiz, Kalaignar Karunanidhi Institute of Technology, Aeronautical Engineering, Coimbatore, India G. A. Sivasankar Kit, Kalaignar Karunanidhi Institute of Technology, Aeronautical Engineering, Coimbatore, India Argha Ghosh, Suresh Gyan Vihar University Jaipur Rajasthan, CSE, Jaipur, India M. D. Rajib Hossain, Suresh Gyan Vihar University Jaipur Rajasthan, CSE, Jaipur, India Hussain, Mohammad Equebal, Suresh Gyan Vihar University Jaipur Rajasthan, CSE, Jaipur, India Rikhi Ram Jagat, National Institute of Technology Raipur, Computer Science and Engineering, Raipur, India Nikita Jain, Tripura University, Information Technology, Agartala, India Shruti Jain, Tripura University, Information Technology, Agartala, India Shelendra Jain, Tripura University, Information Technology, Agartala, India Dr. Kalaiselvi K., Tripura University, Information Technology, Agartala, India Haribabu Kotakula, Tripura University, Information Technology, Agartala, India Hananya Kampa, Tripura University, Information Technology, Agartala, India Prema Kirubakaran, Tripura University, Information Technology, Agartala, India Siva Krishna, Tripura University, Information Technology, Agartala, India Ashish Kumar, Tripura University, Information Technology, Agartala, India Vinay Kumar, Tripura University, Information Technology, Agartala, India Thangavel M., Tripura University, Information Technology, Agartala, India Anil M. A, Tripura University, Information Technology, Agartala, India Majumder, Swanirbhar, Tripura University, Information Technology, Agartala, India Priyanka Mishra, Sanjivani College of Engineering, Kopargaon, Information Technology, Kopargaon, India Rajiv Misra, Sanjivani College of Engineering, Kopargaon, Information Technology, Kopargaon, India Prasad Mutkule, Sanjivani College of Engineering, Kopargaon, Information Technology, Kopargaon, India N. A. Aishwarya , ViMEET, Computer Science and Engineering (AI & ML), Raigad, India P. Appala Naidu, ViMEET, Computer Science and Engineering (AI & ML), Raigad, India Dr. Surya Kant Pal, ViMEET, Computer Science and Engineering (AI & ML), Raigad, India Shashikant Patil, ViMEET, Computer Science and Engineering (AI & ML), Raigad, India Rathidevi R. Rajendran, Sri Vasavi Engineering College, ECE, tadepalligudem, India

Organization

vii

Amita Sharma, Sri Vasavi Engineering College, ECE, Tadepalligudem, India Dr. Pooja Sapra, Sri Vasavi Engineering College, ECE, Tadepalligudem, India Vinita Shah, Sri Vasavi Engineering College, ECE, tadepalligudem, India Purnima K. Sharma, Sri Vasavi Engineering College, ECE, Tadepalligudem, India Dr. Santosh Kumar Sharma, United University, Computer Science and Engineering, Prayagraj, India Vinay Singh, Vivekananda Institute of Professional Studies—Technical Campus, Affiliated to GGSIPU, Delhi, Information Technology, Delhi, India Dr. Shweta Taneja, Vivekananda Institute of Professional Studies—Technical Campus, Affiliated to GGSIPU, Delhi, Information Technology, Delhi, India Pooja Thakar, Vivekananda Institute of Professional Studies—Technical Campus, Affiliated to GGSIPU, Delhi, Information Technology, Delhi, India Ashish Tiwari, NIT Kurukshetra, CSE, Kurukshetra, India Jaykumar Vala, Vivekananda Institute of Professional Studies–Technical Campus, Delhi, India Jitendra Kumar Verma, Vivekananda Institute of Professional Studies–Technical Campus, Delhi, India Aditya Verman, Vivekananda Institute of Professional Studies–Technical Campus, Delhi, India Deepali Virmani, Vivekananda Institute of Professional Studies–Technical Campus, Delhi, India Anuj Yadav Bhanu Chander Vishal Khand Sumit Sar Sonakshi Vij

Reviewers Dr. E. Arul, Coimbatore Institute of Technology, Information Technology, Coimbatore, India Shaymaa Amer Abdul Kareem, Indian Institute of Technology Guwahati, Electronics and Electrical Engineering, Guwahati, India Raktim Acharjee, Indian Institute of Technology Guwahati, Electronics and Electrical Engineering, Guwahati, India Ghazi Alkhatib, Computer Science and Engineering, Indian Institute of Information Technology Bhagalpur, Bihar, India C. Siva Balaramudu, Computer Science and Engineering, Indian Institute of Information Technology Bhagalpur, Bihar, India Deepshikha Bhatia, Computer Science and Engineering, Indian Institute of Information Technology Bhagalpur, Bihar, India Anupama Chadha, Computer Science and Engineering, Indian Institute of Information Technology Bhagalpur, Bihar, India

viii

Organization

Dr. Dilip Kumar Choubey, Computer Science and Engineering, Indian Institute of Information Technology Bhagalpur, Bihar, India G. A. Sivasankar, KIT Kalaignar Karunanidhi Institute of Technology, Aeronautical Engineering, Coimbatore, India Argha Ghosh, National Institute of Technology Raipur, Computer Science and Engineering, Raipur, India Rikhi Ram Jagat, National Institute of Technology Raipur, Computer Science and Engineering, Raipur, India Nikita Jain, Tripura University, Information Technology, Agartala, India Shruti Jain, Tripura University, Information Technology, Agartala, India Shelendra Jain, Tripura University, Information Technology, Agartala, India Haribabu Kotakula, Tripura University, Information Technology, Agartala, India Hananya Kampa, Tripura University, Information Technology, Agartala, India Prema Kirubakaran, Tripura University, Information Technology, Agartala, India Siva Krishna, Tripura University, Information Technology, Agartala, India M. Thangavel, Tripura University, Information Technology, Agartala, India M. A. Anil, Tripura University, Information Technology, Agartala, India Swanirbhar Majumder, Tripura University, Information Technology, Agartala, India Priyanka Mishra, Sanjivani College of Engineering, Kopargaon, Information Technology, Kopargaon, India Prasad Mutkule, Sanjivani College of Engineering, Kopargaon, Information Technology, Kopargaon, India N. A. Aishwarya, ViMEET, Computer Science and Engineering (AI & ML), Raigad, India P. Appala Naidu, ViMEET, Computer Science and Engineering (AI & ML), Raigad, India Shashikant Patil, ViMEET, Computer Science and Engineering (AI & ML), Raigad, India Rathidevi R. Rajendran, Sri Vasavi Engineering College, ECE, Tadepalligudem, India Amita Sharma, Sri Vasavi Engineering College, ECE, Tadepalligudem, India Dr. Pooja Sapra, Sri Vasavi Engineering College, ECE, Tadepalligudem, India Shah Vinita, Sri Vasavi Engineering College, ECE, Tadepalligudem, India Purnima K. Sharma, Sri Vasavi Engineering College, ECE, Tadepalligudem, India Vinay Singh, Vivekananda Institute of Professional Studies–Technical Campus, Affiliated to GGSIPU, Delhi, Information Technology, Delhi, India Dr. Shweta Taneja, Vivekananda Institute of Professional Studies–Technical Campus, Affiliated to GGSIPU, Delhi, Information Technology, Delhi, India Pooja Thakar, Vivekananda Institute of Professional Studies–Technical Campus, Affiliated to GGSIPU, Delhi, Information Technology, Delhi, India Ashish Tiwari, NIT Kurukshetra, CSE, Kurukshetra, India Jaykumar Vala, Vivekananda Institute of Professional Studies–Technical Campus, Delhi, India Jitendra Kumar Verma, Vivekananda Institute of Professional Studies–Technical Campus, Delhi, India

Organization

ix

Aditya Verman, Vivekananda Institute of Professional Studies–Technical Campus, Delhi, India Deepali Virmani, Vivekananda Institute of Professional Studies–Technical Campus, Delhi, India Vishal Khand Sumit Sar Sonakshi Vij

Preface

The 7th International Conference on Internet of Things and Connected Technologies (ICIoTCT) 2022 presents key ingredients for the 5th Generation Revolution. The recent adoption of a variety of enabling Wireless communication technologies such as RFID tags, BLE, ZigBee, etc. and embedded sensor and actuator nodes, and various protocols such as CoAP, MQTT, DNS etc. have made IoT step out of its infancy. The ICIOTCT 2022 was organized on Sept 29–30, 2022 by the Indian Institute of Technology, Patna, VKONEX (India) in collaboration with the International Association of Academicians (IAASSE) USA and it provided a platform to discuss advances in the Internet of Things (IoT) and connected technologies (various protocols, standards etc.). Bihta, India London, UK Singapore, Singapore Ajmer, India Lakeland, USA 2022, ICCM

Rajiv Misra Muttukrishnan Rajarajan Bharadwaj Veeravalli Nishtha Kesswani Ashok Patel

xi

Contents

An Efficacious Classifier for Recognition of Traffic Symbols . . . . . . . . . . . Deepali Virmani and Ketan Parikh Assisted Living Robots: Discussion and Design of a Robot for Elder Care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Garvita Ahuja, Shivansh Sharma, Maanik Sharma, and Srishti Singh Localization Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anshika Jain, Bhumika Jain, Arohi Singhal, and Srishti Singh A Comparative Study Between Various Machine-Learning Algorithms Implemented for the Proper Detection of Fraudulent and Non-fraudulent Transactions Through Credit Card . . . . . . . . . . . . . . . Surya Kant Pal, Nazneen Alam, Rita Roy, Preeti Jawla, and Subhodeep Mukherjee Smart Grid and Energy Management System . . . . . . . . . . . . . . . . . . . . . . . . Ishan Sharma, Priyal, Ananya Tyagi, Radhika Chawla, Aditya Khazanchi, Aaryan Bhatia, and Srishti Singh To Foresight and Formulate Development (FFD) of Robot of Things (RoT) and Drone of Things (DoT) for Revolutionizing Agriculture Ecosystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chandrani Singh, Sunil Khilari, and Anchal Koshta About a Practical Approach for Smart Building by Using Internet of Things . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kumudini Manwar, Dushyant Bodkhey, Chandrani Singh, Girish Mogalgiddikar, and Pratiksha Mahamine IoT-Based Storage Management System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Milind Godase, Chandrani Singh, and Akshay Tanpure

1

11 27

39

49

63

77

89

xiii

xiv

Contents

Venture Analyzer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Aditi Bhole, Anshuta Kakuste, Sudiksha Mullick, Rakhi Kalantri, and Shagufta Rajguru Stress Reliving Application for Personal Wellbeing . . . . . . . . . . . . . . . . . . . 113 Aaryan Rastogi, Nidhi Shrivastav, Atharva Suryavanshi, Palak Wadhwa, Rakhi Kalantri, and R. Shagufta M-Lens an IOT-Based Deep Learning Device . . . . . . . . . . . . . . . . . . . . . . . . . 123 Dheeraj Kallakuri, Nikhil Londhe, Sharon Laurance, Vinayak Kurup, and Rakhi Kalantri Analysis of Electromagnetic Pollution in Buildings and Its Impact Specially on Human Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 Pallav Dutta and Rumpa Saha IoT-Based Smart Notice Board & Class Schedule Notification System with Real-Time Classroom Environment Monitoring Facility for Educational Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 Ashim Mondal, Pallav Dutta, and Rumpa Saha Affordable Smart Kit for Coconut Farm Management Using IoT . . . . . . 165 S. Sri Sankar, S. Viswesh, T. Ramya, and G. Balasubramanian Machine Learning Based Model to Find Out Firewall Decisions Towards Improving Cyber Defence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 Madhab Paul Choudhury and J. Paul Choudhury Two Fold Extended Residual Network Based Super Resolution for Potato Plant Leaf Disease Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197 P. V. Yeswanth, Rachit Khandelwal, and S. Deivalakshmi Analyzing the Tweets of the Patients During the COVID-19 Pandemic Using Machine Learning Techniques . . . . . . . . . . . . . . . . . . . . . . 211 Routhu Shanmukh, Rita Roy, Kavitha Chekuri, Rowthu Lakshmana Rao, and Subhodeep Mukherjee Load Profile Oriented Balanced Cluster Assignment in 5G IoT Based Sensor Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 B. Dey, Sivaji Bandyopadhyay, and Sukumar Nandi KnowSOntoWSR: Web Service Recommendation System Using Semantically Driven QoS Ontology-Based Knowledge-Centred Paradigm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233 R. Dhanvardini, Gerard Deepak, J. Sheeba Priyadarshini, and A. Santhanavijayan Secure Encryption Using Bit Shuffling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 Uday Kumar Banerjee, Anup Kumar Das, Rajdeep Ray, and Chandan Koner

Contents

xv

Predictability of Spells of Maximum Precipitation in the UP East Region with Antarctic Sea Ice Concentration Forcing . . . . . . . . . . . . . . . . . 255 Rashi Aggarwal, Manpreet Kaur, and K. C. Tripathi IoT-Based Real-Time Water Quality Monitoring System Using a RC Boat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269 Utkarsh Asari, Raj Desai, Rutu Parekh, and Udit Meena Tamil Language Automatic Speech Recognition Based on Integrated Feature Extraction and Hybrid Deep Learning Model . . . 283 Akanksha Akanksha A Comprehensive Review of Conversational AI-Based Chatbots: Types, Applications, and Future Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293 M. Vishal and H. Vishalakshi Prabhu Blockchain Based Tourism Recommender . . . . . . . . . . . . . . . . . . . . . . . . . . . 305 M. Aneerudh, S. Shane Rex, and M. Vijayalakshmi Quantum-defended Digital Signature on Lattice for IoT-enabled Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315 Daya Sagar Gupta, Lacchita Soni, and Harish Chandra Securing Digital Ownership Using Non-Fungible Tokens(NFTs), an Application of BlockChain Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327 Suhas Harbola, Jyotsna Yadav, Rahul Johari, Ekta Verma, and Deo Prakash Vidyarthi Battery Optimization of Electric Vehicles Using Battery Management System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339 Simran Khanna, Vansh Bhandari, Tanmay Mishra, Yash Shrivastav Yashas Bajaj, and Srishti Singh Electronic Voting Machine as a Service on the Cloud—Azure for EVM (A4EVM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353 Mohammad Equebal Hussain, Mukesh Kumar Gupta, and Rashid Hussain Internet of Bio-nano Things for Diabetes Telemedicine System with Secured Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365 Lokavya Gabrani, Rajeev Kumar Singh, Sonali Vyas, Sunil Gupta, and Goldie Gabrani Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375

Editors and Contributors

About the Editors Rajiv Misra is an Associate Professor of Computer Science and Engineering at the Indian Institute of Technology Patna, India. His research focuses in distributed systems, cloud computing, big data computing, consensus in blockchain, cloud IoTedge computing, ad hoc networks, and sensor networks. He has contributed significantly to these research areas of distributed and cloud computing and published more than 80 papers in reputed journals and conferences, with an impact of 999 citations and an h-index of 14. Muttukrishnan Rajarajan is currently the Director of the Institute for Cyber Security at City University of London and carries out research in the areas of privacy preserving data management, Internet of Things privacy, network intrusion detection, cloud security and identity management using blockchain. Raj has received funding from EPSRC, Royal Academy of Engineering, European Commission, Innovate UK, British Council and industry to carry out research in cyber security. He has supervised several Ph.Ds. jointly with British Telecommunications, UK in the area of data analytics for cyber security and network intrusion detection. Bharadwaj Veeravalli is currently with the Department of Electrical and Computer Engineering, Communications and Information Engineering (CIE) division, at The National University of Singapore, Singapore. His main stream research interests include cloud/grid/cluster computing (big data processing, analytics and resource allocation), scheduling in parallel and distributed systems, Cybersecurity, and multimedia computing. He is one of the earliest researchers in the field of Divisible Load Theory (DLT). He did Ph.D. degree from the Indian Institute of Science, Bangalore, India. He received gold medals for his bachelor degree overall performance and for an outstanding Ph.D. thesis (IISc, Bangalore India) in the years 1987 and 1994, respectively.

xvii

xviii

Editors and Contributors

Nishtha Kesswani has received prestigious awards, including the UGC Raman Postdoctoral Fellowship tenable in USA and the Young Teacher Award. She received the M.Tech. degree from the Malaviya National Institute of Technology (MNIT). She has a vivid teaching experience at several reputed universities, including California State University at San Bernardino and the University of Ljubljana, Slovenia. She has visited more than 15 countries and delivered invited talks at several conferences and workshops. She is currently with the Central University of Rajasthan, India. Ashok Patel is a faculty at computer and information science, at UMass Dartmouth USA. Before joining UMass, he taught at the Department of Computer Science of Florida Polytechnic University, USA. He primarily teaches cybersecurity courses and is researching an improved efficient fingerprint recognition algorithm and web usage mining. He’s particularly interested in personalizing the web experience for users and individuals using IoT. He has nearly 30 years of teaching experience. Before immigrating to the United States, he was a professor in the Department of Computer Science at North Gujarat University in India.

Contributors Rashi Aggarwal Maharaja Agrasen Institute of Management Studies, Delhi, India; Department of CSE, Manav Rachna University, Gurugram, India Garvita Ahuja Vivekananda Institute of Professional Studies-Technical Campus, Delhi, India Akanksha Akanksha Computer Science, National Institute of Technology, Kurukshetra, Haryana, India Nazneen Alam Department of Mathematics, School of Basic Sciences and Research, Sharda University, Greater Noida, India M. Aneerudh Department of Computer Science and Engineering, Thiagarajar College of Engineering, Madurai, India Utkarsh Asari Dhirubhai Ambani Institue of Information & Communication Technology, Gandhinagar, Gujarat, India G. Balasubramanian School of Electrical & Electronics Engineering, SASTRA Deemed University, Thanjavur, Tamil Nadu, India Sivaji Bandyopadhyay NIT Silchar, Silchar, Assam, India Uday Kumar Banerjee Dr. B. C. Roy Engineering College/MCA/ECE, Durgapur, India Vansh Bhandari Vivekananda Institute of Professional Studies—Technical Campus, New Delhi, India

Editors and Contributors

xix

Aaryan Bhatia Technical Campus, Vivekananda Institute of Professional Studies, Delhi, India Aditi Bhole Department of Computer Engineering, Fr. C. Rodrigues Institute of Technology, Vashi, Maharashtra, India Dushyant Bodkhey Sinhgad Institute of Management, Pune, India Harish Chandra Department of Mathematics and Scientific Computing, Madan Mohan Malaviya University of Technology, Gorakhpur, UP, India Radhika Chawla Technical Campus, Vivekananda Institute of Professional Studies, Delhi, India Kavitha Chekuri Department of Computer Science and Engineering, Raghu Engineering College, Visakhapatnam, Andhra Pradesh, India J. Paul Choudhury Kalyani Government Engineering College, Kalyani, Narula Institute of Technology, Kolkata, India Madhab Paul Choudhury NIT Jamshedpur, Jamshedpur, Jharkhand, India Anup Kumar Das Dr. B. C. Roy Engineering College/MCA/ECE, Durgapur, India Gerard Deepak Department of Computer Science Engineering, National Institute of Technology, Tiruchirappalli, India S. Deivalakshmi Department of Electronics and Communication Engineering, National Institute of Technology Tiruchirappalli, Tiruchirappalli, India Raj Desai Dhirubhai Ambani Institue of Information & Communication Technology, Gandhinagar, Gujarat, India B. Dey NIT Silchar, Silchar, Assam, India R. Dhanvardini Healthcare Informatics Division, Optum UnitedHealth Groups, Hyderabad, India Pallav Dutta Electrical Engineering Department, Aliah University, Kolkata, India Goldie Gabrani Vivekananda Institute of Professional Studies, New Delhi, India Lokavya Gabrani Department of Computer Science and Engineering, Shiv Nadar University, Noida, India Milind Godase Sinhgad Institute of Management, Pune, India Daya Sagar Gupta School of Management, Indian Institute of Technology Mandi, Kamand, Mandi, Himachal Pradesh, India Mukesh Kumar Gupta Suresh Gyan Vihar University, Jaipur, India Sunil Gupta School of Computer Science, UPES, Dehradun, Uttarakhand, India

xx

Editors and Contributors

Suhas Harbola Computer Vision and Image Processing lab, University School of Information, Communication and Technology (USICT), Guru Gobind Singh Indraprastha University, Dwarka, Delhi, India; National Informatics Centre, New Delhi, India Mohammad Equebal Hussain Suresh Gyan Vihar University, Jaipur, India Rashid Hussain Moti Babu Institute of Technology, Bihar, India Anshika Jain Vivekananda Institute of Professional Studies—Technical Campus, New Delhi, India Bhumika Jain Vivekananda Institute of Professional Studies—Technical Campus, New Delhi, India Preeti Jawla IIMT College of Engineering, Greater Noida, India Rahul Johari SWINGER : Security, Wireless, IoT Network Group of Engineering and Research, University School of Information, Communication and Technology (USICT), Guru Gobind Singh Indraprastha University, Dwarka, Delhi, India Anshuta Kakuste Department of Computer Engineering, Fr. C. Rodrigues Institute of Technology, Vashi, Maharashtra, India Rakhi Kalantri Department of Computer Engineering, Fr. Conceicao Rodrigues Institute of Technology, Vashi, Navi Mumbai, Maharashtra, India Dheeraj Kallakuri Department of Computer Engineering, Fr. C. Rodrigues Institute of Technology, Vashi, Navi Mumbai, India Manpreet Kaur Department of CSE, Manav Rachna University, Gurugram, India Rachit Khandelwal Department of Electronics and Communication Engineering, National Institute of Technology Tiruchirappalli, Tiruchirappalli, India Simran Khanna Vivekananda Institute of Professional Studies—Technical Campus, New Delhi, India Aditya Khazanchi Technical Campus, Vivekananda Institute of Professional Studies, Delhi, India Sunil Khilari Sinhgad Institute of Management, Pune, India Chandan Koner Dr. B. C. Roy Engineering College/ECE/CSE, Durgapur, India Anchal Koshta ABCROB Technologies Pvt Ltd, Jabalpur, M.P, India Vinayak Kurup Department of Computer Engineering, Fr. C. Rodrigues Institute of Technology, Vashi, Navi Mumbai, India Sharon Laurance Department of Computer Engineering, Fr. C. Rodrigues Institute of Technology, Vashi, Navi Mumbai, India

Editors and Contributors

xxi

Nikhil Londhe Department of Computer Engineering, Fr. C. Rodrigues Institute of Technology, Vashi, Navi Mumbai, India Pratiksha Mahamine Sinhgad Institute of Management, Pune, India Kumudini Manwar Sinhgad Institute of Management, Pune, India Udit Meena Dhirubhai Ambani Institue of Information & Communication Technology, Gandhinagar, Gujarat, India Tanmay Mishra Vivekananda Institute of Professional Studies—Technical Campus, New Delhi, India Girish Mogalgiddikar GM Soft Pvt Limited, Pune, India Ashim Mondal Electrical Engineering Department, Aliah University, Kolkata, India Subhodeep Mukherjee Department of Management, GITAM (Deemed to Be University), Visakhapatnam, Andhra Pradesh, India; Department of Operations, GITAM SCHOOL OF BUSINESS, GITAM (Deemed to be University), Visakhapatnam, Andhra Pradesh, India Sudiksha Mullick Department of Computer Engineering, Fr. C. Rodrigues Institute of Technology, Vashi, Maharashtra, India Sukumar Nandi IIT Guwahati, North Guwahati, Assam, India Surya Kant Pal Department of Mathematics, School of Basic Sciences and Research, Sharda University, Greater Noida, India Rutu Parekh Dhirubhai Ambani Institue of Information & Communication Technology, Gandhinagar, Gujarat, India Ketan Parikh Department of Computer Science Engineering, Bhagwan Parshuram Institute of Technology, New Delhi, India J. Sheeba Priyadarshini Department of Data Science, Manipal Institute of Technology Bengaluru, Bengaluru, India; Department of Data Science, Manipal Academy of Higher Education, Manipal, India; CHRIST (Deemed to Be University), Bangalore, India Priyal Technical Campus, Vivekananda Institute of Professional Studies, Delhi, India Shagufta Rajguru Department of Computer Engineering, Fr. C. Rodrigues Institute of Technology, Vashi, Maharashtra, India T. Ramya School of Electrical & Electronics Engineering, SASTRA Deemed University, Thanjavur, Tamil Nadu, India

xxii

Editors and Contributors

Rowthu Lakshmana Rao Department of Computer Science and Engineering, Centurion University of Technology and Management, Vizianagaram, Andhra Pradesh, India Aaryan Rastogi Department of Computer Engineering, Fr Conceicao Rodrigues Institute of Technology, Vashi Navi, Mumbai, India Rajdeep Ray Dr. B. C. Roy Engineering College/ECE/CSE, Durgapur, India S. Shane Rex Department of Computer Science and Engineering, Thiagarajar College of Engineering, Madurai, India Rita Roy Department of Computer Science and Engineering, GITAM Institute of Technology, GITAM (Deemed to be University), Visakhapatnam, Andhra Pradesh, India Rumpa Saha Electrical Engineering Department, Aliah University, Kolkata, India A. Santhanavijayan Department of Computer Science Engineering, National Institute of Technology, Tiruchirappalli, India R. Shagufta Department of Computer Engineering, Fr Conceicao Rodrigues Institute of Technology, Vashi Navi, Mumbai, India Routhu Shanmukh Department of Computer Science and Engineering, Centurion University of Technology and Management, Vizianagaram, Andhra Pradesh, India Ishan Sharma Technical Campus, Vivekananda Institute of Professional Studies, Delhi, India Maanik Sharma Vivekananda Institute of Professional Studies-Technical Campus, Delhi, India Shivansh Sharma Vivekananda Institute of Professional Studies-Technical Campus, Delhi, India Yash Shrivastav Yashas Bajaj Vivekananda Institute of Professional Studies— Technical Campus, New Delhi, India Nidhi Shrivastav Department of Computer Engineering, Fr Conceicao Rodrigues Institute of Technology, Vashi Navi, Mumbai, India Arohi Singhal Vivekananda Institute of Professional Studies—Technical Campus, New Delhi, India Chandrani Singh Sinhgad Institute of Management, Pune, India Rajeev Kumar Singh Department of Computer Science and Engineering, Shiv Nadar University, Noida, India Srishti Singh Technical Campus, Vivekananda Institute of Professional Studies, Delhi, India

Editors and Contributors

xxiii

Lacchita Soni Department of Mathematics and Scientific Computing, Madan Mohan Malaviya University of Technology, Gorakhpur, UP, India S. Sri Sankar School of Electrical & Electronics Engineering, SASTRA Deemed University, Thanjavur, Tamil Nadu, India Atharva Suryavanshi Department of Computer Engineering, Fr Conceicao Rodrigues Institute of Technology, Vashi Navi, Mumbai, India Akshay Tanpure Sinhgad Institute of Management, Pune, India K. C. Tripathi Department of IT, Maharaja Agrasen Institute of Technology, Delhi, India Ananya Tyagi Technical Campus, Vivekananda Institute of Professional Studies, Delhi, India Ekta Verma National Informatics Centre, New Delhi, India Deo Prakash Vidyarthi School of Computer and Systems Sciences, Parallel and Distributed System Lab JNU, Delhi, India M. Vijayalakshmi Department of Computer Science and Engineering, Thiagarajar College of Engineering, Madurai, India Deepali Virmani School of Engineering & Technology, Vivekananda Institute of Professional Studies-Technical Campus, New Delhi, India M. Vishal Department of Computer Science and Engineering, R.V. College of Engineering, Bengaluru, Karnataka, India H. Vishalakshi Prabhu Department of Computer Science and Engineering, R.V. College of Engineering, Bengaluru, Karnataka, India S. Viswesh School of Electrical & Electronics Engineering, SASTRA Deemed University, Thanjavur, Tamil Nadu, India Sonali Vyas School of Computer Science, UPES, Dehradun, Uttarakhand, India Palak Wadhwa Department of Computer Engineering, Fr Conceicao Rodrigues Institute of Technology, Vashi Navi, Mumbai, India Jyotsna Yadav Computer Vision and Image Processing lab, University School of Information, Communication and Technology (USICT), Guru Gobind Singh Indraprastha University, Dwarka, Delhi, India P. V. Yeswanth Department of Electronics and Communication Engineering, National Institute of Technology Tiruchirappalli, Tiruchirappalli, India

An Efficacious Classifier for Recognition of Traffic Symbols Deepali Virmani and Ketan Parikh

Abstract Information gathering refers to collecting information from different sources in order to process it to obtain useful information. In particular, image classification refers to the use of several computer algorithms to classify images to extract useful information and as to what they symbolize. It is one of the deep learning methods. Traffic symbols are important to be noticed while driving any vehicle. This paper proposes an image classifier to classify and detect traffic symbols from the dataset of traffic symbols. The proposed algorithm for image classification provides an efficient and accurate way to classify the traffic signal images into classes and the obtained results have great precision. This image classifier can detect the traffic symbols in real time with an accuracy of 88.76%. Various datasets can be chosen to perform experiments with the resulting image classifier. Keywords Traffic symbols · Image classifier · TensorFlow · Deep learning

1 Introduction Road accidents are increasing day by day. Most of the accidents that occur are either due to human error or negligence. Following statistics on annual global road crash as observed from [1] shed a light on how dangerous road accidents are: approximately 1.3 million people die in road crashes each year. Further 20–50 million are injured or disabled. More than half of all road traffic deaths occur among young adults. Road traffic crashes rank as the 9th leading cause of death and account for many deaths. Each year nearly 400,000 people under 25 die on the world’s roads. Over 90% of all D. Virmani (B) School of Engineering & Technology, Vivekananda Institute of Professional Studies-Technical Campus, New Delhi, India e-mail: [email protected]; [email protected] K. Parikh Department of Computer Science Engineering, Bhagwan Parshuram Institute of Technology, New Delhi, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. Misra et al. (eds.), Internet of Things (IoT): Key Digital Trends Shaping the Future, Lecture Notes in Networks and Systems 616, https://doi.org/10.1007/978-981-19-9719-8_1

1

2

D. Virmani and K. Parikh

road fatalities occur in low- and middle-income countries. Road crashes cost USD $518 billion globally. Road crashes cost low- and middle-income countries USD $65 billion annually. If necessary, actions are not taken, road traffic injuries are predicted to become the fifth leading cause of death by 2030. At various places on the road, sign boards are set up to make roads safer and inform the driver and pedestrians about the surroundings, for example, a speed-breaker on the road or the maximum speed limit on the road, all are indicated to the driver through sign boards. But due to many factors, the driver is not always able to read these signs. Knowing the signs and gaining information from that, at by large, can reduce the road accidents that happen every year. This paper pro poses to develop a system that would help a car driver gather all the useful information from the sign boards. Hence, this information can make roads safer and reduce car accidents by a great margin. The system consists of a camera that would gather input about the outside world. This input is then fed to image processing algorithms that try to detect if any sign boards are approaching. The sign board is detected via complex and sophisticated image recognition and classification algorithms. Then from the sign board we try to find out the type of symbol present in the sign board. Machine Learning helps a lot in recognizing the signs in the sign boards. Then when we get a match of a sign, we translate that sign into voice signals via natural language processing and give it as output to the driver of the car. The proposed algorithm will solve a great problem with high accuracy and precision. It is very important that the roads are safe to travel as traveling is a necessary activity in people’s day-to-day life. With this system, roads can be a better and safe place as it will help in reducing the total number of road accidents.

2 Related Work This section focuses on the ongoing work in the field of image processing and analysis. The major mechanisms covered are as follows—TensorFlow, the inception model, image processing and classification, and comparison between the two most popular libraries used in machine learning and computer vision, i.e., TensorFlow and OpenCV. TensorFlow TensorFlow is an open-source software library used for programming of the flow of data across a variety of tasks. With reference to [2], it is a representative mathematics library and is used to implement various machine learning applications. It is developed by the Google Brain team who has released stable versions though is still under development and is an emerging concept in machine intelligence. As seen from [3], one of the applications of TensorFlow is the automated image captioning software. The image captioning software uses the image captioning model which is a machine learning system to automatically produce captions that precisely describe

An Efficacious Classifier for Recognition of Traffic Symbols

3

the images. TensorFlow can be highly recommended for the accurate description and processing of images producing efficient results. The Inception Model The Inception models are the ones which are used as image classification models and can be used for a variety of purposes. It can be observed from [4] that Google has produced various image classification models with different versions referred to as the inception model with their corresponding version number. These image classification models use image encoders to achieve accuracy in the results. Various inception models developed are Inception V1, Inception V2, Inception V3, and Inception V4. This model was able to classify the objects in images with an accuracy of 89.6%. With reference to [5], this inception model was used along with two other models to produce a fusion model for image classification. The Inception V2 image classification model came next in 2015 with an increased accuracy of 91.8% on the same task of encodings for images to produce captions. The most popularly used image classification model is the Inception V3 model with an accuracy of 93.9% on the ImageNet classification task which has the highest accuracy of all the three inception models. Yim et al. [6] uses the Inception V3 model to classify numerous plant images incorporated with the convolutional neural networks. They were able to achieve an accuracy of 98% with reference to the chosen dataset. Also, from [7], it is observed that a modified version of the Inception v3 model can be used to classify multi-label images with a better accuracy as compared to the logistic regression methods. Lately, there has been ongoing research on the Inception V4 image classification model. From [8], it can be acknowledged that Inception V4 model’s recognition performance is same as that of Inception-ResNet-v2 without residual connections chosen as the criteria. Image Classification With reference to [9], image classification can be considered as a type of deep learning method which can help in gathering information from multiple images and be used for further analysis. Nowadays, image classification is being widely used to extract information from images. As said by [10], contextual image classification is a type of pattern acknowledgment in computer vision. It is a method of classification based on related information in images. This approach highlights the relationship of nearby pixels, called neighborhood. The goal of this approach is to classify the images by using the contextual information. Techniques through which image classification can be implemented are • Artificial neural network: It uses a non-parametric approach. The accuracy and performance depend upon the network structure and number of inputs. Advantage of using ANN is that it is a universal functional approximator and is a data-driven self-adaptive technique.

4

D. Virmani and K. Parikh

• Decision tree: Decision trees are based on hierarchical rule-based method and use non-parametric approach. DT is used because it can handle non-parametric training data and does not require an extensive design and training. • Support vector machine: SVM uses binary classifier approach and can handle more input data very efficiently. Performance and accuracy depend upon the hyperplane selection and kernel parameter. SVM provides a good generalization capability and is simple to manage decision rule complexity and error frequency. A modified version of the support vector machine can also prove to be efficient as observed in [11].

3 Proposed Work This section focuses on choosing the required image dataset and performing image classification. The dataset has been created for traffic signals. Various traffic signals, such as no parking, no horn, and danger, have been incorporated to create datasets. This image dataset of traffic symbols and signals is converted into .pb file in order to check the execution accuracy of the TensorFlow in case of image detection. Different varieties of images are collected in order to improve the object recognition precision. This dataset is created by using Docker and created into tensors which implements convolutional neural network based on edge detection, shape detection, and color detection. With reference to [12], it can be observed that convolutional neural networks can be used for classifying very large number of images. Docker is a software technology which provides containers with layers of abstraction for virtualizations mentioned in [13]. This implementation is done in three steps.

3.1 Data Preprocessing This section focuses on creation of image dataset used for training dataset. Datasets for various traffic symbols are collected, i.e., batch download which is used as a training dataset for creating a retraining graph for classification [14]. 200 images per each symbol are used for particular image dataset. These images can also be obtained from [15, 16]. Figure 1 and Fig. 2 show the datasets used for no horn and no parking traffic symbols, respectively. Figure 3 shows the control flow for the traffic signal image classification using Tensor Flow image classifier.

An Efficacious Classifier for Recognition of Traffic Symbols

5

Fig. 1 Traffic signal dataset for no horn sign

Fig. 2 Traffic signal dataset for no parking sign

3.2 Creating Classifier TensorFlow is used to create a retrained graph as a classifier which results in predictions. This is implemented using Docker and using Inception model. This model can be extensively used in video object detection as well as hardware for controlling traffic.

6

D. Virmani and K. Parikh

Fig. 3 Control flow of the traffic signal image classifier

3.3 Testing Data Images for particular traffic symbols are tested against a trained classifier, i.e., retrained graph created and gives predictions in the form sorted confidence value. The proposed algorithm consists of building an image classifier for a dataset of traffic symbols which is depicted in Fig. 4. After creating the image dataset for traffic symbols and signals, a container is initialized using Docker. The inception model is downloaded and converted into a .pb file and finally into a .py file. Now, this image is read with the TensorFlow in order to calculate predictions corresponding to the traffic symbol image. So, the confidence values are generated for each type of available traffic symbol in the dataset. Hence, results with corresponding accuracies are obtained. Using Docker, a particular image is being taken from dataset of flowers which is untrained and is fed into the TensorFlow library implementing the CNN internally. The TensorFlow traffic symbol image classifier is implemented in Python.

4 Results Following are the results obtained from the traffic signal image classifier. Each result corresponds to a particular traffic signal and so the corresponding accuracy of classifying the traffic signal is obtained.

An Efficacious Classifier for Recognition of Traffic Symbols

7

Fig. 4 Proposed algorithm for recognizing traffic symbols

Figure 5 depicts the image classification of no horn sign as traffic signal. It can be observed that the traffic signal image classifier has a very high accuracy in recognizing the symbol as a no horn. Figure 6 depicts the image classification of no parking sign as traffic signal. It can be observed that the traffic signal image classifier has a very high accuracy in recognizing the symbol as a no parking. Table 1 depicts the comparison in obtained confidence values for the three traffic symbols, i.e., no horn, no parking, and u-turn. Hence, from the experimental results, it is observed that the accuracy of image classification and detection of various traffic symbols datasets is found to be around 88.76% as per Fig. 7.

8

D. Virmani and K. Parikh

Fig. 5 Image classification for no horn sign

Fig. 6 Image classification for no parking sign Table 1 Displays the confidence values as obtained for different traffic symbols on testing the classifier Traffic symbols

Test image I1

Test image I2

Test image I3

Test image I4

Test image I5

Test image I6

No parking

0.125

0.003

0.016

0.920

0.033

0.96

U-turn

0.715

0.016

0.006

0.023

0.700

0.027

No horn

0.160

0.980

0.976

0.056

0.266

0.008

An Efficacious Classifier for Recognition of Traffic Symbols

9

Fig. 7 Confidence values for various traffic symbols

5 Conclusion Violation of traffic rules is mainly due to ignorance of traffic symbols which has resulted in an increase in the number of accidents. This scenario is hampering the credibility of the traffic system and its maintenance. The proposed algorithm provides a novel approach to maintain the credibility of the traffic system. The above implementation is an aid which can be used to detect any traffic symbol and based on the results from the image classifier a text-to-speech conversion takes place which could avoid accidents till some extent. The execution of the proposed algorithm clearly predicts the relevant results. The experimental outcomes of the proposed implementation are based on three datasets, i.e., no horn dataset, no parking dataset, and u-turn dataset of 200 images each with an accuracy of 88%, manually tested by the annotators. Future possibilities lie in the implementation of such device with improved accuracy. Also, the size of the dataset shall be increased.

References 1. Newnam S, Von Schuckmann C (2012) Identifying an appropriate driving behaviour scale for the occupational driving context: the DBQ vs. the ODBQ. Saf Sci 50(5):1268–1274 2. Senserrick TM, Swinburne GC (2001) Evaluation of an insight driver-training program for young drivers (No. 186). Monash University Accident Research Centre, Melbourne 3. Wu Y, Schuster M, Chen Z, Le QV, Norouzi M, Macherey W, Dean J (2016) Google’s neural machine translation system: bridging the gap between human and machine translation. arXiv: 1609.08144 4. Jmour N, Zayen S, Abdelkrim A (2018) Convolutional neural networks for image classification. In: 2018 International conference on advanced systems and electric technologies (IC_ASET), Hammamet, 2018, pp 397–402. https://doi.org/10.1109/ASET.2018.8379889 5. Lavinia Y, Vo HH, Verma A (2016) Fusion based deep CNN for improved large-scale image action recognition. In: 2016 IEEE International symposium on multimedia (ISM). IEEE 6. Yim J, Ju J, Jung H, Kim J (2015) Image classification using convolutional neural networks with multi-stage feature. In: Kim JH, Yang W, Jo J, Sincak P, Myung H (eds) Robot intelligence technology and applications 3. Advances in intelligent systems and computing, vol 345. Springer, Cham. https://doi.org/10.1007/978-3-319-16841-8_52

10

D. Virmani and K. Parikh

7. Sun Z, Li F, Huang H (2017) Large scale image classification based on CNN and parallel SVM. In: International conference on neural information processing. Springer, Cham, Manipal, pp 545–555 8. Szegedy C et al (2017) Inception-v4, inception-ResNet and the impact of residual connections on learning. AAAI 9. Lu D, Weng Q (2007) A survey of image classification methods and techniques for improving classification performance. Int J Remote Sens 28(5):823–870. https://doi.org/10.1080/014311 60600746456 10. Loncomilla P, Ruiz-del-Solar J, Martínez L (2016) Object recognition using local invariant features for robotic applications: a survey. Pattern Recogn 60:499–514 11. Maji S, Berg AC, Malik J (2008) Classification using intersection kernel support vector machines is efficient. In: IEEE Conference on computer vision and pattern recognition, 2008. CVPR 2008. IEEE 12. Krizhevsky A, Sutskever I, Hinton GE (2017) ImageNet classification with deep convolutional neural networks. Commun ACM 60(6):84–90 13. Virmani D, Sharma T, Garg M (2021) GAPER: gender, age, pose and emotion recognition using deep neural networks. In: Pandey VC, Pandey PM, Garg SK (eds) Advances in electromechanical technologies. Lecture notes in mechanical engineering. Springer, Singapore. https://doi.org/10.1007/978-981-15-5463-6_26 14. Sankaran A, Goswami G, Vatsa M, Singh R, Majumdar A (2017) Class sparsity signature based restricted Boltzmann machine. Pattern Recogn 61:674–685 15. Lamba PS, Virmani D (2021) DCNN-based facial expression recognition using transfer learning. In: Bansal P, Tushir M, Balas V, Srivastava R (eds) Proceedings of international conference on artificial intelligence and applications. Advances in intelligent systems and computing, vol 1164. Springer, Singapore. https://doi.org/10.1007/978-981-15-4992-2_48 16. Girdhar P, Johri P, Virmani D (2020) Incept_LSTM: accession for human activity concession in automatic surveillance. J Discrete Math Sci Cryptogr. https://doi.org/10.1080/09720529.2020. 1804132

Assisted Living Robots: Discussion and Design of a Robot for Elder Care Garvita Ahuja, Shivansh Sharma, Maanik Sharma, and Srishti Singh

Abstract The exponential increase in the population of older generation has made it imperative to introduce advancements in their healthcare. Thus, the technologies of Robotics, Artificial Intelligence, Machine Learning, and Internet of Things are employed to fabricate a model of assisted living for the elderly. The objective is to create robotic caregivers to increase the independence of seniors. A human healthcare professional cannot be available at all times, whereas a resident robot would take care of every need of the elder. The human caregiver could delegate tasks to the robot. Therefore, a model of division of labor in eldercare is to be enforced to offer the best care to the senior. In this research paper, we first discuss the day-to day tasks that would require robotic assistance. Such robots also act as a companion, give medication reminders, suggest exercises, and support socialization. A service robot prevents the senior from falling and calls authorities in case of emergencies. We also deliberate about the abilities, limitations, and hazards of eldercare robots. The evolution of robots from desktop to telepresence and the growing advantages of the same are pondered upon too. The intent is not to replace humans as caregivers but to destigmatize senior care and make it accessible and affordable for the masses. As a part of our contribution to the study, we propose a robot design for family assistance, senior care, and care of disabled people. We envision an autonomous mobile robot to be personalized according to the user’s needs. Keywords Assisted living · Robots · Elder care · Healthcare · Artificial intelligence · Human–robot interaction

G. Ahuja (B) · S. Sharma · M. Sharma · S. Singh Vivekananda Institute of Professional Studies-Technical Campus, Delhi 110034, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. Misra et al. (eds.), Internet of Things (IoT): Key Digital Trends Shaping the Future, Lecture Notes in Networks and Systems 616, https://doi.org/10.1007/978-981-19-9719-8_2

11

12

G. Ahuja et al.

1 Introduction Recent years have recorded an exponential increase in the population of senior citizens. The birth rate faced a demographic decrement and the growth in number of older adults has caused a healthcare crisis. Before the arrival of assisted living technology, the care of old members of the society was entrusted to family, friends, and neighbors. There has been a cultural reset since the inception of technologies that contribute to the wellness of humans. The latest paradigm suggests that there has been a reduction in pensions and social security [1]. The conception of assistive robots brought on a new era of eldercare. Older generation have a decline in physical functions and impairment of senses. They have limitations of mobility (walking), deterioration of overall health, problems in vision and hearing, teeth decay and are vulnerable to dementia [2]. Assisted living robots fall under the category of Quality-of-Life Technologies (QOL). They have evolved over the years to cater the needs of humans, mainly the ones plagued by diseases and the ones too old to care for themselves. These technologies monitor and assess the subject, run a diagnosis of their condition and devise a treatment plan for their stability/recovery. They also offer companionship to the subject. QOL technologies have done successful diagnosis of ailments and monitoring the health of the senior [1]. The development of information technology has enabled to detect health issues in home environment. These robots can be introduced in home settings and also nursing homes that have a hospital type aesthetic (Fig. 1). Assistive robots can help the senior achieve daily tasks such as providing food, medicines, bathing, and safely lifting the patient. Their company helps the subject with conditions like Dementia, Alzheimer, and Parkinson’s Syndrome. However, people were not accepting to assistive robots in the initial years. Studies quoted that they dehumanized eldercare and observed that caregivers should not be replaced by robots. The motive is not to replace human caregivers with robots but to add them to the organizational structure of eldercare staff. Fig. 1 Mabu, the eldercare robot by Catalia Health [3]

Assisted Living Robots: Discussion and Design of a Robot for Elder Care

13

Assistive robots have not been that prevalent in India except during the COVID19 pandemic, and they were used in hospitals to provide things to the patients. Due to the shortage of working-age health professional, it would be better for a robot to be at the beck and call of the old person. This would also reduce care costs as the robot would offer a good number of services in a single unit. For older adults live in non-traditional housing, enhancing their quality of life using the latest technology is the next step. Robots for elderly care consist of companion and family robots and pet-like companion robots both of which could be mobile or immobile. However, these robots lie under the category of Autonomous Mobile Robot (AMR). One such robot is Mabu which is a non-mobile elder companion, capable of interaction, facial recognition, and regularly checks with patient [4]. Gerijoy is a mobile caretaker and also offers emotional support [5]. Elli-Q, in addition to being a standing robot that gives reminders, provides entertainment, and suggests treatment and exercise routine. It also helps the elder to communicate with family members or doctors [6]. Looking at the pet-like companion robots we have, Paro is seal-like interactive robot and had marvelous therapeutic results on the elders [7]. MiRo is a Dog-like autonomous interactive robot [8]. JustoCat is a cat-like interaction robot companion which responds to touch, petting, and voice. JustoCat helped out dementia patients and progresses interaction between patient and health professional [9]. There are also some robots designed to prevent falling of elders like Hobbit [10]. Giraff, Vgo, Pepper, and Kompai are such telepresence robots which help in communication, task manipulation, and home assistance [11–14]. RIBA I and II are for interactive body assistance [15]. Stevie is a dancing robot that caters to the physical, social, and emotional needs of older adults (Fig. 2) [16]. In this paper, we will be addressing the nuances of eldercare and our contribution toward it.

Fig. 2 Stevie interacting with an old man

14

G. Ahuja et al.

The manuscript is organized as follows: Need of assisted living robots is discussed in Sect. 2. Sections 3 and 4 discuss the abilities and limitations of assistive robots, respectively. Section 5 addresses the hazards that come with employing robots for eldercare. In Sect. 6, we talk about the design of our robot. Section 7 provides insights to our vision. Further, we discuss the contribution of our study and then comes the conclusion.

2 Need of Assisted Living Robots Robotics as a field involves the commencement, designing, manufacturing, and the operation of an automated machine that can perform particular tasks with little or no human intervention. Robots are the need of the hour. As the world population ages, the need for these automated machines is increasing. According to the official data of WHO, currently there are more than 1 billion people over the age of 60 and numbers are expected to rise about 1.4 billion by 2030. As the figure increases, the need of assistances like nurses and caregivers increases. Thus, robotic assistants can be very useful in such a case. Robots can help them to care for themselves and provide emotional support in retirement or old-age homes. These machines can help the people in cleaning, entertaining, etc. In hospitals, the robots can be used for performing basic tasks such as provide food, medicines to patients, assist doctors in surgeries, etc. In the year 2018, McGinn and his colleagues built a white robot with rolling base and named it Stevie. It had short movable arms and a head that displayed animated eyes and a mouth. The idea of the team was to provide automated robotic facility to the elderly people living in retirement or old-age homes. The robot was designed to communicate in different languages, patrol corridors at night, perform cleaning tasks, and so on [16]. Several such types of assisted robots are available in the market and further researches are also being conducted on the same. These robots are not only confined to do chores and entertain but also to help people, either disabled or normal, in walking, standing, and sitting. These machines can be used for providing time-to-time medicines and set reminders. The more the number of aged people, the more nurses, and caretakers are required. And to cope with the huge demand of these caretakers, robots can be used to divide the work force. The future without robots is unpredictable and thus further advancements in the field are required (Table 1) [17].

3 Abilities of Assisted Living Robots The literature gathered the statistics that older generation preferred to receive care from a robot rather than a human. The plus point of this is that a robot will be available for the whole day while a caregiver will have other errands and might not be free

Assisted Living Robots: Discussion and Design of a Robot for Elder Care

15

Table 1 Previous works on assisted living robots Companion and family robots

Elder companion robots

Robots

Functions

Buddy

Telepresence, personal assistant, patrols the house, hands free helper, social interaction, connects family, playmate for kids, fall detection, and medication reminders for seniors [18]

Jibo

Stationary, patient care, eldercare and education, helper robot, suggests exercise, reminds of medication, helps through stress, social interaction [19]

Aido

Telepresence, home safety, playmate and teacher for kids, handles schedule, recognizes the members of household and learns about them, Handywork [20]

Lynx

Smart home robot, telepresence, voice interaction, facial recognition, entertainment, can teach anyone [21]

Kuri

Autonomous mobile robot, Facial and vocal recognition, LED mood lights, touch sensors, robotic nanny When low on charge, automatically retreats to charging station [22]

Zenbo

Education and healthcare, IoT functions, telepresence, perception AI [23]

Miko

Telepresence, education, and entertainment [24]

Mabu

Stationary, personal healthcare companion, Facial recognition, learns about patient, checks in with the senior [4]

Elli-Q

Desktop, social interaction, entertainment, Sets exercise routine, medication reminders [6]

Gerijoy

Telepresence, healthcare, communicates through avatars, emotional support [5]

Telenoid

Humanoid telepresence, improves communication within humans [25]

Pet-like Paro companion robots

Mobile with no arm

Robotic baby seal, helpful for people with dementia or mental health issues [7]

JustoCat

Robotic cat, interaction and communication, does not need power [9]

AIBO

Robotic dog, improved senior communication, facial recognition [26]

MiRo

Very versatile, can be employed in multiple settings, detection sensors, interactive and expressive [8]

Nabaztag

Robotic rabbit, stationary, interactive [27]

Joyforall

Robotic cat, soft, respond to touching, petting, and voice [28]

Hobbit

Telepresence, prevents falls [10]

Vgo

Telepresence, versatile, can be used for eldercare, patient monitoring, people with disabilities, pediatrics [12] (continued)

16

G. Ahuja et al.

Table 1 (continued)

Mobile with arms

Robots

Functions

Kompai

Telepresence, security, health monitoring, physical autonomy, also acts as a walker [14]

Giraff

Telepresence, screen interface [11]

Pepper

Humanoid, telepresence, facial, speech and emotional recognition, touch sensors, autonomous navigation, fully programmable [13]

RIBA I and II Robot for interactive body assistance, lift up or set down humans, strong arms, Tactile sensors [15] Care-O-Bot

Mobile robot, customized arms, or tray [29]

Personal Robot 2

Can drive cars and run errands outside the home setting, grasps objects [30]

Robear

Humanoid, telepresence, safely lifts patients, emotional support [31]

which would leave the old person without help for daily activities. Essential activities for a senior are as follows: (a) Personal hygiene such as bathing, toileting, dressing, brushing hair, and cutting nails (b) Feeding (c) Medication (d) Socialization To be realistic, not all of these activities can be comfortably made possible for an immobile senior solely by a robot. However, looking at the organizational aspect of assistant living, healthcare can be made more efficient in a retirement home or a hospital by task delegation to the resident robot and thus service provision is improved. Whether the chore is performed by the human or the robot, priority of the action is superior. If an elder is in a medical emergency or just in need of some assistance, the rate at which help is provided matters more. In the future, robots could be programmed to administer medical treatment to patients. They are quicker to train as compared to a human nurse, not expensive to maintain, can be easily charged in a plug-in outlet and are able to do odd and repetitive jobs without complaint. Robots would minimize the possibility of errors in prescribing treatment. Also, there would be no hazardous effects of being exposed to chemicals. Involving robots in the healthcare dynamic reduces the cost effectively almost by 65%. The connotations of the aged as being negative or vulnerable must be shot down. This must be destigmatized. Robots are multifunctional and they do check all the boxes instead of the personal hygiene category. Robotic walkers have also been developed to support the older person to be more mobile. For immobile patients, they can also safely lift the person to be taken to another room or the facilities. Domestic

Assisted Living Robots: Discussion and Design of a Robot for Elder Care

17

robots are proficient in household chores such as cleaning, cooking, washing, and laundry and can very well help around the retirement home or hospital. Robots with detection sensors to save the elder from falling and correct their posture are also in the picture. Since the inception of Artificial Intelligence and Machine Learning, robots are capable of storing information about the subjects and learn about their various habits and disabilities. A resident robot also improves safety of the senior in case if an old person lives alone and they undergo some domestic accident or fall dangerously ill and aren’t able to contact anyone. In such a case, no one would come to know for days until there is a family visit or health professional comes for a scheduled checkup. If there is a robot responsible for the elder, it would immediately do the needful as well as call for help. The robot would also alert the authorities and protect the senior in case of home intrusion. These robots can not only be used for the older generation, but also the specially abled young persons who don’t have cognitive problems. As previously discussed, robots are programmed to be excellent communicators which is even better because of facial and vocal recognition and emotional intelligence. Emotional intelligence comes in handy to console the senior if they are sad or in a bad mood. They can connect the elder on video call with their friends and family and also the doctor. Robots can conduct a dance party for all the elders and accompany them for exercises. Acceptance of this new technology will be hard in India but it will do wonders to the eldercare field if done judiciously. This will increase the safety, quality of life, and independence of seniors [2].

4 Limitations of Assisted Living Robots Some people believe that introduction of robots in the healthcare dynamic could dehumanize it. It could also reinforce isolation of the elder and since the robot is performing all the errands, it would laze the senior and hence decline mobility. If the robot is always instructed to do some basic tasks which could be painlessly done by the elder, then it would inflict an impact on memory. It would enforce a habit of forgetting. Thus, there is a possibility of negative health repercussions (Table 2). Mishaps cannot be foreseen every time. The robot could malfunction and breakdown with only the helpless elder present. This could result in a disastrous situation. Some seniors stated that they had difficulty imagining talking with a computer that asks them to take their pills. They felt that a “metal box with a speaker” shouldn’t dictate their life. Even though a robot has an andromorphic shape and is programmed to act like a human, the older adult would not feel the same connection for the machine as they do for their peers or the generation younger than them. The fact that a robot will be taking care of their well-being took out the emotional component of caring. The seniors would be confronted by the psychological implication that they do not have to care back for someone and just be on the receiving end of it. The seniors

18

G. Ahuja et al.

Table 2 Comparing robotic assistance and human assistance Human assistance

Robotic assistance Abilities

Limitations

Abilities

Available 24/7

Professional healthcare

Emergency alerts

Dignity of the elder remains intact

Learns about the patient

Essential human contact

Prevents falls

Actual friendships, responsive interaction

Medication reminders

Divided attention of the human caregiver among other elders

Could irritate the elder Trust and give unsolicited suggestions

Performs household chores Personal assistance

Limitations

No isolation Negative health repercussions; could trigger mobility decline

Eases the burden of human caregiver

Incorporates care

Share experiences and folklore

Social interaction

Dehumanize care

Emotional comfort

Reinforce isolation

would also stop feeling the need to go out and do some self-care tasks since they have a robot at their beck and call. The human aspect of aging is also put at risk as you cannot share your memories, personal advice, or life anecdotes with a robot. All the folklore that our society stands on would slowly cease to exist. This could result in decline of culture. There are multiple reasons why elders feel that relying on a robot for care could prove to be inconvenient or unsafe. The presence of a human is extremely important to reassure the senior and give emotional comfort. For a nursing home setting, a caregiver should be present for most hours of the day to attend all the resident elders whereas in a home setting, health professional visits should be scheduled as per the senior’s ease. To conclude, a fair balance of human care and robot care should be provided to the older generation. The model of assistance would work in such a way so as to reduce the burden of the health professional and tasks are delegated to the robot for efficiency [2].

Assisted Living Robots: Discussion and Design of a Robot for Elder Care

19

5 Hazards of Employing Assisted Living Robots The use of robots for caring of elder people is bringing a revolutionary change in the care industry. There may be many benefits of using robots as caretakers and helpers, but at the same time, this brings along various issues. Privacy issues: The robots equipped with surveillance equipment can be used as nursing robots and can turn out to be very useful but at the same time, it can lead to a violation of patient’s privacy. Without proper guidance and instructions, these robots can turn out to be a threat to the patient’s life. Social interaction: Though robots can be used for various purposes sometimes, they fail to interact. Humans most of the time prefer to interact with other humans and not machineries. Thus, replacing human caregivers with robots can lead to issues related to dignity and happiness of patient. Aged citizens or patients most of the time enjoy interacting with humans only but with advancements in this field and the development of human-like interaction robots, the issue can be resolved. Due to the mechanical, pre-programmed, neutral way of interaction, robots can never replace human nurses and caretakers. Liability issues: Because these assistance robots are mechanical, programmed devices, it is impossible to hold them responsible for any malfunction or any harmful event. Thus, it is very difficult to assign civil and criminal blame to the machine. At the end the manufactures, programmers, suppliers, and technicians are held accountable for all the consequences. The advancements in the field of robots for assistance living such as giving them artificial intelligence are one of the key ways of giving them autonomy and decision-making capabilities. Consequently, in case of wrong decision made by the robot can lead to serious legal challenges. Employment issue: Since robots assure high productivity and high efficiency at low cost, these can easily replace the people working in healthcare sector or people providing assistance to the aged citizens. As a result, concerns are raised about the use of robots in workplace. In particular, a decline in the demand for nursing services delivered by people might hurt an industry that is already in trouble. The lack of core medical assistant skills in our communities could result in a further decline in the number of people interested in pursuing nursing careers. Response time: Though robots may be faster than human in the work field, but when it comes to interaction with humans or giving response, robots can be found slow. A robot works on the command given by the person, in case of an emergency, a nurse or a caretaker can act according to the situation but a robot will work according to the program or information stored inside it. In case of artificially intelligent robots, the decision made by the robot itself can be wrong and can worsen the situation. Safety and security issues: The increased usage of robots for assisted living brings along a few safety and security concerns. Safety and prevention of injury from robots should be the utmost concern for any manufacturer or provider. How to prevent robots from being compromised is another specialized security issue that needs to be addressed. It is very important to keep in mind that with technology becoming so powerful, it is very crucial to monitor any hacker activity that could use

20

G. Ahuja et al.

it for illegal purposes. As these robots are automated machines, they are vulnerable of getting hacked. So, a proper security network for robots needs to be developed [32].

6 Design This manuscript is focused on the build out of an intelligent assistance which includes autonomous robots with sensors and also external actuators. Robots add to traditional IoT structures a variety of factors to consider better user interaction, complex functions, workability, and recovery, etc. (Fig. 3).

6.1 General Structure of the System The complete whole structure of the system can be seen. The system is made up Power Source which includes 9 V battery, Controller, Control and Task Program, Sensor, Actuator which includes 2 Hobby Gearmotors and has an LCD Screen (Fig. 4). (a) Power Source: Power source supplies energy to the robot. It can be electricity, combustible materials, or a battery.

Fig. 3 Circuit of the proposed robot

Assisted Living Robots: Discussion and Design of a Robot for Elder Care

21

Fig. 4 Flowchart of the structure

(b) Controller: The controller is responsible for coordinating the movements of the robot. The microcontroller also manages to collect the input material first through its sensor. (c) Control and Task Program: A set of instructions from the creator of the robot’s controller is known as management system, whereas the series of questions usually provided by the user is called Task Program. To conduct a task, the manipulator must perform its movement activity organized by the Task program. (d) Sensor: The sensor in robots takes specific measurements of the environment. This sensory data helps the robot to take the decision accordingly. This is done to ensure the safety of the robot and the user. (e) Actuator: The actuator is responsible for producing and controlling the motion of the robot. It is the device which causes movement (it can be movement of joints or movement of wheels), emits sound (by a loudspeaker), or shows some output on an LCD (Liquid Crystal Display) Panel with relevant code.

6.2 Obstacle Alert by the Robot The Ultrasonic sensors detect the distance by using ultrasonic waves. The system radiates an ultrasonic wave and receives the wave sent from the target. It measures the distance by measuring the duration between emission and reception. The LCD (Liquid Crystal Display) will show “COLLISSION ALERT” along with the distance remaining in the obstruction. The LED lights on the robot will also start blinking red so as to alert the users with partial eyesight or hearing impairment (Fig. 5).

22

G. Ahuja et al.

Fig. 5 Circuit displaying collision alert

6.3 Verbal Interaction For the robot to communicate verbally, it takes voice input from the user, which is then converted to the text via the SpeechToText Module. The converted text is then processed using the Natural Language Processing library. Further, the robot processes the user’s request and formulates a response in text form which is then converted to speech using a TextToSpeech Module (Fig. 6).

7 Our Design We have planned to design a robot called “Keva” for the Indian nursing homes. We envision it to be of about 4 feet and with a white outer and a screen with additional LED lights. Keva would have a circular face, triangular kind eyes, and a headset placed on its head. Family and Companion robots have already infiltrated the Indian market but there hasn’t been much progress in the assisted living eldercare sector. The LCD placed near Keva’s stomach would proudly display the words “Hello! I’m Keva, your new best friend.” Keva would be fluent in almost all languages of the world but especially all Indian languages. It will have facial, vocal, and emotional recognition.

Assisted Living Robots: Discussion and Design of a Robot for Elder Care

23

Fig. 6 Verbal interaction: speech and language processing

We reckon Keva to have telepresence and it should detect and prevent falling of the elder. It would remind the elder to take medicines and check-up about having eaten their food. Keva should be able to detect blood pressure, diabetic levels, and cardiac health. It would also be an exercise or walk buddy. We want it to be witty and clever so that it humors the elder and they are never bored of talking to their robot. Keva should also be able to entertain the old person, connect their family for video calls and conduct a dance party with their favorite songs. The LED lights would now come handy to create a party vibe and keep the elder happy. The same LED lights would be useful when the robot is interacting with a person with impaired sight or hearing as both the kinds can sense laser light. It would help the elder signal Keva about their mood or health (Fig. 7). Keva could also be used for the disabled people or for educating children. It would be a multi-purpose robot ready to be personalized for the family’s needs. In [1], an autonomous mobile robot for eldercare was proposed and they argued that it needed to be affordable. Similarly, when we get to building Keva and releasing it in the market, we will make sure that it is cost-efficient and accessible to the needful.

24

G. Ahuja et al.

Fig. 7 Our robot design, Keva

8 Contribution of Our Study Our study contributes to the field of assisted living robotics by making observations from the existing literature and highlighting the perceptions that come with the dynamics of assistive healthcare robots and an older adult. During the course of our study, we have come across the medical, personal, and societal claims that debate about caregiver robots. Our manuscript started with discussing the tasks that would require an elder to seek the help of a caregiver and the inception of QOL technologies that would employ a robot to take the burden off the healthcare professional. The presence of a helper robot would increase the independence of the user. Then we ruminated on the abilities and limitations of assistive robots. The hazards that come with enlisting robots in a personal setting are also pondered upon. Following Lehoux and Grimard, there should be a division of labor between the human caregiver and the robot since the seniors need human presence in their lives. Then we discuss the design of the robot “Keva” and the significance of LED lights for an elder with impaired hearing or weak sight.

9 Conclusion During the course of our study, we have established that it is important to keep the logistics of care in mind while designing technologies for the seniors or for the specially abled. It is also essential to do in-depth analyses of the pros and cons of introducing robotics and IoT into the picture of family life and healthcare. Thus,

Assisted Living Robots: Discussion and Design of a Robot for Elder Care

25

our research was dedicated to understanding the needs and criticisms of the users of such a technology. Medical policymakers should look into the dynamics of a patient, healthcare professional, and caregiver robot in order to improve the healthcare system that surrounds the treatment of an elder, a terminally ill person, or a disabled person. Delegation of tasks among the human and the robot caregiver would prove to be efficient and be cost-redundant. The social, economic, and medical compatibility of the robot must be high. Our future work will highlight the affordability of mobile robots in eldercare.

References 1. Johnson MJ, Johnson MA, Sefcik JS, Cacchione PZ, Mucchiani C, Lau T, Yim M (2020) Task and design requirements for an affordable mobile service robot for elder care in an all-inclusive care for elders assisted-living setting. Int J Soc Robot 12(5):989–1008 2. Lehoux P, Grimard D (2018) When robots care: public deliberations on how technology and humans may support independent living for older adults. Soc Sci Med 211:330–337 3. https://robots.ieee.org/robots/mabu/ 4. http://www.cataliahealth.com/platform-ai/mabu-experience/ 5. https://www.ge.com/news/reports/a-robotic-companion-for-the-elderly 6. https://elliq.com/ 7. http://www.parorobots.com/ 8. https://www.miro-e.com/robot 9. https://robots.nu/en/robot/justocat 10. Fischinger D, Einramhof P, Papoutsakis K, Wohlkinger W, Mayer P, Panek P, Vincze M (2016) Hobbit, a care robot supporting independent living at home: first prototype and lessons learned. Robot Auton Syst 75:60–78 11. https://telepresencerobots.com/robots/giraff-telepresence/ 12. https://robots.ieee.org/robots/vgo/ 13. https://www.softbankrobotics.com/emea/en/pepper 14. https://kompairobotics.com/robot-kompai/ 15. https://newatlas.com/riba-robot-nurse/12693/ 16. https://time.com/longform/senior-care-robot/ 17. Savage N (2022) Robots rise to meet the challenge of caring for old people. Nature 601(7893):8– 10 18. https://buddytherobot.com/en/news/buddy-the-emotional-robot-blue-ocean-awards/ 19. https://jibo.com/ 20. https://www.robot-advance.com/EN/actualite-domestic-robot-aido-159.htm 21. https://www.theverge.com/2017/11/20/16681396/amazon-alexa-powered-lynx-robot-ubtechrobotics 22. https://robots.ieee.org/robots/kuri/ 23. https://zenbo.asus.com/ 24. https://in.miko.ai/ 25. https://robots.ieee.org/robots/telenoid/ 26. https://us.aibo.com/ 27. https://www.inceptivemind.com/internet-connected-rabbit-nabaztag/12143/ 28. https://joyforall.com/products/companion-cats 29. https://www.care-o-bot.de/en/care-o-bot-4.html 30. https://robots.ieee.org/robots/pr2/ 31. https://www.theguardian.com/technology/2015/feb/27/robear-bear-shaped-nursing-carerobot

26

G. Ahuja et al.

32. Grasso C (2018) Challenges and advantages of robotic nursing care: a social and ethical analysis. The Corporate Social Responsibility and Business Ethics Blog

Localization Technologies Anshika Jain, Bhumika Jain, Arohi Singhal, and Srishti Singh

Abstract Whether the goal is simply to aid shoppers in a mall or prevent accidents in warehouses, the solution is increasingly spelled Localization. Given that the literature on localization is vast and spans many different disciplines, we conducted a comprehensive review of the dominant technologies used in indoor as well as outdoor environments. Surveying localization technologies, including healthcare, mapping, navigation, indoor environment, and wireless sensor networks, the survey examines the technologies used in localization systems. Moreover, we classify the existing approaches into a structure so that they can be reviewed and discussed in an organized manner. Localization techniques each have their pros and cons, and cannot be used independently. Keywords Localization technologies · Indoor positioning · Tracking and detection · Wireless sensors · Navigation · Mapping

1 Introduction Localization based services are becoming increasingly popular with people around the world. In localization, imagery, messages, features, and products are tailored to fit the cultural, linguistic, and geographical preferences of a specific audience. Localization plays an important role in the expansion of international businesses. New technology allows localization to become more personalized, more targeted, more efficient, and more effective. A major goal of today’s communications systems is the availability of location information [1]. The positioning systems could be used either indoors or outdoors. The outdoor localization is mostly achieved by making use of the Global Navigation Satellite System. In localization, several common techniques are used where the variety of these techniques are based on Received Signal Strength Indicator (RSSI) A. Jain · B. Jain (B) · A. Singhal · S. Singh Vivekananda Institute of Professional Studies—Technical Campus, New Delhi, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. Misra et al. (eds.), Internet of Things (IoT): Key Digital Trends Shaping the Future, Lecture Notes in Networks and Systems 616, https://doi.org/10.1007/978-981-19-9719-8_3

27

28

A. Jain et al.

signal measurements, and Time of Flight (TOF) measurements. While utilizing these measurements or localization techniques has its advantages, it has certain limitations that will make the process of localization difficult. With the use of inertial sensors or wearable devices, dead-reckoning measurements can also provide location information.

2 Healthcare and Localization 2.1 Professional Healthcare Services As society rushes to prevent infectious diseases and pandemic pests, all available capabilities are mobilized to push that epidemic in many ways, including medical treatment of patients when appropriate treatment is available or isolating patients until treatment is available. Users can travel to any nation for affordable medical treatment as the world becomes more interconnected [2]. Medical translators are critical in helping patients and their families communicate when they are not proficient in the local language. In addition, they provide substantial support to doctors, nurses, technicians, and deaf patients.

2.2 Body Area Networks (BANs) Currently, emergency healthcare systems are underdeveloped, particularly in developing countries. There can be a sudden relapse of Cardiovascular diseases (CVDs). Time is of utmost importance for patients. By building an emergency healthcare system, we hope to save more lives in emergencies. Telemedicine has grown in popularity with the rapid development of Body Area Networks (BANs) and wireless communication systems. Sensors on the body surface or implanted sensors are used to monitor several vital signs parameters and all signals are gathered and transmitted by a receiver (for example, a mobile phone or a computer) to a doctor. In recent years, advances in mobile technology and BANs have facilitated the use of mobile-based health monitoring and alert systems. In such systems, patients can receive real-time feedback about their health condition, as well as alerts in case of health-threatening conditions.

Localization Technologies

29

Fig. 1 Framework for emergency healthcare system based on MCC

2.3 Mobile Cloud Computation (MCC) Mobile cloud computing is a type of cloud computing that is specifically designed for mobile devices. Massive computing, storage, and software services can be implemented in a scalable and virtualized way with much lower energy consumption. Mobile devices can execute many applications with high computational complexity using MCC. As shown in Fig. 1, based on the MCC framework, more accurate offsite personalized medical diagnosis and treatment can be provided.

2.4 Localization Technologies for Radiotherapy Delivery The motion inherent in normal organs and target structures has been a major area of technological advancement in radiation oncology treatment delivery [3]. As

30

A. Jain et al.

Fig. 2 Structure of Sensor Node

radiation therapy technologies have advanced over the last decade, such as Intensity Modulated Radiation Therapy (IMRT), stereotactic radiosurgery/radiotherapy, Stereotactic Body Radiotherapy (SBRT), and image-guided radiation therapy, it has become increasingly important to position patients accurately and reproducibly in the treatment positions. To address these localization issues, implantable devices have been developed and can be tracked during and after treatment such as Calypso 4D localization system, AlignRT, and Radiocameras.

3 Mapping 3.1 Simultaneous Localization and Mapping (SLAM) Computers, devices, robots, drones, and other autonomous vehicles can simultaneously detect and map the environment using Simultaneous Localization and Mapping (SLAM). Autonomous vehicles and other robots rely on SLAM to determine where they are and determine the best route to their destinations. SLAM creates its maps, making them faster, more autonomous, and more adaptable than preprogrammed routes. SLAM works by combining several emitters, sensors, and AI for a single purpose. For example, SLAM robots employ various types of cameras and sensors to understand their environment, such as radar, Lidar, ultrasonic, and others. The better a robot understands its environment, the more effectively it can map, navigate, avoid obstacles, and adapt to changes [4]. SLAM has been reduced in some applications as a result of highly accurate GPS modules [5]. In some outdoor environments, GPS can almost entirely replace SLAM. The GPS may suffer from a reduction in performance or outages, in which case SLAM can step in and fill in the gaps in navigation where more detail is required as well as take over in these circumstances.

Localization Technologies

31

3.2 Detection and Tracking of Moving Objects When physically navigating around an obstacle is difficult or dangerous, being able to detect motion and track a moving object provides a critical advantage [6]. An object hidden from view can be reconstructed using previous methods. The process of object tracking usually involves the detection of objects. Here is a brief overview of the process: 1. Creating a bounding box around an object to classify and detect it. 2. Identifying each object (ID) uniquely. 3. Detecting and storing the relevant information about the detected object as it moves through frames. Object detection is the process of detecting a target object in an image or a single frame of the video. Object detection will only work if the target image is visible on the given input. Any interference will prevent it from detecting the target object [7].

3.3 Sensor Map-Based Localization Depending on the map patterns, localization in the known environment can be divided into four categories: (1) Localization based on feature maps: A feature map, as its name suggests, contains features that can be used to model the environment. A vehicle localization application has two types of features. Most studies use artificial features or global features in coarse localization or topological localization procedures. Computer vision or image processing methods are commonly used to extract these features, which do not exist in reality. A few examples of this type of feature are Scale-Invariant Feature Transform (SIFT), Speeded Up Robust Features (SURF), Haar-like, and map grid. Secondly, natural features (appearance) are often used in fine or metric localizations. Methods based on computer vision or machine learning can detect these features in the real world. These features include lane lines, landmarks, and buildings. (2) Localization based on occupied grid maps: A grid map is another type of map. Each cell on the map has its unique property. Earlier research used a relatively simple grid map. Each cell contains only one binary value indicating whether there are any obstacles in this cell, if so, vehicles cannot pass through it, if not, they can. Often, binary grid maps are referred to as occupied grid maps. In this case, occupied indicates if there are obstacles in the cell. (3) Localization based on topological maps: Among other kinds of maps, there are topological maps. A topological structure represents the environment through links and junctions. Links represent connected

32

A. Jain et al.

Table 1 Summary of model of maps Models of maps

Amount of information

Map’s robustness

Difficulty of mapping

Accuracy of map

Data size

Feature map

★★★

★

★★

★★★

★✩

Grid map

★★

★★

★

★✩

★★

Topological map

★

★★✩

★★✩★

★✩

★

Point cloud map

★★✩

★✩

★✩

★★★

★★★

relationships between junctions (corners, intersections, feature points, etc.), while junctions are important positions in the environment (corners, intersections, feature points). Maps such as Open Street Map (OSM) are typical topological maps. (4) Localization based on point cloud maps: Lastly, we have point cloud maps. In the context of vehicles, a point cloud is a collection of points sensed by laser scanners or cameras. Light variance is insensitive to this type of method. However, localization can fail when there are new buildings or obstructions in the scene. The use of low-cost sensor configurations, such as monocular cameras or low-cost laser scanners, was widespread in recent studies. Low-cost sensors can achieve an accuracy of 10–20 cm, which allows intelligent vehicles to make big strides in localization accuracy compared to traditional methods (like GPS, IMU) [8]. Stereovision sensors and high-beam laser scanners are relatively expensive, but they are capable of providing depth information. The map models can be compared in Table 1 to provide high-precision 3D localization. Whether a map represents the surrounding environment well or conveys enough surrounding information depends on the amount of information it contains. The high-precision maps can then be used to implement accurate localization. With sensors (such as cameras, LIDARs, etc.) on the vehicle, real-time information about the surroundings can be collected. Live data can be compared with prior maps to estimate the vehicle’s position and pose.

4 Navigation 4.1 Satellites Satellite Navigation is based on a global network of satellites that transmit radio signals from a central orbit in the middle. A combination of signals from at least four satellites is necessary to determine the location and time of each of the 31 satellites. Each satellite emits signals that allow receivers to determine the location and time of each satellite. The most accurate time is provided by GPS satellites equipped with

Localization Technologies

33

atomic clocks. GPS satellites that carry atomic clocks provide the most accurate time. Time information is entered into the satellite broadcast codes so that the recipient can continuously determine when the signal is being broadcast. The signal contains data used by the recipient to calculate satellite locations and to make some necessary changes for optimal configuration. The receiver calculates the distance between the receiver and satellite- based on the difference between the reception time and the broadcast time [9]. Signal speed should be decreased or delayed by the ionosphere and troposphere if the distribution is delayed. With information about the range from three satellites and the satellite location when the signal is being sent, the recipient can calculate his or her three-dimensional location. GPS atomic clock coordination is needed to calculate the range from these three signals. The receiver, however, does not require an atomic clock because it measures the fourth satellite. Latitude, longitude, altitude, and time are calculated by the receiver using four satellites.

5 Sound Due to the nature of the sound equipment, the Sound Source Location (SSL) can be used in multiple locations, including targeted local performance inside, outside, and even underwater. Therefore, local audio source creation has many applications within several areas of engineering, such as biology audio courses, smart conference rooms, traffic monitoring, local underwater source acoustical source, local speaker, and military industries such as tracking flying objects, sniper fire localization, and local shooting. For guidance on the audio source, two basic methods are used. The first method is based on Time Variations Arrival (TDOA) and the second is how to light. On the other hand, consumption of quadcopters has recently grown significantly, and today you may easily find the right quadcopter with low flight capacity. Therefore, these flying objects are becoming more and more important.

5.1 Wi-Fi A research that is based on Wi-Fi as a tool for indoor positioning was the indoor triangulation system. In this method, the location of various Wi-Fi devices was measured by using the system’s own Wi-Fi nodes within a given space. The concentration and location determine the localization precision. Another noteworthy example of Wi-Fi localization systems is about various radio beacons. Since they all used protocols, beacons were given unique or semi-unique identification numbers. Inspection systems use this identifier to recognize the user’s internal location. Indoor location applications have improved detection capabilities to detect the current state of the mobile device. Released for smartphones, this app includes both offline and online

34

A. Jain et al.

fingerprint scanning levels. Meanwhile, the authors created an application demonstrating Wi-Fi trilateration for internal positioning using an Android mobile device [10]. This was achieved by applying a series of signal strength measurements that improve location accuracy. Finally, the author introduced the latest systems and solutions developed for wireless localization of Indoor-71 applications and their algorithms. In this study, the authors presented existing wireless solutions for indoor tracking and tried to classify different methods and systems.

5.2 Optical To solve outside navigation problems, autonomous navigation algorithms have been implemented in diverse robotic application studies. There are two major causes for the big error of their local construction: incorrect odometry and limit in the use of the Global Positioning System. Against the above problems, a new one has been created, optical navigation sensors can solve critical issues in encoder-based odometry problems. A computer mouse sensor mouse has a very similar function to the optical navigation sensor. Because the visual wandering sensor is neither self-confident nor is affected by the drift of a moving and inaccurate robot movement model, its function of measuring the position of a robot is significantly higher than that of odometry based on coding. A visual navigation sensor and an IMU limit the robot’s state in the forecasting step, and GPS measurements adjust its predicted condition. Depending on GPS communication quality, Differential GPS (DGPS) or Single GPS modes are selected in the configuration. In addition, the measuring parameter of the visual navigation sensor is autonomously tuned.

5.3 Visible Light View light positioning (VLP), a local technology based on visual acuity light (VLC), has gained much attention in recent years due to its ability to provide high accuracy based on existing lighting infrastructure. The modulated LED broadcast unique identity (ID) by changing the high frequency invisible to human eyes, but invisible to photodiodes (PD) and rolling shutter effect (RSE) cameras. The LED ID can be mapped/modified once as it is normally stopped and is not easily vulnerable to environmental changes. Therefore, the “Last Mile Problem” local performance is solved with VLP and a pre-built LED map of LED. Design a PD-based indoor positioning system using a reading machine to provide approximately 10 cm vertical accuracy. In addition, some high-resolution camera-based VLP (STOA) programs can provide upto-date accuracy on portable smartphones or mobile robots. However, VLP systems still face real challenges despite their promising performance.

Localization Technologies

35

5.4 Inertial Inertial Navigation Systems (INS) roaming systems can calculate location, either about a specific system/point or complete links. The basic idea behind the INS system is to measure changes in related movements (at a rate of acceleration) to produce a variable area in another reference framework that does not work over time. Inertial measurement units (IMUs) are the core of an INS system. A gyroscope and three accelerometers are arranged in an orthogonal pattern on the device. Many studies report that the INS-only system has been successfully used in setting up, for example, the pedestrian computation system (PDR), based on the kinematics of human mobility to track the location and posture of a pedestrian in the house places. This can be done by using waist-mounted IMUs or smartphone IMU.

6 Indoor Environment Localization 6.1 Inertial Measurement Unit (IMU) Nowadays, the demand for internal navigation systems is increasing rapidly based on location services, such as shopping and advertising. From a technical point of view, internal navigation programs can be divided into two categories: infrastructure-based programs, and those without infrastructure. The first stage requires installing the devices in place, and the pedestrian area is calculated using those devices. Inertial Measurement Unit (IMU)—based Pedestrian Counting Systems (PDR) can provide related stops in the first place. They need the first position to be known in other ways, whereas both the sample size and the relative position ratio are very high. IMU sensors in PDR systems include accelerometers, magnetometers, and gyroscopes, which are used to determine the location of a user based on his or her title and step length. A home-based image is localized using the camera, which may be a monocular, stereo, or RGB-D camera. Using local photographic algorithms, the camera captures the test area [11].

7 Radio Frequency Identification (RFID) RFID technology is a new technology that can track the mobility of objects or people. There are three types of RFID systems: passive, semi-passive, and active. RFID systems consist of tags (also called transponders, smart tags, wireless barcodes, or smart tags), readers (also called decoders, interrogators, transmitters, receivers, or transceivers), and a host computer, and software. A wireless or wired connection is available between a reader and a host computer. Semi-passive RFID systems work on the same principle as passive ones, but their Batteries are built into semi-passive

36

A. Jain et al.

tags. The battery provides built-in power for the tag’s telemetry and sensor control circuitry, giving the tag more power to communicate [12]. However, the onboard power source is not directly used to generate Radio Frequency (RF) electromagnetic energy. RFID systems also operate in several areas of the radio frequency spectrum [13]. Different regions are typically used for different applications, and there is no single frequency that is suitable for all applications, all regions, or all types of operating environments.

8 Wireless Sensor Networks (WSNs) The use of Wireless Sensor Networks (WSNs) is widespread in industries such as manufacturing, government, defense, and the environment. In the monitoring area, WSNs are formed by several sensor nodes communicating wirelessly, resulting in a self-organizing multi-hop network, intended to collaborate perception, acquisition, and processing of network coverage area information for monitoring and onward transmission to the observer. There is no question that GPS is currently the most mature of the most widely used positioning systems, yet it is only suitable for outdoor environments without shelter due to its high accuracy, real-time capability, anti-interference ability, etc [14]. Furthermore, the user nodes of GPS have many disadvantages such as high energy consumption, large volumes, high cost, and the need for extensive infrastructure. WSN nodes are being improved rapidly in terms of their ability to work around the clock both indoors and outdoors, as well as their low power and cost consumption [15]. The localization function of WSNs plays an important role in the effectiveness of sensor networks. WSNs consist of sensor nodes, sink nodes, and manager stations. Monitoring data is collected, transmitted, and processed by multi-hop sensor nodes randomly distributed in the monitoring area. Through the Internet or satellite management station, the data is transferred to the sink node. Sensor nodes are usually microembedded systems with limited processing power, storage capacity, and communication capabilities. The processing power, storage capacity, and communication capacity of the sink node are relatively strong. A sensor network is connected to the Internet and other external networks, protocol stacks are converted between the two, management nodes are monitored and data is collected. A sensor node consists of a sensing unit, a communication unit, a processing unit, an energy unit, and other components [16].

9 Conclusion Localization-technology services are drawing near an inflection factor. The continuing rollout of the technological infrastructure, the supply of LBS(Location-based services) packages, and the marketplace’s growing awareness of their capacity cost

Localization Technologies

37

must result in growing enterprise opportunities. However, there is nevertheless a high diploma of uncertainty in the LBS space. Challenges are emerging to the cell community operator-centric LBS version. Hardware groups, software companies, competing infrastructure technology (together with WiMax and satellite networks), and new competition from the pc and net industries are all vying for marketplace roles. Clients are becoming interested in regional services, but the uptake has been gradual. New LBS enterprise fashions and new techniques want to be taken into consideration. This bankruptcy evaluates the future of vicinity-based total offerings via an important assessment of the era, carrier applications, market developments, and strategic problems. As a result of the COVID-19 pandemic, the digital transformation of all activities and industries around us has been accelerated and intensified, which has directly affected localization firms.

10 Future Scope It is predicted that wireless localization technologies will become increasingly common in the future, allowing for more effective service to people who have a variety of needs. Tracking in real-time, activity recognition, health care, navigation, emergence detection, and locating targets of interest are among the applications they can deploy. New technological improvements in this domain can assist users in their daily lives, easing obligations inside the wireless care device, administrative center, and exercise sectors. Due to the great range of services that can be furnished by using utilizing IoT interior/outside localization has these days won a reputation. Destiny industrial applications include live sports effects tracking in actual time, which might be then utilized in live broadcasts and for staying having a bet and gambling. In addition to a developing array of industrial packages, localization generation can strengthen the place of business safety and protection inside the discipline of warehousing and distribution. It’s an ever-evolving industry with a lot of room for innovations and opportunities. Advancement in this technology would provide great benefits for scientific as well as personal purposes.

References 1. Arlind B, Ibraheem S, Abdulraqeb A, Qazwan A, Mardeni R (2020) An overview of indoor localization technologies: toward IoT navigation services. International symposium on telecommunication technologies (ISTT). IEEE Xplore, Malaysia, pp 76–81 2. Wan L, Guangjie H, Shu L, Feng N (2018) The critical patients localization algorithm using sparse representation for mixed signals in emergency healthcare system. IEEE Syst J 12(1):52– 63

38

A. Jain et al.

3. D’Ambrosio DJ, Bayouth J, Chetty IJ, Buyyounouski MK, Price RA Jr, Correa CR, Dilling TJ, Frankli GE, Xia P, Harris EE, Konski A (2012) Continuous localization technologies for radiotherapy delivery: report of the American society for radiation oncology emerging technology committee. Pract Radiat Oncol 145–50 4. https://www.linkedin.com/pulse/top-benefits-healthcare-localization-services-joydip-roycho wdhury 5. Asaad SM, Maghdid HS (2022) A comprehensive review of indoor/outdoor localization solutions in IoT era: research challenges and future perspectives. Comput Netw 212:1–28 6. Genevieve G, Francesco T, Robert H, Jonathan L, Daniele F (2016) Detection and tracking of moving objects hidden from view. Nat Photon 10:23–26 7. Trung-Dung V, Julien B, Olivier A (2011) Grid-based localization and local mapping with moving object detection and tracking. Inform Fusion 12(1):58–69 8. https://www.techtarget.com/whatis/definition/simultaneous-localization-and-mapping 9. Weipeng G, Linyi H, Shangsheng W, Zihong Y, Wanlin L, Chen Y, Ziyu L (2021) Robot localization and navigation using visible light positioning and SLAM fusion. J Lightwave Technol 39(22):7040–7051 10. Da Z, Feng X, Zhuo Y, Lin Y, Wenhong Z (2010) Localization technologies for indoor human tracking. In: International conference on future information technology. IEEE Explore, Korea, pp 1–6 11. Youngmok Y, Jingfu J, Namhoon K, Jeongyeon Y, Kim C (2012) Outdoor localization with optical navigation sensor, IMU and GPS. IEEE, Germany, pp 377–382 12. Yuntian Brian B, Suqin W, Hongren Wu, Kefei Z (2012) Overview of RFID-based indoor positioning technology. Geospatial Sci Res 2:1–10 13. Poulose A, Han D (2019) Hybrid indoor localization using IMU sensors and smartphone camera. Sensors 19:1–17 14. Mohammad Mahdi F, Saeed Bagheri S, Ensieh I, Bernabe L-B (2020) Sound source localization in wide-range outdoor environment using distributed sensor network. IEEE Sens J 20(4):2234– 2246 15. Ma XH, Bing ZG, Tang YQ (2011) Research on localization technology in wireless sensor networks. In: International conference on computer education, simulation and modeling. Springer, Heidelberg, pp 392–398 16. Huthaifa O, Wafa S, Omar O, Raed AA (2021) A review of indoor localization techniques and wireless technologies. Wireless Pers Commun 119:289–327

A Comparative Study Between Various Machine-Learning Algorithms Implemented for the Proper Detection of Fraudulent and Non-fraudulent Transactions Through Credit Card Surya Kant Pal, Nazneen Alam, Rita Roy , Preeti Jawla, and Subhodeep Mukherjee

Abstract E-commerce has dramatically increased the usage of credit cards for online and offline purchases. A considerable proliferation of transactions done through credit cards has led to a substantial and consequential surge in fraud. To combat the highly imbalanced dataset, we have performed oversampling on the dataset. This paper demonstrates various machine-learning algorithms for credit card fraud detection in this paper. We have also provided a detailed comparison between the performance of four algorithms. The algorithms used here in this paper were LightGBM, DNN, XG Boost, Random Forest, ANN and finally LSTM. As a result of the comparison, it is observed that LightGBM has the lowest accuracy score. So, we implemented other algorithms to improve our accuracy score. After a detailed comparison between six algorithms, it is observed from the graph that LSTM resulted in maximum accuracy as compared to other algorithms. It is observed that after running 30 epochs, LSTM resulted in a 0.9373 accuracy score. The performance of LSTM increases as we run it with more epochs, resulting in maximum accuracy. Keywords E-commerce · Machine learning algorithms · Credit card · Fraud detection · Random forest S. K. Pal · N. Alam Department of Mathematics, School of Basic Sciences and Research, Sharda University, Greater Noida 201306, India R. Roy Department of Computer Science and Engineering, GITAM Institute of Technology, GITAM (Deemed to be University), Visakhapatnam, Andhra Pradesh, India P. Jawla IIMT College of Engineering, Greater Noida, India S. Mukherjee (B) Department of Operations, GITAM SCHOOL OF BUSINESS, GITAM (Deemed to be University), Visakhapatnam, Andhra Pradesh, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. Misra et al. (eds.), Internet of Things (IoT): Key Digital Trends Shaping the Future, Lecture Notes in Networks and Systems 616, https://doi.org/10.1007/978-981-19-9719-8_4

39

40

S. K. Pal et al.

1 Introduction Credit cards are fine rectangular pieces of plastic or metal with a pre-set credit limit issued to users (cardholders) by a financial institution or service provider to forge cashless transactions [1]. A credit card is the most trivial way to acquire a line of credit. In general, it allows the account holders to forge purchases on credit without placing up cash for sale. A credit card can be divided into consumer and business credit cards. Credit cards are extensively used thanks to the universalization of ecommerce and the advancement of mobile intelligent devices. Online and offline transaction has become more convenient and accessible using credit cards, thus making them the most dominant mode of payment [2]. In the former type, they acquaint their card physically to forge payment. In the latter kind, the cardholder must withhold just some crucial card credentials, i.e. card number, card expiration date and secure code, which is obligatory for payment [3]. It can also be defined as deception or false representation of facts, intentionally confining decisive information or providing the perpetrator with erroneous statements for profit or to gain an unfair advantage [4]. Nowadays, online shopping is a day-to-day activity. Recently, due to high internetbased technology dependency, there has been an increase in transactions done through credit cards [5]. Currently, frauds done through a credit card are the biggest headache to companies and business hubs [6]. The cardholder has no further information about where and how their card is being used [7]. To minimize the illegal activity of credit card fraud, various systems, a wide range of models, and processed some preventive measures [8]. Traditional rule-based expert systems work satisfactorily to detect fraudulent transactions [9]. In this technologically advanced era, a tremendous flow of big data that these conventional rule-based expert systems cannot handle might underperform [10]. Many fintech industries have included machine-learning classification algorithms like SVM, neural networks, and a wide range of clustering and treebased approaches to detect fraudulent transactions [11]. This paper seeks to conduct a detailed comparative analysis of six different classifying algorithms with the parameter of our choice on a publicly available dataset. The best-performing algorithm is selected.

2 Review of Literature Credit cards permit cashless transactions and are a popular form of payment that is approved offline and online. Making payments and other exchanges is simple, convenient and fashionable [12]. Credit card fraud is increasing in combination with technological advances. Economic fraud is becoming more widespread as global communication technology progresses. Each year, billions of dollars in losses are noted as just a result of these fraud cases [13]. These operations are carried out

A Comparative Study Between Various Machine-Learning Algorithms …

41

so delicately that they depict open transaction records. As eCommerce grows in popularity, so too does the fraudulent activity. Credit card fraud is detected using a feed-forward neural network (FFNN). To handle commercial databases, the neural network is used for data mining. The rate of detecting fraud was 85% and 100% to identify legal transaction rates [17]—meta-learning methods to identify fraudulent transactions through credit cards. Artificially balanced data classify better than the skewed original credit card fraud data. Meta-learning with BAYES combined with classifiers of the highest accurate favourable rates learned from 50% fraud distribution is best. The information utilized in this article comes from a SQL database that comprises Visa Card transactions. This information is preprocessed before being used to identify credit card fraud. This paper’s data can be divided into three stages: training, prediction and fraud detection. Using a considerable amount of previous training data resulted in GNN producing fewer average training mistakes [14]. Meta classifier is used in credit card fraud detection problems. Naïve Bayesian, C4.5 and Back Propagation neural networks are the classification algorithms [15]. HMM can detect if a transaction through an incoming credit card is a fraudulent transaction or not. The valid transactions are not rejected and also demonstrate the necessity of understanding the spending profile of the cardholders [16]. Various fraud detector methods include Fuzzy Darwinian, Dempster, Bayesian theory and Hybridization of BLAST-SSAHA, HMM, ANN and BNN [17]. A neural network and a genetic algorithm detect credit card fraud. The genetic algorithm is used to create a credit card fraud detection neural network. The supervised learning feed-forward backpropagation algorithm (BPN) was employed for training and parameter selection (weight, number of layers, etc.). Data mining techniques, including ANN and LR, BN techniques, are combined to detect the problem of credit card fraud to reduce the bank’s or financial risks [18]. The three procedures outlined above in Python are applied to raw and preprocessed data. Training the features of fraudulent and non-fraudulent transactions deployed two types of random forests [19].

3 Research Methodology 3.1 Dataset This dataset solely comprises numerical variables that are the products obtained after using Principal Component Analysis (PCA) feature selection transformation. The original features and ground information about the dataset are not released due to privacy concerns. PCA has reduced everything except the parts—‘Time’ and ‘Amount’ for privacy concerns. There are 28 principal components resulting from a PCA Dimensionality reduction, mainly used for hiding customer identities and crucial data and V1, V2, …. V28.

42

S. K. Pal et al.

3.2 Data Acquisition and Pre-Processing The Class feature in the dataset, when it takes value 0, then it results in non-fraud (negative class) cases, and when it takes value 1, it results in Fraud (positive type) cases [20]. To address this problem of the unbalanced dataset, we need to use oversampling. Oversampling is the technique used to combat the issue of unbalanced classes in a dataset. The main aim of oversampling is to increase the minority class in training [21]. Oversampling results in no loss of information from the original training set. It is a tremendous advantage of oversampling [22, 23]. The number of samples in classes is distributed equally, i.e. 284,315 samples in both categories.

3.3 Methodologies DNN weighted neural network DNN stands for the deep neural network. It is an artificial neural network with more than two hidden (i.e. multiple) layers between the input and output layers. DNNs use mathematical models to process and manipulate the data in complex ways. DNN, as of definition, contains multiple hidden layers; that’s why the name ‘deep’ networks. Each node is connected with the other and has associated weights and thresholds [24]. Light GBM Light Gradient Boosting Machine (LightBGM) is a gradient boosting framework that uses decision tree-based learning algorithms to improve the model efficiency and lower memory usage for better accuracy and faster training speed. The leaf which results in the largest decrease loss is chosen by LightGBM [25]. After performing this fore-mentioned algorithm, a testing accuracy score of 0.9975 is observed. The following algorithms are performed to increase and develop our accuracy score [26]. Random Forest Random Forest, a supervised machine-learning algorithm, is a type of ensemble learning [26]. It is used in Classification and Regression problems [27]. At first, a Random Forest classifier is imported, and a random forest classifier is created with 100 estimators. Then, it is used to fit with X_train and y_train. After that, predictions are computed with the training and testing set of X. It has been observed that the accuracy score of the testing set is 0.996. For further better accuracy, the following algorithms are being applied. XG Boost XG Boost implies Extreme Gradient Boosting [28]. At first, a model from training data is constructed, and then the second model is built sequentially, in which errors were present in the first model. XG Boost is an implementation of Gradient Boosted

A Comparative Study Between Various Machine-Learning Algorithms …

43

decision trees. In this algorithm, models are created sequentially. Then as a final output, it gives a strong and precise model. After performing the algorithm, a 0.996 accuracy score is obtained in the testing dataset. Now, to get better accuracy, the following algorithm is being applied. ANN ANN stands for Artificial Neural Network. An ANN is generally a computational network based on biological neural networks which inherit or develop the human brain structure [29]. Then it is processed through the Batch Normalization () function to normalize the pixel values. Then, the Dropout () function is implied to drop the neurons at 0.3 probability. Then, the model is fitted with 300 epochs and 2048 batch size, and 0.996 accuracies are obtained. To obtain better accuracy, the following algorithm is performed. LSTM Long Short-term memory (LSTM) is a kind of recurrent neural network (RNN) in deep learning, which is known for tackling the long-term dependencies of RNN. It slightly modifies the information and can flow that data through the cell state [30]. LSTM selectively forgets things and can cling to long-time information. LSTM was mainly developed to combat the vanishing gradient problem while training traditional RNNs. Output gate and squashing function is used to yield important information. LSTM will perform better, and after running maximum epochs, maximum accuracy will be obtained, and that’s what the graph has depicted, i.e. 0.999 accuracies.

4 Interpretation This section discusses and tabulates the detailed interpretation of the results. The result of all six algorithms is tabulated below (Table 1). The following graph represents the Model accuracy in Fig. 1 and Model loss in Fig. 2, respectively, of the DNN algorithm. By convention, the validation dataset’s history is referred to as a test dataset because that is precisely what it is for the model. The model should be trained a little bit more, given the accuracy trend on datasets has been rising for the previous few epochs, according to the accuracy plot. Additionally, the model displays equivalent proficiency on datasets, indicating that it has not yet overlearned the training set. For accelerated learning, the loss function calculates the difference between the output of the LSTM model and the desired output during training. The user-specified validation data represents the desired result, and we have set the validation data to be 10% of the training data. As a result, overfitting can be avoided by stopping the model during training because the output of the training data is compared to the validation data at the end of each epoch. The graph below in Fig. 3 shows the Loss and Accuracy of the model of the LSTM algorithm.

44 Table 1 Result of the algorithms

S. K. Pal et al.

Algorithms

Train accuracy

Test accuracy

LightGBM

0.9986

0.9975

DNN

0.9980

0.9994

XG Boost

0.9997

0.9995

Random Forest

1

0.9996

ANN

0.9983

0.9996

LSTM

0.9998

0.9999

Fig. 1 Model accuracy

Fig. 2 Model loss

The graph in Fig. 4 is shown to represent the loss function evolution during training, accuracy evolution during training, precision evolution during training and recall development during training: The below graph in Fig. 5 compares test and train results between the forementioned algorithms. We can choose from many LSTM parameters, including

A Comparative Study Between Various Machine-Learning Algorithms …

45

Fig. 3 Loss and accuracy

Fig. 4 Loss function evolution during training, accuracy

learning rates and input and output biases. Thus, there is no need for precise modifications. With LSTMs, updating each weight is more straightforward than with Back Propagation Through Time (BPTT), reducing complexity to O(1). LSTM accuracy is greater than the XGBoost, Random forest and LightGBM.

46

S. K. Pal et al.

Fig. 5 Comparison of the test and train results between the fore-mentioned algorithms

5 Conclusion To improve security, it has become an essential task for financial companies to build a system that is efficient enough to detect credit card fraud. In this research, we demonstrate various machine-learning algorithms to detect credit card fraud and provide a comparative analysis of some of the algorithms. Recently, fraud done by credit card payments is the most popular and essential in every field of daily life. The algorithms used here in this paper were LightGBM, DNN, XG Boost, Random Forest, ANN and finally LSTM. As a result of the comparison, it is observed that LightGBM has the lowest accuracy score. So, we implemented other algorithms to improve our accuracy score. After a detailed comparison between six algorithms, it is observed from the graph that LSTM resulted in maximum accuracy as compared to other algorithms. It is observed that after running 30 epochs, LSTM resulted in a 0.9373 accuracy score. The performance of LSTM increases as we run it with more epochs, resulting in maximum accuracy. The comparative visualization of the performance of each algorithm is based on train and tests. LSTM was the best-performed algorithm we could ever get.

References 1. Izotova A, Valiullin A (2021) Comparison of Poisson process and machine learning algorithms approach for credit card fraud detection. Procedia Comput Sci 2. Alarfaj FK, Malik I, Khan HU, Almusallam N, Ramzan M, Ahmed M (2022) Credit card fraud detection using state-of-the-art machine learning and deep learning algorithms. IEEE Access 3. Rout M (2021) Analysis and comparison of credit card fraud detection using machine learning. In: Artificial intelligence and machine learning in business management. CRC Press 4. Roy R, Baral MM, Pal SK, Kumar S, Mukherjee S, Jana B (2022) Discussing the present, past, and future of Machine learning techniques in livestock farming: a systematic literature

A Comparative Study Between Various Machine-Learning Algorithms …

5. 6. 7.

8.

9.

10. 11. 12. 13.

14.

15. 16.

17. 18.

19. 20. 21.

22. 23.

24.

47

review. In: 2022 international conference on machine learning, Big Data, cloud and parallel computing. IEEE Asha, RB, Suresh Kumar KR (2021) Credit card fraud detection using artificial neural network. Glob Transit Proceed Procopiou A, Chen TM (2022) Security challenges and solutions in IoT networks for the smart cities. CRC Press, InInternet of Things Sadineni PK (2020) Detection of fraudulent transactions in credit card using machine learning algorithms. In: 2020 fourth international conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud). IEEE Mukherjee S, Baral MM, Pal SK, Chittipaka V, Roy R, Alam K (2022) Humanoid robot in healthcare: a systematic review and future research directions. In: 2022 international conference on machine learning, big data, cloud and parallel computing. IEEE Mukherjee S, Chittipaka V, Baral MM (2022) Addressing and modeling the challenges faced in the implementation of blockchain technology in the food and agriculture supply chain: a study using TOE framework. In: Blockchain technologies and applications for digital governance. IGI Global, pp 151–179 Malik EF, Khaw KW, Belaton B, Wong WP, Chew X (2022) Credit card fraud detection using a new hybrid machine learning architecture. Mathematics Al-Hashedi KG, Magalingam P (2021) Financial fraud detection applying data mining techniques: a comprehensive review from 2009 to 2019. Comput Sci Rev Mukherjee S, Chittipaka V (2021) Analyzing the adoption of intelligent agent technology in food supply chain management: empirical evidence. FIIB Bus Rev Malini N, Pushpa M (2017) Analysis on credit card fraud identification techniques based on KNN and outlier detection. In: 2017 third international conference on advances in electrical, electronics, information, communication and bio-informatics. IEEE Mukherjee S, Nagariya R, Baral MM, Patel BS, Chittipaka V, Rao KS, Rao UA (2022) Blockchain-based circular economy for achieving environmental sustainability in the Indian electronic MSMEs. Manage Environ Qual Int J Madhurya MJ, Gururaj HL, Soundarya BC, Vidyashree KP, Rajendra AB (2022) Exploratory analysis of credit card fraud detection using machine learning techniques. Glob Trans Proceedi Mukherjee S, Chittipaka V, Baral MM, Srivastava SC (2022) Can the supply chain of Indian SMEs adopt the technologies of industry 4.0?. In: Advances in mechanical and industrial engineering. CRC Press Roseline JF, Naidu GBSR, Pandi VS, alias Rajasree SA, Mageswari N (2022) Autonomous credit card fraud detection using machine learning approach. Comput Electr Eng Baral MM, Mukherjee S, Nagariya R, Patel BS, Pathak A, Chittipaka V (2022) Analysis of factors impacting firm performance of MSMEs: lessons learnt from COVID-19. Benchmark Int J Pal SK, Mukherjee S, Baral MM, Aggarwal S (2021) Problems of big data adoption in the healthcare industries. Asia Pac J Health Manage Itoo F, Singh S (2022) Comparison and analysis of logistic regression, Naïve Bayes and KNN machine learning algorithms for credit card fraud detection. Int J Inf Technol Carneiro EM, Dias LAV, Da Cunha AM, Mialaret LF (2015) Cluster analysis and artificial neural networks: a case study in credit card fraud detection. In: 2015 12th international conference on information technology-new generations. IEEE Pal SK, Baral MM, Mukherjee S, Venkataiah C, Jana B (2022) Analyzing the impact of supply chain innovation as a mediator for healthcare firms’ performance. Mater Today: Proceed Baral MM, Mukherjee S, Chittipaka V, Srivastava SC, Pal SK (2022) Critical components for food wastage in food supply chain management. In: Advances in mechanical and industrial engineering Lenka SR, Pant M, Barik RK, Patra SS, Dubey H (2021) Investigation into the efficacy of various machine learning techniques for mitigation in credit card fraud detection. In: Evolution in computational intelligence. Springer, Singapore

48

S. K. Pal et al.

25. Mukherjee S, Baral MM, Chittipaka V, Srivastava SC, Pal SK (2021) Discussing the impact of industry 4.0 in agriculture supply chain. In: Recent advances in smart manufacturing and materials. Springer, Singapore 26. Moumeni L, Saber M, Slimani I, Elfarissi I, Bougroun Z (2022) Machine learning for credit card fraud detection. Springer 27. Sadgali I, Nawal SA, Benabbou F (2019) Fraud detection in credit card transaction using machine learning techniques. In: 2019 1st international conference on smart systems and data science (ICSSD). IEEE 28. Mukherjee S, Baral MM, Chittipaka V, Pal SK (2022) A structural equation modelling approach to develop a resilient supply chain strategy for the COVID-19 disruptions. In: Handbook of research on supply chain resiliency, efficiency, and visibility in the post-pandemic era 29. Arfeen, AA, Khan BM (2022) Empirical analysis of machine learning algorithms on detection of fraudulent electronic fund transfer transactions. IETE J 30. Ashfaq T, Khalid R, Yahaya AS, Aslam S, Azar AT, Alsafari S, Hameed, IA (2022) A machine learning and blockchain based efficient fraud detection mechanism. Sensors

Smart Grid and Energy Management System Ishan Sharma, Priyal, Ananya Tyagi, Radhika Chawla, Aditya Khazanchi, Aaryan Bhatia, and Srishti Singh

Abstract This paper gives insights about smart grid and energy management systems. With the rise in industries, the consumption of electricity has been increasing everyday which will lead to deterioration of fossil fuels. To avoid this and with the “go green” mission, scientists are working on various technologies thus smart grid comes into play. Smart Grid is an electricity grid which is responsible for production, organization and utilization of energy using various techniques. The paper also outlines about the energy management system and CO2 emission. The paper also gives the ecological benefits and cost benefits regarding Energy Management System (EMS). Different types of EMS with different techniques are outlined in the paper. Keywords Energy management system · Smart grid system · Smart grid opportunity ecological benefits · Information system · Automation

1 Introduction There has been a heightened requirement for electricity, due to which several solutions for efficient energy consumption, environmental sustainability, generation of energy from renewable sources and new power distribution business models for active energy control have been considered. However, it is necessary to find solutions that can balance power demand at the levels of the home, block, and neighbor through intelligent control, and interactive energy monitoring. By reducing the number and duration of high demand periods grid maintenance expenses, energy losses, reduction in need of stand-by generators, and better quality and reliable services would be provided to the end user. “Smart grid” is a term used to name an electrical energy delivering system combining both digital technologies and long transmission networks to enhance the I. Sharma (B) · Priyal · A. Tyagi · R. Chawla · A. Khazanchi · A. Bhatia · S. Singh Technical Campus, Vivekananda Institute of Professional Studies, Delhi 110034, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. Misra et al. (eds.), Internet of Things (IoT): Key Digital Trends Shaping the Future, Lecture Notes in Networks and Systems 616, https://doi.org/10.1007/978-981-19-9719-8_5

49

50

I. Sharma et al.

energy consumption, and open up new processes for energy production and distribution. Reliability, efficiency, and safety improvements in power distribution are the three things that can be achieved through communication computing technologies. That’s why smart grid can optimize the energy efficiency of the grid being beneficial to the end users by coordinating as well as scheduling low priority home devices. This way, in order for the power consumption to benefit from the most effective energy sources and/or pricing at a certain period. By rapid detection and analysis of data coming from the distribution network, the smart grid could take correct measures to restore the power stability when required. Electrical grids are expected to lower the amount of CO2 emissions by reducing end user energy consumption during peak hours when electricity is generated through power plants that produce a lot of CO2 emissions. Communication is the fundamental key in the development of the grid as it gives real-time expense and incentive signals and allows the end user to become the source of energy production and to provide effective means for energy saving by giving them the choice to either consume the energy or give it to the grid. Individual electricity-consuming devices can turn off at appropriate times to flatten and eliminate spikes during peak hours by continuously monitoring the frequency of the power grid. Due to the fact that any power source can measure frequency on the grid, it is therefore feasible to include software that acts as a real-time frequency meter to monitor the behaviour of an appliance. Depending on its value, the appliance would react to the measured signal. The smart appliance would turn off until the frequency returned to the predicted level when the grid frequency decreased. The smart appliance would turn on and boost its power consumption when the grid frequency went over its normal value, taking advantage of a period of low power demand.

2 Smart Grid Smart Grid is an electrical supply network which is used in digital communications technology. It is used to detect changes in local usage. It is a key component in energy strategy. The contribution of smart grid in the electrical distribution system is to liberalize the energy market, reduce environmental pollution, promote sustainable and renewable development, increase energy efficiency and it helps in cost reduction. It also helps the distributed generation to grow and penetrate [1]. The Distributed Generation (DG) introduces benefits in electrical distributive systems which are: • • • •

Pliability and management of electrical load Range of Local load peaks Change of energy resource supply Depletion of electrical energy loss

Smart Grid and Energy Management System

51

Problems Caused by DG: • • • •

Short circuits Hampers the protection system Making Voltage regulation complex Unwanted islanding of MV system

Its main objective is to modify the functionality of electricity transmission to get a more user oriented service. It enables the attainment of 20/20/20 targets. It also helps to attain a competitive market environment by providing high security, power quality and by increasing the regulation of electricity supply. It also has 3 integrated parts such as Active Grids, Micro Grids and Virtual Utility.

2.1 Active Grids It is a network which plays a non-active role of supplying final consumers. It also plays a submissive role in operator controls and rules which the power requires. They coordinate in dispatching and voltage optimization.

2.2 Micro Grids It is a set of generators, storage systems and loads which are connected to work independently from the electrical grid. It recreates energy manufacturing and the distributed system. It provides a local control system that exchanges energy amongst the loads [2, 3].

2.3 Virtual Utility It is a virtual power plant that optimizes management and manages distributed energy resources. They include generators, loads and storage systems.

52

I. Sharma et al.

3 Smart Grid Overview 3.1 Architecture Smart grid is an intuitive web formulated on the principle of the latest gears, radar, and machinery to lead power resources and it increases the safety, authenticity, and effectiveness of the energy value chain [4]. The reason why smart grid is such a hit these days is its capacity to improve renewable Electricity Consumers (EC) from system and then govern the utilization of energy due to its dual flow of energy and data as shown below. Initial step in the smart grid is the power generation. It consists of power sources like. • NUCLEAR POWER • HYDRO POWER • RENEWABLE GENERATION SOLAR AND WIND. It communicates with power distribution. The foundations of power distribution are laid on proximity networks that interconnect the users with the client grid and pass on the data using high machinery. The last step is the power consumption. This step demands the EC both from households and industries. Some of the users produce energy from solar panel biomass etc. and it is the reason for managing the generation and utilization of the energy for evaluating the service. The framework of the smart grid is made of many layers that constantly link with each other [5]. • The component layer is used for physical devices which are used to get information and communication. • Communication layer is used for transferring data using modern methods • Information layer reports communication system which is used to interchange the information • Function layer gives logical functions free from physical architecture • Business layer gives the copy of business (Fig. 1). a. Smart Grid Opportunity It has been seen that more and more people are shifting to smart grid due to its authenticity, safety, well organized EMS with the increase in quality of power. It assures power quality management developed from smart transmission and positioning and pricing for real-time power markets [6, 7]. Thus, the electrical system should be worked to make it cost-effective. Smart grid brings value for utility and customers.

Smart Grid and Energy Management System

53

Fig. 1 Smart grid value chain

Utility- to assist the grid run better with good decisions, smart grid uses many technologies and monitoring systems for utilities to give more information. Thus, the advantages of smart grid are as follows. • • • •

Improving management of all the systems Decreasing energy dependence using renewable energy Perpetuating safety of grid Enhancing quality of resources.

Customers-grid offers many alternatives to its consumers as they utilize energy and more, they generate electricity by themselves. Optimization of electricity becomes convenient with the help of real-time communication. Grid also keeps a check on the consumption in real-time with the help of smart grid. b. Smart Grid Systems Smart grid systems depend on the latest and best figures with transmission framework for better production, organization and accommodation of energy which makes it cost-effective. Communication infrastructure is prime in keeping the transition between the various parts of the smart grid. A safe high-speed network for communication is

54

I. Sharma et al.

essential for a smart grid as it joins data to vehicle information. Communication infrastructure consists of Home Area Networks (HANs), Business Area Networks (BANs) and Neighborhood Area Networks (NANs) [8]. Network techniques are of two types. • Wireless technology-Smart grid is made of various gadgets and a lot of them transmit information through wireless medium. Some devices are facing speed and scalable issues with large amounts of data. To name a few out of IEEE 802.11, IEEE802.15.4, [9] Bluetooth, Infrared, ZIGBEE and Radio frequency, only IEEE802.15.4 and ZigBee are widely used as other devices have constraints related to energy consumption. Radio frequency lacks protocols and telecast signal and Bluetooth limits to a few nodes, so to overcome this Multiple-Input MultipleOutput (MIMO), Orthogonal Frequency-Division Multiplexing (OFDM) was developed recently • Wired technology- It is more efficient due to wide range; less delay and it works with higher capacity. However, it needs more investment than wireless. The under structure is based on fiber optics, and thus Power Line Communication (PLC) is mostly used by electrical companies as it helps them reduce money and has an edge over the energy power grids (Fig. 2).

Fig. 2 Smart grid communication technologies

Smart Grid and Energy Management System

55

4 Information Systems Information Systems (IS) are integral elements in the smart grid that interconnects for an expandable and efficient grid. As shown in Fig. 3 the utility information system carries and packs the data from substation or electricity consumers which consists of commercial, residential, and industrial gateway. It is then utilized to uproot the information about ES, state of the lines etc. Utility IS consisting of Supervisory Control and Data Acquisition (SCADA)—takes facts and figures from the utility field to supervise the electrical grid framework. It communicates with different IS and tells about the network. It consists of Geographic Information System (GIS), Advanced Metering Infrastructure (AMI) and Meter Data Management System (MDMS) which reports data from EC and swapping of data takes place. Demand Response Management System (DRMS) and Outage Management System (OMS) are important parts of the grid as they interconnect with other parts of the system in conjunction to provide the user with the best experience.

5 Energy Management Techniques in Smart Grid In the past few years with the advancement of technology, smart grid has been widely used in home automation. It is used to locally generate electrical energy, transmit, distribute, consume the required energy and store the rest for future application. Smart grid can be used in smart homes in the following ways.

5.1 Zigbee Network Interface with Microcontroller In this method, a smart home energy management system is developed using a Zigbee network interfaced with a microcontroller. Zigbee network and Peripheral Interface (PIC) microcontroller [10] is used to control and generate electrical energy from renewable resources such as solar panels and wind turbines. Also, ZigBee network is used to manage energy consumption. The data about the energy consumption of the home appliances and the energy generated by renewable resources is collected from servers using ZigBee sensors which are used to control energy consumption. Passive Infrared Sensor (PIR) is linked with a PIC microcontroller. If there is no human being in the home the PIR is ON and all the electrical devices are OFF and vice versa. This system can be controlled using an RS232 [11] interface. Details about the power generated and consumed are continuously uploaded to the server. During peak hours electric appliances are turned OFF to prevent overuse of energy. This technology helps to reduce electricity bills (Fig. 4).

56

I. Sharma et al.

Fig. 3 Information system in smart grid

5.2 Zigbee Network Interfaced with Field Programmable Gate Array (FPGA) The smart energy management technique is used to manage and enhance the energy efficiency of the photovoltaic power system. Using these techniques users can manage their energy consumption. Details of daily energy consumption are stored in a file and users can view this file anytime [12]. The Energy Management Center (EMC) is applied using Graphical User Interface (GUI) for monitoring and data logging of voltage and current of photovoltaic systems and utilities. Data is updated every second and shown on the voltmeter, ammeter, and power meter. It consists of three

Smart Grid and Energy Management System

57

Fig. 4 Block diagram for GSM based automatic trip control system

switches that are used for the selection of efficiency-based technique, user-based technique, and normal load control scheme respectively as shown Fig. 5.

5.3 Bluetooth Smart home automation can also be achieved by developing a lighting control system. In this, a small piconet is developed using microchip and Bluetooth modules [13]. Android software is used to monitor and control home lighting enabled through Bluetooth. The microchip controller checks the input every 500 ms from the android phone application and changes the lighting status as per the command received by the microchip controller (Fig. 6).

6 Energy Management Analysis in Smart Grid Increased energy consumption typically results in rising CO2 emissions and has a long-lasting effect on global warming; energy efficiency is a crucial problem for sustainable development initiatives. Energy management, in fact, encompasses all acts that may have an impact on the demand for energy, including those to reduce wasteful energy use and energy consumption on a large or medium scale [14].

58

I. Sharma et al.

Fig. 5 Block diagram for smart energy management system using FPGA

Fig. 6 Block diagram for Bluetooth module for lighting control system

The primary goals of energy management system are: • to understand the environmental benefits of energy management through connected household appliances with clear figures— to include stand-by usage as well as the fact that it prevents demand peaks and lowers energy losses.

Smart Grid and Energy Management System

59

• To calculate the reduction in CO2 emissions enabled by improved energy load and consumption control. • To address the concerns of a home energy management system’s costs and benefits. Daily routines for common human activities in individual homes follow rather predictable patterns. When a single house or a group of houses connected to the same power grid are taken into account, these patterns have a major impact on the utilization of electricity. The mentioned pattern’s synchronicity causes the immediate energy demand to vary significantly throughout the day, which has a detrimental impact on the overall effectiveness of the entire energy supply chain. This is mostly due to peak demand conditions, which need more current to travel via the electrical grid and through residential wiring systems, increasing energy waste due to power losses. Additionally, in contrast to flat energy profiles, energy demand is typically higher during the day than at night. At both the per-house and grid levels, Distributed Side Management (DSM) and Distributed on Site (DoS) control measures are used to reduce energy demand peaks and flatten the energy demand pattern. This is primarily accomplished by avoiding the synchronous activation of devices or loads, optimizing their activation to reach a greater overall level of energy efficiency, and carrying out this process in the most open setting possible with regard to the end users. Utilizing some type of local energy generation, such as Combined Heat and Power (CHP) systems, DoS actions are complementary to DSM actions and enable one to obtain even greater performance. These techniques need to be supported by particular home infrastructures or appliances, by appropriate control algorithms carried out by a centralized “Load/Energy Manager” device that can handle communication with the appliances and the electricity grid. The control measures for DSM that are offered here are implemented on the per-house load. These actions accomplish two goals: • Peak shaving, applied on a country-wide level: To reduce energy losses in transmission and distribution power systems and prevent issues with power quality and energy output, the utility manages the energy consumption of large groups of homes. The energy/load manager receives a directive from the utility to carry out the peak shaving function. • Power leveling applied on a per-house level: In order to reduce internal power losses, move appliance activation to times when energy costs are lower, and keep the instantaneous house load under the agreed-upon maximum, an energy/load management is responsible for leveling power consumptions below a specified threshold in each home. DoS control actions include: 1. Generation of electric energy from Photovoltaic (PV) systems, biomass, microwind turbines, etc.; 2. Generation of thermal energy from Instantaneous Gas Water Heaters (IGWH), Electric Storage Water Heaters (ESWH), heat pumps, solar panels, etc.;

60

I. Sharma et al.

3. Generation of both electric and thermal energy from Combined Heat and Power (CHP) and Combined Heat and Cool and Power (CHCP) systems. A house simulator has been built to assess instantaneous residential power losses and consumptions. The simulator can provide precise numerical data for a variety of scenarios, including homes with conventional appliances, homes with intelligent appliances, and homes that can implement DoS features. This method is designed because different homes—or even the same home on different days—display various appliances and load utilization patterns. Therefore, it is important to run a significant number of simulations using random inputs in order to determine power peaks and losses (one of the key components in this study) that result from the simultaneous activation of appliances. In order to closely mimic user behavior and traditions, these inputs (such as usage patterns or daily activation numbers) might be quantitatively described using statistical data or surveys. The simulator structure accepts information on the wiring configuration of an apartment, load profiles for the appliances, statistical patterns of appliance use, and present factors as input. Data is then automatically evaluated to extract the distribution and the most common data, as well as certain derived numbers that are used by the grid simulator. The data comprise information such as the peak power consumption and power losses. This scenario ensures maximum energy, financial, and CO2 emission savings. When the entire grid and industrial system is taken into account, this significant decrease in energy usage and power peaks results in additional savings.

6.1 Evaluation of Ecological Benefits The environmental benefits of implementing energy management control actions at home have been assessed in this section. To meet societal and ecological objectives, an examination has been mandated. Each nation has a unique electricity generation system, which has an impact on the emission of CO2 and other undesirable gasses. The carbon dioxide coefficient varies according to the type of power plant. This coefficient [15, 16] takes into account energy lifecycle emissions, which include emissions related to plant construction, fuel mining and processing, regular plant operations, disposal of used fuel and other waste by-products, and plant disposal. The reduction in the demand of electric energy will lead to deduction in the amount of CO2 emissions also further reducing the local impact on air pollution, public health and ecosystem.

Smart Grid and Energy Management System

61

Fig. 7 Cost–benefit analysis (CBA) results

6.2 Cost Benefit Analysis The analysis is split into two sections: 1. A private study of finances at the household level that only considers the costs and benefits that apply to one household; 2. A socioeconomic analysis of the community, at the country level, which considers all the costs and benefits associated with the introduction of the DSM and DoS actions [17]. These costs and benefits concern not only the households but also the electricity chain (from production to distribution), as well as the environment. The results of the cost–benefit analysis are reported in Fig. 7.

7 Conclusion With an increase in population and industrialization, the energy demand is also increasing which is giving rise to environmental problems caused because of energy production. The smart grid is a user-friendly web system designed with the idea of using the newest tools, machinery, and radar to manage power resources. It improves the security and efficiency of the energy value chain. Some users generate energy using solar panels, biofuels, etc. In this paper, we have shown ways in which smart grids can be used in energy management and their ecological and economic benefits. Future scope Smart grid is the future. If used with its full potential, it will meet the increasing energy demand as well as control the energy bills preventing blackouts. It will also prevent blackouts and will give clean electricity to the supplier. With absolute certainty, we can say that smart grid will be the need of the future and will be economical.

62

I. Sharma et al.

References 1. European Commission (2006) European smartgrids technology platform vision and strategy for Europe’s electricity networks of the future. Technical Report, European Commission, Louxembourg 2. Di Tommaso AO, Favuzza S, Genduso F, Miceli R, Ricco Galluzzo G (2010) Development of diagnostic systems for the fault tolerant operation of micro-grids. In: Proceedings of the international symposium on power electronics, electrical drives, automation and motion. Pisa, Italy. https://doi.org/10.1109/SPEEDAM.2010.5542262. cs:77956596092 3. Genduso F, Miceli R, Favuzza S (2013) A perspective on the future of distribution: Smart grids, state of the art, benefits and research plans. Energy Power Eng 5:36–42. https://doi.org/ 10.4236/epe.2013.51005 4. Wang W, Lu Z (2013) Cyber security in the smart grid: survey and challenges. Comput Netw 57:1344–1371 5. McGranaghan M, Schmitt DHL, Cleveland F, Lambert E (2016) Enabling the integrated grid: leveraging data to integrate distributed resources and customers. IEEE Power Energy Mag 14:83–93 6. Agarwal V, Tsoukalas LH (2010) Smart grids: importance of power quality. In: Proceedings of frst international conference on energy-efcient computing and networking. Berlin, pp 136–143 7. Amin SM (2011) Smart grid: overview, issues and opportunities. Advances and challenges in sensing, modeling, simulation, optimization and control. Eur J Control 17:547–567 8. Yan Y, Qian Y, Sharif H, Tipper D (2013) A survey on smart grid communication infrastructures: motivations, requirements and challenges. IEEE Commun Surv Tutor. 15:5–20 9. Xu J, Wang J, Xie S, Chen W, Kim J-U (2013) Study on intrusion detection policy for wireless sensor networks. Int J Secur Appl. 7:1–6 10. Sethuraman M, Jayanthi S (2014) Low cost and high efficiency smart HEMS by using Zigbee with MPPT techniques. Int J Adv Res Comput Sci Softw Eng 4(11):46–49 11. Ekshinge JV, Sonavane SS (2014) Smart home management using wireless sensor network. Int J Adv Res Electrtonics Commun Eng 3(4):453–456 12. Khattak YH, Mahmood T, Ullah I, Ullah H (2014) Smart energy management system for utility source and photovoltaic power system using FPGA and ZigBee. Am J Electr Power Energy Syst 3(5):86–94 13. Yan M, Shi H (2013) Smart living using Bluetooth—based android Smartphone. Int J Wirel Mob Netw 5(1):65–72 14. Rajan C (2003) Demand side management using expert system. In: Proceedings of the conference on convergent technologies for Asia-Pacific region TENCON 2003, Bangalore, India 15. IEEE Distribution Commitee (2003) Planning for effective distribution. IEEE Power Energy Mag 3:54–62 16. Newborough M, Augood P (1999) Demand-side management opportunities for the UK domestic sector. IEE Proc Gener Trans Distrib 146:283–293 17. International Energy Agency Data Services (2002) Guide to cost-benefit analysis of investment projects. Technical Report; European Commission, Paris, France

To Foresight and Formulate Development (FFD) of Robot of Things (RoT) and Drone of Things (DoT) for Revolutionizing Agriculture Ecosystem Chandrani Singh, Sunil Khilari, and Anchal Koshta

Abstract Robot of Things (RoT) is a robotic middleware platform and solution for intelligent cognitive systems for smart agriculture which is handling repetitive, mundane, and complicated tasks. Mostly, robots are deployed to perform repetitive and mundane tasks which are typically done by humans working on lower pay and on grass root of agriculture sector. Robot of Things (RoT) platform, increasingly trending in the world of robotics, is becoming commonplace in agricultural robotics. Robots pick apples, gather strawberries, harvest crops, and strip away weeds. This RoT platform and solution can communicate and transmit signals and data to other devices like sensors and intelligent switches and hubs. This cutting-edge technology can be revolutionizing today’s agriculture in innovative ways. Drones gather aerial images that help farmers quickly assess crop health and many more functionalities. Due to less attention to agriculture domain as compared to other domains to an aim agriculture farming sustainable. This paper will present case scenarios and an innovative solution to make drone technology has to be utilized in this domain. With respect to Drone of Things (DoT), this paper will propose interesting and costeffective solutions to increase the efficiency and productivity of various stakeholders of agriculture. The stakeholders who will be facilitated through the usage of DoT & RoT are the farmers, start-ups, Farmepreneurs, governments, Agri-entrepreneurs, equipment suppliers, agronomists, etc. In order to emphasize and implement, the IoT way of doing business which will encompass research by the researcher to showcase that the issue of sustainability can be addressed by usage of the new and innovative mechanism. Keywords Robot of Things (RoT) · Drone of Things (DoT) · Internet of Things (IoT) · Intelligent Cognitive System C. Singh · S. Khilari (B) Sinhgad Institute of Management, Pune, India e-mail: [email protected] A. Koshta ABCROB Technologies Pvt Ltd, Jabalpur, M.P, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. Misra et al. (eds.), Internet of Things (IoT): Key Digital Trends Shaping the Future, Lecture Notes in Networks and Systems 616, https://doi.org/10.1007/978-981-19-9719-8_6

63

64

C. Singh et al.

1 Role of Information and Communication Technology (IT) in Agriculture At the global level, there are many Information and Communication Technology applications in agriculture, and forums and active organizations that facilitate the exchange of information and knowledge between farmers, policy makers, agricultural and implementation experts and organizations. These applications render cognizance about how ICT can contribute to the agriculture sector. The forums and programs raise awareness on the implementation of ICT across agribusinesses and increase the focus on the need to create a progressive and conducive environment to alleviate growth. They promote the exchange of knowledge through the use of ICT applications, provide a platform for enriched communication between staff, and contribute to the execution of projects that can add desired value to the society. They also promote cooperation and collaboration between international organizations and ensure usage of advanced technologies. Despite these efforts, new strategies are failing to evolve and mature, especially in developing countries. Many global equipment manufacturers, engaged in roll out of new ICT applications specific to agriculture, provide services to their online customers that are mobile compliant (UNCTAD, 2021). These companies have made significant investments in ICT systems, and through their ICT services gather information about their clients’ farming activities, procedures and examine them, and share insights with the former to enhance their farm’s productivity. In many cases they provide greater benefits, such as conserving natural resources [2, 6, 18].

1.1 Current Status of on Agriculture Farming Issues Commodity marketing and the provision of information sharing and exchange to farmers through ICT generates profits for companies and these innovations are often protected by patents, trademarks, and copyrights motivating organizations to engage more in research and development. However, many aspects of knowledge dissemination, benefit the community and with proper governmental intervention aspects like the negative impact of climate change on agriculture needs to be taken into serious cognizance for the benefit of future generations [20]. International organizations have demonstrated their unique role in producing and disseminating this information to governments, recognizing its positive relevance, and ensuring the value of teaching, grooming, public consciousness, public engagement, public admittance to information, and cooperation at all levels of change applicable to agriculture sector.

To Foresight and Formulate Development (FFD) of Robot of Things …

65

1.2 Projects for Monitoring Climate Change Presently, ICT applications specific to agribusinesses and their adaptation to environmental changes are available on a large or small scale. They extend from major projects such as satellites, meteorological data evaluation, and large IOT based networks to projects that enhance cognizance of endangered assemblages and individuals to help them determine and manage their adaptability requirements. Climate change will pose a major obstacle to the most widely used agricultural systems [16]. Agricultural activities, based on soil, water sourcing, distribution and management and cuss control, combined with better market access, can lead to amendments in agricultural production, climate change and increased resilience, either due to change in environmental conditions or due to resource distribution and utilization.

1.3 Restrictions to Strengthen Sustainable Agriculture Technology to strengthen sustainable industries requires knowledge [1]. Climate Smart Agriculture manages many of the goals of growth and development in the field of agriculture under certain constraints, while lowering and eliminating greenhouse gases [30]. It requires information of numerous characteristics of farming considering, for example, minor soil disturbances, lasting soil coverage, and crop rotation. Construction of indigenous knowledge by international establishments and research institutes working can strengthen sustainable production. It also requires a large number of well-trained agents to reach native and remote farmers, contact them, and guide them. Utilizing state-of-the-art ICT applications to promote sustainable agricultural development has been the key initiative that is being promoted and encouraged by governments of various countries. New claims such as visual enhancement programs, online forums and/or websites, can empower a farmer to share the expert information. Building on knowledgesharing communities and integrated ICT systems can be achieved through a collaborative technology platform, and with intervention of agri-tech labs. This collaborative initiative will help in developing high-quality testing applications, and create prototypes (Sensor, 2022), [24] test solutions that can be adapted within a multistakeholder environment [11]. The Information Technology professionals in the labs will enable the linking of farmers, farmers’ clusters, claim developers, agronomists, and agents using a concept known as Innovation Lab Approach. This can ensure gathering and distribution of timely and precise information about the weather, markets, and rates. Some of the broader areas where ICT will continue to play an important role are as follows: – Improving market access – Improving agricultural production

66

C. Singh et al.

– Capacity building and empowerment ICT in agriculture enables farmers to choose what to invest and how and incur costs associated with identifying and evaluating an agent who will negotiate, and monitor the agreement. In many developing countries, agriculture makes up the bulk of the economy; there is a need for more agents to reach out to local farmers, contact them, and recommend them the new technologies that may be critical to their farming. ICTs have the potential to decrease these cost using technology. For example, Kenya’s DrumNet assisted to connect financial organizations, farmers group, retailers, and consumers with agricultural products through a small cash loan system (Marina Ruete, 2015). Farmers have gained access to inputs from local suppliers through a pre-established credit mark from banks, where DrumNet has delivered the bank with credit score for each farmer. Increased usage of digital technology has created paybacks for everyone through easy communication and information sharing, as well as increasing community engagement [14].

2 Role of Internet of Things in Agriculture From high-tech greenhouses to cloud seeding, agricultural robots are enabling farmers to fill labor insufficiency and the shelves of the supermarket [12]. While Internet of Things aim at restricting human participation in any communication network, it is evident that either devices are made to communicate with each other or devices that can be an intermediary between two entities may be made to connect the two. Robotics on the other hand also considerably reduces human engagement. So both, when they co-exist together become Internet of Robotic Things (IoRT), which has become an established field of research that describes the collaboration between these two segments. These two segments have lots of overlaps and an amalgamation of the said disciplines is anticipated to do tremendous societal good. In recent times farming has become an automatized procedure where sensors, computers, robots and drones and other machines can execute the work of the laborer [10]. On the other hand, the Internet of Drone Things (IoDT) is visualized as progression in the Drone technology through Internet of Things, Computer vision, Cloud Computing, Wireless Communication, Data Science, and end-to-end security techniques [17]. The current era sees significant growth in use of drones in miscellaneous areas. With rise in IoDT-based operations, monitoring, surveillance, and investigation can happen at ease. Further with usage of global positioning system (GPS) machinery, along with geographic information system (GIS) techniques and tools, precision agriculture practices have advanced significantly enabling micro-monitoring and mapping of yield from plants (esterháziZs. PeczeZs. Stépán. 2013) [21]. Fertilizer type and quantity required or crop disease identification are a few monitoring activities that can be done using Internet of Drone Things. Smart agriculture is becoming a major necessity and is attracting big investors. Therefore, the agricultural sector has begun to adapt to the latest and most advanced technologies such as Internet

To Foresight and Formulate Development (FFD) of Robot of Things …

67

of Things (IoT), Drone of Things (DoT), Robot of Things (RoT), Data Science, Machine Learning (ML), Deep Learning (DL), Cloud Computing (CC), Artificial Intelligence (AI), etc. Among them, the robotic centered and drone established environmental observing and field management arrangement has been extensively used in contemporary farming everywhere in the world [29].

2.1 Internet of Things (IoT) Applications in Agriculture The use of Internet of Things (IoT) in agriculture has the prospective to transform the world. The world’s population is poised to hit the 9.6 billion mark by 2050 (Poonam Gorade, 2018). Therefore, to meet the hunger of the people, the modern agriculturist must utilize the power of IoT [15]. The use of smart IoT-based agriculture not only focuses on huge scale farming activities such as organic and open farming, but also enables efficient water usage, required plant and soil management and many more [13]. The various IoT applications in agriculture are as follows: the following eight kinds of field robots have become quite standard (Table 1). A comparative study of the capabilities of Robots and Drones utilized to a large extent by the agricultural enterprises is described for the readers to have a fair idea on the extent of deployment of these modern machineries in the agriculture segment (Oannis Malounas, E. Fthymios Rodias, Christoph Hellmann Santos and Erik Pekkeriet, Sensors-2020).

3 Challenges and Opportunities for Sustainable Development of Agricultural Eco-System 3.1 Challenges and Opportunities Strengthening agricultural systems is less about operational recommendations and policies and more about impending sustainability. Food production needs to be addressed through rapid population growth, resource utilization, soil degradation, land use reduction, and growing water scarcity. Except for the network performance and the bandwidth availability that are greatly enhanced, the operations of digital farming will continue a challenge as cloud-based infrastructure also needs durability in terms of transfer and storage [19]. Lack of access to reliable and timely market information for farmers, lack of forecasting of service delivery, poor and inefficient service delivery, inadequate refrigeration facilities and lack of adequate food processing facilities, poor communication between farmers and consumers are some of the issues that hinder sustainability. A farmer’s business decision can be quite complex aligning to global economic conditions, such as commodity, volatility

68

C. Singh et al.

Table 1 Types of field robots and their usage [5], [7–9], [23, 26, 28] Serial number

Type of robot of things (RoT)

Image

Usage

1

Pollution control robot

These field robots are nowadays armed to observe the pollution at the ground level. These robots quantify carbon dioxide and nitrous oxide discharges so that farmers could shrink their soil and land footprint

2

Weed controlling robot

These field robots can autonomously model a farm and deliver targeted sprays of herbicides to eliminate weeds. This process decreases crops’ exposure to herbicides and assists to avoid the increase of herbicide-resistant weeds

3

Precision agriculture robot

These field robots are used on small farms and are used to autonomously monitor soil respiration, photosynthetic movement, leaf area indexes (LAI), and other biological issues

4

Livestock ranching robot

This is a novel kind of field robot that is used to ruck farm animal on big cattle farms. These robots also proctor the animals and ensure they are growing and have adequate arena to feed and move

5

Crop harvesting robot [27]

This robot is used for harvesting plants, particular field robots can work round the clock for quicker harvesting, in some circumstances finishing the load of allocated task quickly

6

Seeding robot

Planting and seeding an evolving exercise can be done by field robots with 3D vision systems who can with precision plant and seed crops for better development, mainly for lettuce farming and vineyards

7

Fruit harvesting robot

These field robots are used to harvest fruit in addition to crops using advanced vision systems to recognize fruits and pluck them without destroying them

(continued)

To Foresight and Formulate Development (FFD) of Robot of Things …

69

Table 1 (continued) Serial number

Type of robot of things (RoT)

Image

Usage

8

Nursery automation robot

These field robots can be used in nurseries, move plants around large greenhouses and are very useful against growing labor shortage

9

Equipment navigation robot

These types of robots are remote controlled and heavy tilling equipment’s can be steered mechanically from home using Global Positioning System. These machines work with precision, and adjust when they observe change in terrains

10

Material handling robot

These types of robots can lift heavy instruments and execute tasks like plant spacing with great precision therefore improving the space and plant quality, and dropping manufacturing costs

and trade issues, and the fact that yields can be affected by weather, diseases, etc. Following are some biggest problems faced by farmers in developing countries in contrast to the relevant opportunities for technology innovations in agriculture sector (Table 2). Opportunities for development include diversified agricultural conditions, high domestic demand and a large domestic market, fertile soil. Other challenges that Table 2 Challenges and Opportunity for sustainable development of agricultural [3, 4] Sr

Challenges

Opportunities

1

Small and fragmented land-holdings

Adopting technology and innovation

2

Irrigation

Get IP for farming prior art

3

Soil erosion

Adoption of mechanism Big Data is only for Big Farming (Smart Data)

4

Agricultural marketing

Best practices—right products, at the right rates, and at the right time

5

Inadequate storage facilities

ROI for suitable investment

6

Changing tastes and expectations

Organic fertilizer development

7

Adopt and learn new technologies

Hydroponic Agriculture

8

Inspire young people to stay in rural areas

Harnessing Rainwater

9

Scarcity of capital

Start-up & Agropreneurship

10

Actionable weather forecasts

Precision Agriculture

70

C. Singh et al.

farmers face in adopting the latest technologies in agriculture are as follows: Lack of Infrastructure—Farmers using IoT technology will not be able to use the benefits of this technology due to a not-so-strong communication infrastructure. Farms are located in remote areas with weak internet access in many parts of the world.

3.2 Case: Real-Time Time Series of Remotely Sensing Images of Crop and Soil Using R-Language R-Language (RStudio Ver: 1.0.153): a free and open source software (FOSSE), with no license constraints, is cross platform and can find many users from the agritech zone. This platform enables the analysis of crop and soil images. The authors using the skeletal application source tried to present a case of real-time crop monitoring. Temporal datasets of 30 different fields for 30 farmer’s crops with 11 spatial parameters have been considered for this case.

3.3 The Software Application Source Code #Clear the screen shell("cls") #set source data path location setwd("c://agri/") getwd() #Load the package for excel data bulk upload using FTP server install.packages("readxl") library(readxl) farmpath=read_excel("c:/agri/Farming.xls") print(farmpath) plot(farmpath$Max_Temp,farmpath$Soil_Temp,zlim=c(0,1),col=c("red","blue")) library(ggplot2) ggplot(farmpath, aes(Max_Temp, Soil_Temp, group = 1,),)+ geom_bar(stat = "identity")+ geom_line(color = "green")+ geom_point(color = "blue")

To Foresight and Formulate Development (FFD) of Robot of Things …

71

3.4 Runtime Output

Data Source: Bulk import from excel-structured format after refining the data gathered by Drone. Parameter values captured by drone: FarmerName, Location of Crop, Year, Month, Day, Date-wise minimum temperature, Date-wise maximum temperature, Date-wise soil temperature, Latitude, Longitude, Image (Table 3). Further, this can be analyzed with the help of RStudio package “raster” and image captured by drone. Also assign date code to each image captured and assign value to the variable of raster package of R and raster can visualize data by using the plot function. #"raster package" lines show how install the package install.packages(’raster’) #Load the package require(raster) fHR , only depends on the values Oˆ is < D, Pb , ET − sT · r · wT > and f . Since the private w has a uniform distribution in ⎡DS (w, E). Then the values must be disseminated by the algorithm of collision resistance and ˆ to an identical chance of identifying a collision the revised values determined by O as the initial algorithm’s values. Thus, for the following event, we examine the probability that the original method will detect a collision: E'T − s'T · r' · wT √= S'g .

(3)

Now, the probability of eq.(3) is determined by Pr(3) = Pr[(3) ∧ E = E' ]+Pr[(3) ∧ E √= E' ∧ (S'g √=E'T − sT · r · wT )] + Pr[(3) ∧ E √= E' ∧ (S'g = E'T −sT · r · wT )]. If (E = E' ), then (3) ∧ (E = E' ) holds. Thus, Pr [(3) ∧ E = E' ] ≥ Pr [(2) ∧ E = E' ] ≥ [(2) ∧ E = E' ] ·

(1 − α) . (2 − α)

If (E √= E' ∧ (Sg' √= E 'T − s T · r · w T ) ∧ ( f = 0)), then w = w' , r = r' , and s = s' .Then, we have Pr [(3) ∧ (E √= E' ) ∧ (Sg' = E 'T − s T · r · w T )] ≥ Pr [E √= E' ∧ (Sg' = E 'T − s T · r · w T ) ∧ ( f = 0)] (1−α) .Pr [(2) ∧ E √= E' ∧ (Sg' = E 'T − s T · r · w T )]. ≥ (2−α) Hence, if E √= E' and S'g = E'T − sT · r · wT then E'T − s'T · r' · wT √= S'g are similar to E'T − s'T · r' · wT √= E'T − sT · r · wT , respectively. Therefore, for every w' , / ⎡DS (w, E' ). where w'T · DT = wT · DT and E'T − s'T · r' · wT √= S'g are true iff w' ∈ m m ¯ Let W ⊆ W, S ⊆ Z q , and R ⊆ Z q be the set of all private keys of w, r, and s, respectively, such that ∀E √= E' |⎡DN (w, E) ∩ ⎡DS (w, E' ) ≤ α · ⎡DS (w, E)|.

Quantum-defended Digital Signature on Lattice for IoT-enabled Systems

323

¯ , r ∈ Z qm , s ∈ Z qm ] ≥ α. As a Under (α, β)-Hiding property, we have Pr [w ∈ W result, based on the proportion of the success is achieved: ¯ ∧ r ∈ R ∧ s ∈ S ∧ ( f = 1) (4) (E √= E' ) ∧ (Sg' = E 'T − s T · r · w T ) ∧ w ∈ W and by utilizing f ’s independence, we have Pr (4) = Pr [( f = 1)].Pr [(E √= E' ) ∧ (Sg' = M 'T − s T · r · w T )∧ ¯ ) ∧ (r ∈ R) ∧ (s ∈ S)] (w ∈ W Pr [(E√=E' )∧(Sg' =E 'T −s T ·r·w T )]−Pr [(w∈W / ¯ )∧(r∈ / R)(s∈S)] / ≥ (2−α) ≥

Pr [(2)∧(E√=E' )∧(Sg' =E 'T −s T ·r·w T )]−1+β . 2−α

Thus, we have Pr [(3)(4)] = Pr [w' ∈ / ⎡DS (w, E' )|(4)] ≥ 1 − maxD,w∈W¯ ,s∈S,r∈R,E√=E' ≥ 1 − α.

|⎡DS (w,E)∩⎡DS (w,E' | |⎡DS (w,E)|

Hence, Pr [(3) ∧ (E √= E' ) ∧ (Sg' = M 'T − s T · r · w T )] ≥ Pr [(3) ∧ (4)] ≥

Pr [(3) ∧ (E √= E' )] ∧ (Sg' = E 'T − s T · r · w T ) − 1 + β 2−α

· (1 − α).

Adding the above cases, we get Pr [(3)] ≥ (Pr [(2)] − 1 + β). = (η − 1 + β).

1−α 2−α

1−α . 2−α

Therefore, the values of Pr (3) and η are negligible if closure, collision resistance, and (α, β)-hiding features satisfy and the signature scheme is strongly unbreakable ˆ against O.

6 Conclusion The signature is among the most secure pieces of information used in online transactions. In IoT systems, the signature has become an important instrument. Traditional signature schemes have been subjected to several attacks as a result of quantum

324

D. S. Gupta et al.

computing and grid technologies. In recent years, a lattice has emerged with a rich platform, as the lattice has become a potent weapon in the face of attacks. So, in this article, we have devised a lattice-based digital signature protocol for IoT-enabled systems. The suggested protocol is based on lattice-based cryptography. The security assessment of these suggested protocols reveals that they are capable of withstanding quantum computing technology. Thus, by replacing the security of existing protocols, the suggested protocols create a new security basis in the time of quantum computers.

References 1. Diffie W, Hellman ME (1976) New directions in cryptography. IEEE Trans Inf Theory 22(6):644–654 2. Gupta DS (2022) A mutual authentication and key agreement protocol for smart grid environment using lattice. In: Proceedings of the international conference on computational intelligence and sustainable technologies 2022. Springer, Singapore, pp 239–248 3. Rivest RL, Shamir A, Adleman LM (1978) A method for obtaining digital signatures and public-key cryptosystems. Commun ACM 21(2):120–126 4. Zhang H, Yuan Z, Wen Q-Y (2007) A digital signature schemes without using one-way hash and message redundancy and its application on key agreement. In: IFIP international conference on network and parallel computing workshops 5. Chang CC, Chang YF (2004) Signing a digital signature without using one-way hash functions and message redundancy schemes. IEEE Commun Lett 485–487 6. Qin Y, Li C, Xu S (2010) A fast ECC digital signature based on DSP. In: International conference on computer application and system modeling, vol 7. IEEE 7. Wang G (2005) An abuse-free fair contract signing protocol based on the RSA signature. In: Proceedings of the 14th international conference on world wide web. pp 412–421 8. Wang Guilin (2009) An abuse-free fair contract-signing protocol based on the RSA signature. IEEE Trans Inf Forensics Secur 5(1):158–168 9. Prabu M, Shanmugalakshmi R (2009). A comparative analysis of signature schemes in a new approach of variant on ECDSA. In: International conference on information and multimedia technology. IEEE, pp 491–494 10. Khalil MN, Hau YW (2008) Implementation of SHA-2 hash function for a digital signature System-on-Chip in FPGA. In: 2008 international conference on electronic design. IEEE, pp 1–6 11. Even S, Goldreich O, Micali S (1996) On-line/off-line digital signatures. J Cryptol 9(1):35–67 12. Bleichenbacher D, Maurer UM (1996) On the efficiency of one-time digital signatures. In: ASIACRYPT. pp 145–158 13. Hevia A, Micciancio D (2002) The provable security of graph-based onetime signatures and extensions to algebraic signature schemes. In: ASIACRYPT. pp 379–396 14. Gupta DS, Islam SH, Obaidat MS, Hsiao KF (2020) A novel identity-based deniable authentication protocol using bilinear pairings for mobile ad hoc networks. Adhoc & Sens Wirel Netw 47 15. Gennaro R, Gertner Y, Katz J, Trevisan L (2005) Bounds on the efficiency of generic cryptographic constructions. SIAM J Comput 35(1):217–246 16. Ajtai M (1996) Generating hard instances of lattice problems (extended abstract). In: STOC. pp 99–108 17. Micciancio D (2007) Generalized compact knapsacks, cyclic lattices, and efficient one way functions. Comput Complex 16(4):365–411

Quantum-defended Digital Signature on Lattice for IoT-enabled Systems

325

18. Peikert C, Rosen A (2006) Efficient collision-resistant hashing from worst-case assumptions on cyclic lattices. In: TCC. pp 145–166 19. Gupta DS, Islam SH, Obaidat MS, Vijayakumar P, Kumar N, Park Y (2020) A provably secure and lightweight identity-based two-party authenticated key agreement protocol for IIoT environments. IEEE Syst J 15(2):1732–1741 20. Lyubashevsky V, Micciancio D (2006) Generalized compact knapsacks are collision resistant. ICALP (2):144–155 21. Gupta DS, Biswas GP (2018) Design of lattice-based ElGamal encryption and signature schemes using SIS problem. Trans Emerging Tel Tech 29(6):e3255 22. Gupta DS, Ray S, Singh T, Kumari M (2022) Post-quantum lightweight identity-based twoparty authenticated key exchange protocol for Internet of Vehicles with probable security. Comput Commun 181:69–79 23. Lyubashevsky V, Micciancio D (2008) Asymptotically efficient lattice- based digital signature. In: Proceedings of the conference on theory cryptography. pp 37–54 24. Gupta DS, Biswas GP (2018) A novel and efficient lattice—based authenticated key exchange protocol in C–K model. Int J Commun Syst 31(3):e3473 25. Gupta DS, Karati A, Saad W, da Costa DB (2022) Quantum-defended blockchain-assisted data authentication protocol for internet of vehicles. IEEE Trans Veh Technol 71(3):3255–3266 26. Mukherjee S, Gupta DS, Biswas GP (2019) An efficient and batch verifiable conditional privacy-preserving authentication scheme for VANETs using lattice. Computing 101(12):1763–1788

Securing Digital Ownership Using Non-Fungible Tokens(NFTs), an Application of BlockChain Technology Suhas Harbola , Jyotsna Yadav , Rahul Johari , Ekta Verma , and Deo Prakash Vidyarthi

Abstract Non-Fungible Token(NFT) is the latest tech-jargon buzzing in IT Industry. The continuous work in the area of securing contents with BlockChain has given rise to Non-Fungible Tokens(NFTs). One of the major problems of digital assets is the claim of ownership, as to how to verify ownership of digital assets (like audios, music videos, digital art, streaming contents on the web, etc.) which is accessed/copied by billions. NFTs are unique and distinct from each other thus securing the ownership. These tokens are used to represent a digital asset and contain a proof of ownership. As the BlockChain technology grows, its applications have been designed, developed and deployed in a variety of industries. In this research, we intend to give an idea of NFT and related technology. Not only this, an effort has been made to explore Indian NFT marketplaces, and the process of selling and buying the NFTs has also been discussed. Keywords Cryptoapp · Non-fungible token · BlockChain · NFT trading · Ownership · NFT · Digital assets

S. Harbola (B) · J. Yadav Computer Vision and Image Processing lab, University School of Information, Communication and Technology (USICT), Guru Gobind Singh Indraprastha University, Sector-16C, Dwarka, Delhi, India e-mail: [email protected] R. Johari SWINGER : Security, Wireless, IoT Network Group of Engineering and Research, University School of Information, Communication and Technology (USICT), Guru Gobind Singh Indraprastha University, Sector-16C, Dwarka, Delhi, India D. P. Vidyarthi School of Computer and Systems Sciences, Parallel and Distributed System Lab JNU, Delhi, India S. Harbola · E. Verma National Informatics Centre, New Delhi, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. Misra et al. (eds.), Internet of Things (IoT): Key Digital Trends Shaping the Future, Lecture Notes in Networks and Systems 616, https://doi.org/10.1007/978-981-19-9719-8_27

327

328

S. Harbola et al.

1 Introduction Even before the pandemic hit the world, people were connected on the web. But after the pandemic, almost every person is digitally connected sharing contents and within seconds the data moves around the globe. The digital data be it a document, an artwork, music or a video clip which is unique in nature and is usually an asset to an individual or an organisation is known as a digital asset. The data flows through the Internet, and it can be shared/re-shared several times and can be viewed by millions and used by anyone. At times an artwork of an artist is used by somebody in another corner of the world without even giving credit to the artist leading to the issues of plagiarism and copyright violations. Most of the digital assets are intended for a larger audience mainly used by the public for entertainment like a game, music, etc. Encrypting these assets would reduce their use and popularity thus affecting the overall value. As the sharing of digital assets like music, video clips or a document increases, the question of its ownership rights comes into the picture. Blockchain is a cutting-edge technology that is evolving and proving its application in a variety of areas. The Non-Fungible Tokens (NFTs) generated by BlockChain technology are used as a proof of ownership for digital assets. The digital assets are unique in nature, similarly NFTs are unique and these can’t be replaced with each other as their values differ like music can’t be compared with the video clip. A bitcoin is replaceable with another bitcoin. Similarly, a hundred rupee note can be replaced with another 100 rs note as their value is the same thus making them fungible. NFTs are maintained in Ethereum BlockChain. The NFTs can be produced for any digital asset like the following:• • • • • • •

Art GIFs Videos and sports highlights Collectibles Virtual avatars and video game skins Music Trump Cards

The NFTs are unique and thus are scarce in nature due to their high and continuous demand in the market. As well known, a scarce resource always has a growing value. There are several marketplaces where NFTs are traded for cryptocurrency or real currency.

2 Literature Survey As the NFT market grew, new apps or platforms for NFT trading emerged to improve the overall performance of the NFT trading system; one such platform with evaluation mechanism was proposed by the authors of [2]. The results by Wang [16]

Securing Digital Ownership Using Non-Fungible …

329

clearly show anchoring effects and loss aversion in the NFT market at both seller and buyer sides. Yu Wang encountered that in the NFT market both buyer and seller are not acting or following any strategy. Chirtoaca et al. [5] proposed a framework of smart contact with the ERC721 standard including features and extensions which are commonly used by the developers. Kumar et al. [6] propose a decentralised education model where educational assets of all students are cryptographically hashed and NFTs are created to uniquely identify. Arora et al. [14] used a BlockChain-based image matching game for the analysis of non-fungible tokens, BlockChain and cryptocurrency techniques. The authors illustrate how a decentralised application based on Ethereum interacts with ERC721 token. The authors of [15] analysed the use of NFTs in securing intellectual property and patent and presented a framework for NFT-based patent filing system. NFTs are extensively used by game developers to boost the player. The developer may get a royalty if the NFTs are sold in the marketplace. The game-based NFT might give the owner an edge to complete a stage or immunity or increase his wealth in the game. These NFTs may be sold by the organisation in the marketplace or may be given as a reward to the player to boost his performance who in turn may sell it in the marketplace to other players. There are tools [3] available which can analyse social media to know the popularity of the game in which the investor wants to invest, so that he may make a wise decision. Computer aided design(CAD) models are used in several industries around the globe. These 3D CAD models include designs of car, spare part, building or any other intellectual property of an engineer in manufacturing industries. NFTs are used to claim ownership of 3D models [4]. Music NFT gives ownership over the music file. Music that is purchased over the apps like iTunes and subscription to some apps providing access to a variety of songs give only the right to listen and not the actual ownership. By owning a music NFT you may trade it and also get the royalty of its business. Minting a NFT comes with a cost that includes fees of the application that is used to create NFT and an environmental cost which is incurred as a lot of computation is performed and electricity is used to create an NFT. Weijers [7] analyses the environmental hazards of NFTs and suggests using ESC (Environmental Smart Contracts) which uses proof of stake rather than proof of work.

3 NFTs (Non-Fungible Tokens) In this section, we will discuss the BlockChain technology and Ethereum to create NFTs. Later, we will see the places where the NFTs can be bought and sold in India. 1. BlockChain BlockChain is a distributed digital ledger with no central authority and immune to any kind of tamper [1]. The community users record transactions which are shared within the community. BlockChain technology was invented a decade ago with Bitcoin as its first product. Bitcoin and cryptocurrency are now a trillion $ mar-

330

S. Harbola et al.

ket. NFT marketplace is also growing at a good pace. The tamper-free distributed ledger concept of BlockChain technology has drawn the attention of many industries. Several industrial labs are working and integrating various technologies with BlockChain to exploit and fully utilise the inherent features of BlockChain technology. Proofing the ownership of digital assets like artwork, cartoons, avatar, music, videos, etc. has always been a big problem. These assets can be viewed and downloaded and enjoyed by anyone but someone having the digital asset say a music file is not the owner of the music and at times it becomes difficult for the actual owner to prove the ownership. NFTs can solve this problem of ownership. The BlockChain can be used to verify and authenticate the ownership. The BlockChain is composed of blocks containing data which are connected to each other thus forming a chain by adding to each other. The transactions are stored in the form of hash which are authenticated by the distributed network. The hash values stored in the blocks are unique and thus frauds are detected immediately as any change in the block will change the hash value of the block. The distributed system is transparent in nature. 2. Ethereum Ethereum [8] is a BlockChain-based community-run open-source technology software platform with smart contract functionality. A user interactive, immutable and permanent distributed application can be developed and deployed on it. It gives developers programming capabilities over a BlockChain network. The smart contracts have a set of cryptographic rules defined within it, and these are executed only under specified conditions. Say a NFT has a rule to give a royalty of 5% to the creator then each time an NFT is sold 5% of its value would automatically go to the creator of the NFT. The ability to be programmed and be customisable is helpful in building applications for various industrial processes. 3. NFT Marketplace NFT marketplace is a peer-to-peer digital platform to buy or sell NFT. Artists have to create digital tokens for their art. The investor/collector/buyer can purchase the artwork in the form of a digital token. The owner/seller of the digital token can opt to sell the NFT for a fixed price or run an auction. Many artworks/collectibles have already been sold for billions through the bidding process. A few Indian NFT MarketPlaces are as follows: (a) Wazirx WazirX [9] is a well-known India-based cryptocurrency exchange which is owned by Binance. They have started a traditional and regional-based NFT Marketplace. The platform charges 5% as a service charge for every trade. (b) Jupiter Meta Jupiter Meta was launched in 2012 and provides a platform to create Metaverse experience [10] in music, games and videos. They have NFT âŁ˜Icons of Chennai’ which is based on local food, beaches, lifestyle and historical places.

Securing Digital Ownership Using Non-Fungible …

331

(c) Bollycoin NFT market of Bollycoin [10] includes Bollywood characters of films such as famous heroes, heroines and supporting artists. Digital collectibles from the Bollywood Industry are auctioned through the platform, providing an innovative way for Bollywood enthusiasts from around the world to own NFTs of their favourite Bollywood films and celebrities. (d) Beyondlife Beyondlife [12] is the marketplace for celebrities and fans. It has a special collection of NFTs for famous actors like Amitabh Bachchan and Stan Lee’s Chakra the Invincible. An artist can create NFT in only three minutes. It supports auctions of NFT and organises several live auctions NFTS. (e) Colexion Colexion [13] has trump cards for cricket players. These cards are one of a kind and are treated as tokens and are sold on the platform. They have introduced a gaming platform which allows the players to own and trade their digital avatars and other things. It is associated with famous Cricketers like Yuvraj Singh, Glen Maxwell, etc., as they all introduced their trading cards.

4 Minting and Selling Digital Assets Creative artwork, 3D avatar, music, video clips and many more come in the category of digital assets. An artist can gain good profit by selling NFTs of his artwork. Figure 1 depicts various steps for selling NFT. The following steps are followed to sell a digital asset.

Fig. 1 Process of selling NFT in a marketplace

332

S. Harbola et al.

1. Creating an Artwork The initial step is to create a unique artwork or to choose one from one of the artist’s creations. NFT represents a digital file which may include a painting, music, video clip or anything. 2. Minting NFT Ethereum BlockChain is used in NFT creation. One must have some ether in his wallet as minting NFT comes with a service charge. Although there are many BlockChains providing NFT, Ethereum is the most popular. One has to pay the minting fees and NFT is ready to be sold in the marketplace. 3. Choose a Marketplace There are several NFT marketplaces to buy and sell NFTs. These marketplaces have their own process of minting NFT. One can choose any of the suitable marketplaces. 4. Setup an Auction Most of the marketplaces have the option to set up an auction or fixed price for the NFT. Auctions can be timed or unlimited. In timed auctions, the timing of the auction can be set by the creator as a day, week or time specified by the creator. Unlimited time means the bidding for the auction will continue until a bid is accepted. 5. Acquiring Ownership The buyer has an account in the marketplace and should have loaded the requisite amount in the wallet. In case of peer-to-peer transfer, there is no need to load money in the wallet, and the transfer of money can be made directly to the seller’s account. Mostly the wallet money is in the form of cryptocurrency like ether. The buyer participates in the LIVE auction to purchase the digital asset. If his bid is accepted, then he would be the owner of the digital asset. Once a person buys a digital asset, he can resell it or retain the asset.

5 Application of NFT 1. Election-Voting Many countries including India have a process of postal ballot. Many people who are working away from their constituency can cast their vote through postal ballot from their place of posting. There are a lot of people working away from home and their name is listed in the home district but only a few people have access to/are allowed to cast their vote through postal ballot. NFTs can be issued by the government to those people working away from home and listed in home districts so that they can cast their vote through an online process using the NFT, which could be used only once. 2. Real Estate NFTs can be used in real estate to tokenise the property. This will reduce the scams in real estate. NFTs can store the changes in ownership of the

Securing Digital Ownership Using Non-Fungible …

3.

4.

5.

6.

7.

8.

9.

10.

333

property, and it can also be used to transfer land deeds. The transactions would be simple by using smart contracts for properties and rental services. Ticketing system NFTs can be applied to the ticket booking process for movies, live concerts, matches or any other event as the seats are limited and unique, no two people can have the same seat. Ensuring Authenticity of Products Authenticity of a product can be determined by NFTs. NFTs can store the information of the supply chain thus ensuring that the product is authentic and not a fake one. It can be helpful in identifying the authenticity of food products and medicine. Industrial prototyping NFT applications don’t stop at consumer products either. There have already been numerous companies successfully using NFTs for industrial design prototyping purposes. Intellectual Property Patent and Intellectual property filing is a lengthy and costly process. NFTs-based systems have a good potential in this domain. The ownership of intellectual property(IP) can be protected by unique features of NFT. Tokenising the IP would induce confidence and assurance to the owner until the government agencies give him formal protection. Academic Credentials NFTs can be used to store certificates, courses completed, attendance, achievements and any other academic record pertaining to the student or faculty member. Degree certificates can be replaced with NFT which can be given to students who are eligible for the degree. Supply Chain NFTs can be applied to a product. The products can be tracked from manufacturing till delivery of the product, i.e. the whole supply chain, that is lifecycle of product. Digital Creativity Digital creativity includes digital art, music, videos, images, Gif(Graphical Image Format), etc. NFTs are used by artists to uniquely identify their work. NFTs are used to track the originality of the digital art. Games NFTs are extensively used by game developers to boost the selling of their digital games in the cyberspace. The developer may get a royalty, if the NFT is sold in the marketplace. The game-based NFT might give the owner an edge to complete a stage or immunity or increase his wealth in the game. These NFTs may be sold by the organisation in the marketplace or may be given as a reward to the player to boost his performance who in turn may sell it in market space to other players.

6 Issues and Challenges 1. When an NFT is sold, the creator(if desires so) may retain the copyright of the asset, that is, the buyer will have a licence to the asset which is represented by the token. Thus the buyer may not be able to exploit the use of the asset for commercial purposes. 2. A user can mint an NFT for a digital asset which he doesn’t own. There is no mechanism for ensuring the actual ownership of the asset. When we are purchasing

334

3. 4.

5.

6.

7.

8.

S. Harbola et al.

NFT we must confirm if the person is owning the NFT or not. There are many cases where people mint NFT for the copied digital asset and sell it claiming themselves as the actual owner. In India, financial institutes like banks do not provide any loan for the purchase of NFTs and they don’t accept NFTs as security for sanctioning loans. Trading in NFTs may be exploited for money laundering and terrorist financing by anti-social elements. NFTs with high values must be scrutinised by the application providers in consultation with the government to mitigate the risk of money laundering and terrorist financing activity. The employees of NFT companies may get involved in the purchase of NFTs to gain more profit and might get involved in illegal and unfair means. Marketplaces must create policies for their employees. Mining process uses a lot of computation thus releasing a lot of heat and consuming electricity making it a very energy-intensive process; proof of work used by Ethereum is not sustainable in the long run, and sooner or later, it has to switch to proof of stake. There is a risk of cyberattacks and cyberfrauds associated with NFT retention. Even a small flaw in a smart contract can be exploited by the attacker to gain access or move the NFTs to different networks. One such example is Poly network which lost $600 million worth of assets due to flaws in the Smart contract. The price of a NFT depends on various factors like creativity, uniqueness and scarcity. As of now, NFTs do not follow any standard and thus the price of the NFT might fluctuate and is usually unpredictable.

7 AI and NFT In the last decade, AI has proved its applicability in daily lives. AI experience has led to the change of software in most of the industries and even influenced NFT. Artificial Intelligence and deep learning algorithms are in prediction and trend analysis of the stock market [17] based on time series data. Models of ARIMA and LSTM [18] are being extensively used by researchers to predict the spread of the Covid Virus thus helping the government agencies to take preventive measures. The AI models and deep learning can be used by the NFT makers to see the trend of NFT marketplaces to help them produce NFTs of a certain type to gain more profit. The convergence of AI and NFT new capabilities has evolved from a simple ownership to intelligent, self-learning ownership which enhances the user experience. New developments in artificial intelligence have led to its applications in the area of blockchain technology and introduced AI NFTs. AI models like GANs (Generative Adversarial Networks) can generate their own content if they are well trained with a good data set, and this enables them to attach with a blockchain technology to eventually generate NFTs. NFTs can be generated by using various models of artificial intelligence. An AI NFT is a non-fungible token embedded with a Generative Pre-trained Transformer 3 (GPT-3) language model prompt as part of its smart contract. This type of NFT

Securing Digital Ownership Using Non-Fungible …

335

is intelligent and has properties like interactivity and animation. The progress in the fields of AI like NLU (natural language understanding), computer vision and speech analysis has merged with NFT and begun a new ecosystem of intelligence NFT. AI can be used to generate NFTS and AI could be embedded within NFT thus giving NFT users a dynamic and intelligent experience. There are mainly three categories of NFT with AI capabilities: 1. AI-generated NFTs The progress in AI has facilitated the creation of NFTs using AI technology thus creating thousands of NFTs in a few seconds. AI models like GANs (Generative Adversarial Networks) can generate their own content if they are well trained with a good data set, and this enables them to attach with a blockchain technology to eventually generate NFTs. The creators can easily create new art using various deep learning methods such as computer vision, language and speech. There are several websites and projects which give a facility to its users to create NFT art by just giving the keyword, and the rest of the work is done by AI engines. In the current trend, AI-generated NFTs are of the form of generative arts and they will diversify in the coming years by using new deep learning techniques. 2. NFTs’ embedded-AI A NFT created by AI and deep learning technology is not intelligent enough to interact with users. Embedding AI features within a NFT makes it intelligent and enhances user experience. Artificial Intelligence within NFTs opens up a whole new field and opportunity for the researchers, artisans and creators to enhance the capabilities of the NFTs in a different direction. This enables the NFT to learn new things as it interacts with new users thus giving a unique experience. These NFTs have self-learning capabilities thus learning new things as they interact with users and create a different and user-specific experience. This enables the artists and creators to develop and design an intelligent and interactive NFT around their favourite characters. Even the existing NFTs can be embedded with AI to improve user experience. The NFTs incorporate speech and language capabilities to provide an interactive user experience. Alethea created one such NFT, which was auctioned for $480,000. 3. AI-first NFT Infrastructures The AI and deep learning methods for NFTs can be used by the whole ecosystem of NFT. Integrating capabilities of AI with the building blocks of NFT, like data platforms and marketplace, would enable intelligence throughout the lifecycle of NFTs. NFT contains intelligent indicators which analyse data present on chain and use computer vision methods to make smart and intelligent suggestions to its users. Data and intelligence APIs are going to become an important component of the NFT market.

336

S. Harbola et al.

8 Conclusion NFTs have solved problems of ownership of digital assets. Before the BlockChain technology, the ownership of digital assets was subject to tempering and it was hard to verify the ownership leading to significant financial losses. NFTs have applicability in a variety of areas from 3D models, engineering designs, digital artwork, scripts, music, video clip and many more. RBI(Reserve Bank of India) has been warning citizens from investing in cryptocurrency but nothing has been stated about NFT trading, as of now. NFT is all about ownership and royalty thus giving the artists a better chance to earn. There are many marketplaces or apps available in India to sell or buy NFT, and these are mostly based on regional and cultural interests. The NFT market is in its initial state in India and is growing very fast, giving opportunities to artists. But can NFT trading be considered safe? Only time would tell, as technology is still continuously evolving.

References 1. Yaga D, Mell P, Roby N, Scarfone K (2019) Blockchain technology overview. arXiv:1906.11078 (2019) 2. Kim JW, Song JG, Lee TR, Jang JW (2022) Performance evaluation of NFT trading platform based on hyperledger fabric blockchain. In: 2022 the 8th international conference on computing and data engineering. pp 65–70 3. Boonparn P, Bumrungsook P, Sookhnaphibarn K, Choensawat W (2022) Social data analysis on play-to-earn non-fungible tokens (NFT) games. In: 2022 IEEE 4th global conference on life sciences and technologies (LifeTech). IEEE, pp 263–264 4. Mouris D, Tsoutsos NG (2022) NFTs For 3d models: sustaining ownership in industry 4.0. IEEE Consumer Electronics Magazine 5. Chirtoaca D, Ellul J, Azzopardi G (2020) A framework for creating deployable smart contracts for non-fungible tokens on the Ethereum blockchain. In: 2020 IEEE international conference on decentralized applications and infrastructures (DAPPS). IEEE, pp 100–105 6. Kumar NN, Senthil Kumar R, Rajshekhar Basale R, Saffath M (2022) Decentralized storage of educational assets using NFTs and blockchain technology. In: 2022 4th international conference on smart systems and inventive technology (ICSSIT). IEEE, pp 260–266 7. Weijers D, Joseph Turton H (2021) Environmentally smart contracts for artists using nonfungible tokens. In: 2021 IEEE international symposium on technology and society (ISTAS). IEEE, pp 1–4 8. Ethereum. https://ethereum.org/. Accessed 9th May 2022 9. WazirX. https://wazirx.com/. Accessed 10th May 2022 10. Jupiter Meta. https://jupitermeta.io/. Accessed 10th May 2022 11. Bollycoin. https://www.bollycoin.com/. Accessed 11th May 2022 12. Beyondlife. https://www.beyondlife.club/. Accessed 11th May 2022 13. Colexion. https://colexion.io/. Accessed 12th May 2022 14. Arora A, Kumar S (2022) Smart contracts and NFTs: non-fungible tokens as a core component of blockchain to be used as collectibles. In: Cyber security and digital forensics. Springer, Singapore, pp 401–422 15. Bamakan Seyed Mojtaba, Hosseini Nasim Nezhadsistani, Bodaghi Omid, Qiang Qu (2022) Patents and intellectual property assets as non-fungible tokens; key technologies and challenges. Sci Rep 12(1):1–13

Securing Digital Ownership Using Non-Fungible …

337

16. Wang Y (2022) Anchoring effect and loss aversion: evidence from the non-fungible token market. Available at SSRN 4097185 17. Wang Y, Guo Y (2020) Forecasting method of stock market volatility in time series data based on mixed model of ARIMA and XGBoost. China Commun 17(3):205–221. https://doi.org/10. 23919/JCC.2020.03.017 18. Harbola S, Jain P, Gupta D (2021) Analysis, visualization and forecasting of COVID-19 outbreak using LSTM model. In: Khanna A, Gupta D, Pólkowski Z, Bhattacharyya S, Castillo O (eds) Data analytics and management. Lecture notes on data engineering and communications technologies, vol 54. Springer, Singapore. https://doi.org/10.1007/978-981-15-8335-3_14

Battery Optimization of Electric Vehicles Using Battery Management System Simran Khanna, Vansh Bhandari, Tanmay Mishra, Yash Shrivastav Yashas Bajaj, and Srishti Singh

Abstract Battery Management System (BMS) is an electronic system that manages a chargeable battery to confirm that it has been operated safely and expeditiously. It monitors parameters like temperature, voltage, and current to confirm safe conditions like acceptable cooling of the battery to forestall warming. Mistreatment of lithiumion batteries has a high power-to-weight magnitude relation, high energy potency, smart high-temperature performance, and low self-discharge. Overcharging degrades the capability of the battery. BMS determines what proportion of current will safely go into the battery. BMS results in reliable management of power and helps within the optimum power performance. Incorrect operations like too high or too low temperature, overcharging, or discharging can speed up the degradation method of battery dramatically. In this manuscript, we will be discussing about the advantages of a BMS in the battery optimization of Electronic Vehicles. The majority of the problems may be resolved by developing advanced BMS in electrical Vehicle (EV) like battery modeling, correct battery, state of charge, and state of health estimation, which can provide an exact driving range of EV. Keywords Electric Vehicle · Optimization · Battery management system · Battery modeling · Battery charging

1 Introduction BMS refers to a management scheme that monitors, controls, and optimizes an individual’s performance or multiple battery modules in an energy storage system. In automobile applications, BMS is used for energy management in different system interfaces and ensures the system’s safety from various hazards. Key technologies in the BMS of EV include battery modeling, state estimation, charging, and discharging. A good BMS should safely protect the driver/operator by detecting unsafe operating S. Khanna (B) · V. Bhandari · T. Mishra · Y. Shrivastav Yashas Bajaj · S. Singh Vivekananda Institute of Professional Studies—Technical Campus, New Delhi, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. Misra et al. (eds.), Internet of Things (IoT): Key Digital Trends Shaping the Future, Lecture Notes in Networks and Systems 616, https://doi.org/10.1007/978-981-19-9719-8_28

339

340

S. Khanna et al.

conditions, protecting the cells from damage in failure cases, prolonging the life of the battery in normal operating regions, and should inform the user about the battery details and its status of operation. The primary goal of the Battery Management System (BMS) is to protect the battery and stop any operations that are outside of its safe limits. It keeps track of the health and state of charge (SOC) of the battery pack (SOH). BMS controls a rechargeable battery to make sure it runs securely. It is intended to keep track of the parameters related to the battery pack and each of its cells, then use the information gathered to reduce safety hazards and improve battery performance. As a result, the present work is a review of BMS with an emphasis on research into BMS optimization for Evs, which will increase BMS reliability and improve the power performance of Evs. According to the results of the literature search, BMS performance needs to be improved for Evs in the future.

2 Battery and Battery Management System The transition of battery and power supply systems to Evs from traditional ICEs is well under progress. However, one of the main reasons why electric vehicles are not more prevalent on the road is their limited range, which is caused by the constrained energy storage capacity of present battery systems. According to research, 100 miles can be covered by Evs with ten minutes of fast charging, and the battery packs can be sufficiently charged. Therefore, the BMS in Evs must play crucial functions as a component. It is a device to control the operation of the battery’s life cycle in order to increase the battery’s lifespan. It is crucial for the BMS to maintain the battery’s dependability and safety, as well as to make sure that the charge regulation, cell balancing, and state monitoring and assessment are all properly working. As with other electrochemical devices, a battery’s chemistry will behave differently depending on the situation. The implementation of these functionalities is difficult as a result of the unpredictability of battery performance. Due to their increased energy density, reduced self-discharge, and longer life cycle, lithium-ion batteries (LIB) have been widely recognized as advanced technology utilized and developed during the last ten years. Additionally, LIB chemistry is widely recognized as the preferred technique for EV energy storage in the direction of sustainable transportation. However, there is still a need for additional study points and room for improvement in the endeavors of the future. These included developing electronic circuits and algorithms for a more efficient battery consumption for Evs as well as selecting the optimum cell materials. Voltage, current, and temperature were the battery parameters that should be taken into account for BMS optimization. Thermal management system would be necessary to maintain the optimal cell performance and also to accomplish a full battery lifespan, which is another obstacle placed on the performance of the vehicle connected to BMS (Fig. 1).

Battery Optimization of Electric Vehicles Using Battery Management …

341

Fig. 1 Graph expressing the relation between specific energy and energy density

2.1 Functionalities Battery packs that are connected internally or externally are handled by BMS. With common data for cell voltages, pack current, pack voltage, and pack temperature, it determines the battery values. The state of charge (SOC), state of health (SOH), depth of discharge (DOD), and operational critical parameters of the cells/battery packs are all estimated by BMS using these measurements. The measurements also aid in extending battery life and keeping up with the original power network’s demand needs. Functional building components and design methodologies are used to construct BMS. The proper architecture, functional unit blocks, and associated electronic circuitry for designing a BMS and BMS charging scheme will be indicated by the battery requirements for various applications. Based on the following characteristics, battery life can be improved (Fig. 2). – Energy management system with a user interface to control and examine battery systems performance in different system blocks. – Battery pack performance and safety features. – Resiliency among the system units in different accident scenarios. – Advanced technologies that integrate batteries with conventional/nonconventional energy sources. – Internet-of-things (IoT), which monitors and controls the energy management system.

342

S. Khanna et al.

Fig. 2 Key functions of BMS

3 Optimization of Lithium-Ion Battery Lithium-ion batteries are used in Electric Vehicles due to their high energy and current density, long life cycle, and low self-discharge; but the Lithium ions present in Lithium-ion batteries are brittle and hence we need a safety-providing device in which they can be stored for each pack. This device is referred to as a Battery Management System or BMS which is used to set the peak and threshold voltage limits of a battery during charging and discharging. It monitors cell temperature and is also used to control maximum charging and discharging currents. Therefore, BMS is important for battery management and protection, prolonging its life and keeping it ready for full power delivery when required. Optimization of Lithium-Ion batteries requires various negative factors to be minimized/dealt with. Self-discharging occurs even without any connections between the electrodes through internal chemical reactions. It usually depends on the battery type and is especially high in Nickel-based rechargeable batteries (~15–30%) and around 2– 3% in Lithium batteries (per month). This is because primary batteries cannot be recharged between manufacturing and use and hence have a lower self-discharging rate. But since self-discharging is a process which occurs more quickly at higher temperatures, storing batteries at lower temperatures could reduce the self-discharge rate and hence preserve the initial energy stored in the battery. Extreme weather conditions (When the ambient temperature and relative humidity are altered beyond the norm)can affect the battery performance and can cause the battery to stop working or/and cause various defects such as causing it to bulge, bubble, melt, become damaged, smoke, sparks, flame, expand, contract, or even explode in extreme cases.

Battery Optimization of Electric Vehicles Using Battery Management …

343

Fading capacity of a battery refers to the property of a battery which causes it to gradually and irreversibly lose its capacity to hold a charge. Since it goes through multiple cycles of charging and discharging, its quality deteriorates over time. Its general life period is 2–3 years (300–500 cycles), and in Electric Vehicles, it is important to know the present condition of the battery and the remaining lifetime. Lithium-ion batteries have been widely used as power cells because of their advantages of high energy density, low self-discharge rate, and no memory. The charging strategy of the battery will affect its charging efficiency, cycle life, and safety performance. The constant-current–constant-voltage (CCCV) charging method is one of the most widely used charging methods; further, it is simple and easy to control. However, the CV phase takes a long time. Accordingly, a large number of optimized charging strategies have emerged. A pulse charging method was proposed to improve the charging efficiency, which allows lithium ions to diffuse more evenly throughout the battery and thus alleviate polarization. In this scenario, the charging time is realized by changing the amplitude and width of the current, and it is difficult to control effectively. As the charging rate will affect its charging time and cycle life, the multi-stage CC charging strategy is widely used. A multi-stage charging method that considered the charging time and energy loss as optimal objectives was proposed in [4, 5], which indirectly controls the cycle life of a battery by controlling energy loss. However, the battery life is not verified at last. In order to realize the online optimization, a charging strategy based on model predictive control was proposed in [6]. It applied system models to predict system responses and to find the best future control sequence by optimizing the user-defined objective function. But the whole process is complex to implement. In summary, previous battery charging strategy studies mainly concentrated on the optimization of the charging time or polarization. Until now, there has been little work done to improve the cycle life of the battery. As the battery’s energy loss during the charging process increases, the corresponding battery capacity degradation will be more serious [7]. Therefore, in order to prolong the cycle life of the battery, the capacity degradation speed and energy loss were taken as two optimal objectives. The cycle life test of the battery at different SOC cycle intervals was used to establish the capacity degradation speed model. In addition, the energy loss was calculated based on the equivalent circuit model of the battery, and the effects of the charging rate and SOC on the model parameters were taken into account. The optimal current sequence was obtained by the dynamic programming algorithm with the average charging rate, maximum charging rate, charging capacity, and cut-off voltage as constraints. And a contrast test with the traditional charging method was made.

344

S. Khanna et al.

3.1 IoT-Based Battery Management System for Electric Vehicles In this, the idea of monitoring the performance of an electric vehicle using Iot techniques is proposed. It has two components: a monitor and an interface. The system is capable of detecting declining battery performance and sending notifications to the user for further action. By combing a GPS system to detect the coordinate and show it on Google Maps, the system is capable of sending information such as position, battery state, and time via the Internet. This concept is frequently used in smartphones by building smartphone applications that assist users in battery monitoring.

4 Battery Model and Parameters Identification The first-order equivalent circuit model was chosen in this paper, as shown in Fig. 3. OCV is the open circuit voltage which has a close relationship with SOC. The Ohmic resistance RΩ represents the internal connection impedance of the battery. The polarization resistance RP and its parallel capacitance CP reflect the polarization phenomenon of the battery, mainly including electrochemical polarization and concentration polarization [8]. IL is the load current (positive for discharge, negative for charge). In this study, the power-type battery was selected as the research object. Its positive and negative electrode material is Li(NiMnCo)O2 and graphite, respectively, nominal capacity is 8Ah, the charging and discharging cut-off voltage is 4.2 V and 2.75 V, respectively, and the maximum charging rate is 15C. In order to obtain the battery parameters more accurately, the pulse charge– discharge test of the battery at different SOC and charging rates was conducted, as shown in Fig. 4. First, a 0.05C charge–discharge test was carried out to obtain the capacity and OCV-SOC curve. Since the polarization is very small with the small charging and discharging rate, the value of OCV is the same as the terminal voltage. The OCV-SOC curve is shown in Fig. 5. Then the battery was tested for internal resistance. The battery has been subjected to a charge process with a pulse of 10 s for every SOC of 5% with charging rates 1C, Fig. 3 First-order equivalent circuit model of battery

Battery Optimization of Electric Vehicles Using Battery Management …

345

Fig. 4 Test flow of battery model parameters

Fig. 5 OCV-SOC curve

3C, 5C, 6C, and 8C, respectively. The discharging pulse rate was 1C, but it should ensure that the discharging and charging capacity was consistent during the pulse test. The parameters of internal resistance were identified by the least squares method mentioned. The Ohmic resistance plays a role in the moment of adding current and the formula is shown in Formula (1). In addition, the polarization voltage at different SOC points and charging rate in the charging process can be expressed by the battery model through Formula (2). To make Equation (2) available for numerical calculation, it should be discretized by Equation: RΩ =

ΔU ΔI

(1)

Vo (S OC, Ic ) = OC V (S OC) − It RΩ (S OC, Ic ) − V p (S OC, Ic )

(2)

It Vp dV p = − dt Cp RpC p

(3)

346

S. Khanna et al.

Fig. 6 (a) Ohmic resistance identification results; (b) Polarization resistance identification results

Vo (K ) = OC V (K ) − RΩ I L (K ) − R p L p (K ) ( ( )) 1 − exp − Δt τ I p (K ) = 1 − I L (K ) Δt ( +

(

τ

1 − exp − Δt τ Δt τ

)

)) ( Δt − exp − I L (K − 1) τ

) ( Δt I p (K − 1) + exp − τ

The identified parameters of RΩ and RP are plotted in Fig. 6a, b. It can be seen from the image that at the same rate, the RΩ and RP decrease with the increase of SOC and are more stable in the 30%-80%SOC interval. That’s because the positive electrode of the battery is in a lithium-rich state at the beginning of charging and it takes more energy to escape. So the polarization is large, resulting in a large internal resistance in the low SOC intervals. In order to verify the accuracy of the model parameters, the equivalent circuit model of a battery was established under Simulink. The voltage under simulation and actual test with different charging rates are shown in Fig. 7. The comparison shows a little difference between the simulation and test results. The average error is within 0.4%. It can be demonstrated that this result can satisfy the precision of the parameters required in this paper.

Battery Optimization of Electric Vehicles Using Battery Management …

347

Fig. 7 Voltage curves of simulation and test

5 Formulation of Battery Charging Optimization 5.1 The Models of Capacity Degradation Speed The cycle life test of the battery was carried out in different SOC intervals at 25 °C. The SOC was divided into 0%-20%, 20%-40%, 40%-60%, 60%-80%, 80%-100%, and 0–100% for a total of 6 intervals for life testing and there are 3 battery samples under each test condition. The charging and discharging rate were 6C (48A). A performance test was conducted every 200 cycles, including 0.05C charge–discharge tests and HPPC tests. The calculation of capacity degradation is shown in Formula (4). The 1 s resistance is regarded as the Ohmic resistance. The difference between the 60 s resistance and Ohmic resistance is taken as the polarization resistance. The changes in the capacity degradation and Ohmic and polarization resistance with the number of cycles under different SOC cycle intervals are shown in Fig. 8. Capacit yloss = 100 × (1 − Q test /Q intial )

(4)

1

Fig. 8 The changes of (a) Capacity degradation; (b) Ohmic resistance; (c) Polarization resistance with the number of cycles

348

S. Khanna et al.

The capacity degradation speed model DS(Ic) of the lithium-ion battery with Li(NiMnCo)O2 cathode is shown in Ref. [10]. Since the positive and negative materials of the battery in this paper are the same as those of the literature. based on this life model, the capacity degradation speed model DS(SOC,Ic) thinking of different charging rates and SOC cycle intervals is shown in Formula (5). The SOC-related parameter K refers to the ratio between the capacity degradation of the each and entire cycle interval. The SOC changes in each charging stage are expressed as Formula (6). η is the Coulomb efficiency and 1 has been taken. DS(soc, Ic ) =

N −1 ∑

K (soc) × ΔS OCk × DS(Ic )

(5)

i=1

∫ ΔS OC =

Δs

η×I L (t)dt/3600×capacit y

(6)

0

Figure 8a shows a linear relationship between the capacity degradation and the number of cycles, such as (7). To identify the parameters h and g, the capacity degradation curves under different SOC cycle intervals are fitted by Formula (7). By deriving Formula (7), the capacity degradation speed (DS) is obtained in Formula (8). The parameters g and K under different SOC cycle intervals are shown in Table 1. (a) shows a linear relationship between the capacity degradation and the number of cycles, such as (7). To identify the parameters h and g, the capacity degradation curves under different SOC cycle intervals are fitted by Formula (7). By deriving Formula (7), the capacity degradation speed (DS) is obtained in Formula (8). The parameters g and K under Fig.8 (a) show a linear relationship between the capacity degradation and the number of cycles, such as (7). To identify the parameters h and g, the capacity degradation curves under different SOC cycle intervals are fitted by Formula (7). By deriving Formula (7), the capacity degradation speed (DS) is obtained in Formula (8). The parameters g and K under different SOC cycle intervals are shown in Table 1. r different SOC cycle intervals are shown in Table 1. Capacit yloss = h + g × cycle

(7)

DS = d(Capacit yloss )/d(cycle)=g

(8)

Table 1 Values of parameters b and K under different SOC cycle intervals 0–20%SOC 20–40%SOC 40–60%SOC 60–80%SOC 80–100%SOC 0–100%SOC 0.00304

0.00469

0.00375

0.00375

0.00563

0.00662

K 0.45922

0.70846

0.56647

0.56647

0.56647

1

g

Battery Optimization of Electric Vehicles Using Battery Management …

349

6 Battery Cooling Systems One way to protect Lithium-Ion Batteries is to control their temperature. The primary objective of a system with thermal management capabilities is to supply a battery pack at an acceptable mean and consistent distribution of temperature (or perhaps with only slight variations among the battery modules of the battery cell), as specified by the battery supplier. The battery module heat management system must, nevertheless, be small, light, inexpensive, portable, and consistent with the location of the car as specified by the car’s manufacturer. Additionally, it must be precise and accessible for maintenance needs. Setting up a suitable thermal management system will effectively remove heat from the battery pack, help to reduce the excessive temperature rise, improve stability, and improve protection during charging and discharging. Some of the ways by which we can improve the temperature conditions by Batterycooling systems are listed below:

6.1 Cooling Methods Used in Battery Cooling Systems The four critical functions of BCS are heat extraction coolant flow from inside the battery, raising the battery temperature by heating whenever the system is at a very low temperature, shielding to avoid rapid fluctuations in battery temperature, and a mechanism to expel possibly dangerous fumes from inside the module. Additionally, BCS must ensure properties like high performance, simplicity, low weight, cheap cost, less usage of parasitic power, quick packing, and easy maintenance in order to adapt to EVs. AIR-BASED BCS The BTM air-cooled technology works by allowing air to travel through the module, passively cooling the pack. Depending on the cooling method, which may include a fan, the air-cooling system can be divided into two categories: forced (artificial) convection cooling and natural convection cooling. The method of air cooling is frequently used in marketable EVs due to the advantages of simple design, easy maintenance, and lower cost. Because air has a low viscosity, parasitic power consumption is minimal throughout the whole system operation cycle. A passive air-cooling system using a cabin air (natural air cooling) and a straightforward forced air-cooling system are used. PLATE-BASED BCS The best option to cool an EV battery pack is liquid cooling, where a cold plate acts as an interface of heat transferbetween battery cells and flowing fluid. Air cooling is the simplest method of maintaining an EV battery pack’s temperature, but because air has a very low coefficient of heat transfer, liquid cooling is the best option. Through the use of the cold plate and the liquid cooling mechanism, which has attracted

350

S. Khanna et al.

numerous researchers recently, heat will be removed from the battery. Thin metallic structures known as “cold plates” have internal channels from which coolant flows. The pace at which heat is transferred from the battery cells to the cooling plate is determined by the rate of heat production in the cell. The coolant then transfers the heat away from the battery and discharges it outside the cell.

7 Advantages BMS is a very important component in Electric Vehicles as it is essential in maintaining the safety and reliability of the battery, used in battery state monitoring and evaluation, controlling the state of charge, balancing cells and controlling the operating temperature, and the management of regenerative energy.

8 Disadvantages Around the world, the demand for electric vehicles is increasing at an exponential rate. With global concern heightened for emission reduction to fight climate change, there is greater support for increasing electric transportation use while phasing out the conventional fuel-driven vehicles. And as more and more electric vehicles are deployed on the roads, attention must be given to the BMS products to ensure public transportation reliability and manufacturing regulations. Typical BMS architecture for electric transportation applications is master–slave architecture, where there are central control and distributed sub-controllers. In a typical BMS architectural topology each slave board manages a group of cells and a master control board interfaces with a slave board to control the overall functionality of the system. BMS safety is one of the most elevated concerns in the battery industry. Several numbers of codes and standards are prepared and followed for different applications to ensure battery safety. However, although batteries are the most convenient form of energy storage, accidents are continuing to happen in battery infrastructure. As a result, all design basis scenarios must be considered to eliminate the risks. On comparing the graph of State of Charge and Open Circuit Voltage curve LFP battery has high voltage and is of high preference over NiMH battery. Cylindrical LFP is of high choice because it is easy to manufacture them. One major drawback of LFP is that it makes State of Charge estimation and balancing cell among the battery system challenging. Because lithium batteries get heated up easily some research work in the field of battery charging estimation and temperature management is needed. Hence other chemical equivalent battery materials are needed to be considered to prolong battery usage to the maximum.

Battery Optimization of Electric Vehicles Using Battery Management …

351

9 Conclusion In this paper, we talk about how a battery management system (BMS) can help in improving the battery of electric vehicles. We can achieve this by bettering the lithium-ion batteries. Traditional ways to improve the battery life of lithium-ion batteries are to use partial-discharge cycles, to avoid overcharging, and to limit the battery temperature. Future BMS may use a hybrid energy storage system (HESS) that combines lithium-ion batteries and ultracapacitors in order to improve predictive technical models for deployment [12]. The charging optimization in lithium-ion batteries based on capacity degradation speed and energy loss is proposed in this paper. The capacity degradation speed model based on the characteristics of the battery life cycle in different SOC intervals is built. By a first-order dynamic equivalent circuit model of the battery, the energy loss is calculated and the optimal objective function is established. The optimal current sequence is obtained by the Dynamic programming algorithm. Compared with the traditional charging method, when the balance coefficient M is 0.5, the loss of capacity and energy are reduced by 3% and 2%, respectively, whereas the average charging rate is almost the same. In addition, the capacity degradation of the traditional and optimized charging method is 3.85% and 2.64%, respectively, after 750 cycles, which effectively prolongs the cycle life of the battery. In order to improve the predictive technical model for deploying performance management of EVs, the hybrid energy storage system (HESS), which should integrate ultracapacitor combined with lithium compound battery, is a viable future view for BMS. Future innovation invention is anticipated to support the EV industries by the year 2020 or beyond. Additionally, it will be possible to achieve the goal of reducing greenhouse gas (GHG) emissions by up to 40% by the year 2020. There is no standard solution setting for BMS performance, however, and many tactics must still be used in order to enhance BMS performance for the next HEVs and EVs.

References 1. Gozdur, Roman, Tomasz, Przerywacz, Dariusz, Bogda´nski (2021) Low power modular battery management system with a wireless communication interface. Energies 14, 19 2. Samanta, Akash, Sheldon S, Williamson (2021) A Survey of wireless battery management system:topology, emerging trends, and challenges. Electronics 10, 1–12 3. Surya, Sumukh, Vidya. Rao, Sheldon S Williamson (2021) Comprehensive review on smart techniques for estimation of state of health for battery management system application. Energies 14, 1–22 4. Divakar BP, Ka Wai Eric Cheng, Wu HJ, Xu J, Ma HB, Ting W, Ding K, Choi WF, Huang BF, Leung CH (2009) Battery management system and control strategy for hybrid and electric vehicle. In: International Conference on Power Electronics Systems and Applications (PESA), pp 1–6. IEEE, Hong Kong 5. Gabbar Hossam A, Ahmed M Othman, Muhammad R Abdussami (2021) Review of battery management systems (BMS) development and industrial standards. Technol 9, 2–23

352

S. Khanna et al.

6. Ghuge Dr NN, Dandge Vishal, Bari Kunal, Mahesh Kulkarni (2021) IoT based battery management system for electric vehicles. In: International Conference on Communication & Information Processing (ICCIP) 2021 7. Karkuzhali V, Rangarajan P, Tamilselvi V, Kavitha P (2020) Analysis of battery management system issues in electric vehicles. In: Conference Series: Materials Science and Engineering, pp 1–7. IOP Publishing 8. Hariprasad A, Priyanka I, Sandeep R, Ravi V, Shekar O (2020) Battery management system in electric vehicles. Int J Eng Res 9:1–3 9. Moulik, Bedatri, Dirk, Söffker (2020) Battery management system for future electric vehicles. Appl Sci 10, 1-3 10. Chen, Yukai, Khaled, Sidahmed Sidahmed Alamin, Daniele, Jahier Pagliari, Sara, Vinco, Enrico, Macii, Massimo, Poncino (2020) Electric vehicles plug-in duration forecasting using machine learning for battery optimization. Energies 13, 1–19 11. Hu Rui (2011) Battery management system for electric vehicle applications 12. Salehen PMW, Su’Ait MS, Razali H, Sopian K (2017) Battery management systems (BMS) optimization for electric vehicles (EVs) in Malaysia. In: AIP Conference Proceedings, AIP Publishing 13. Hoke Anderson, Alexander, Brissette, Dragan, Maksimovi´c, Annabelle, Pratt, Kandler, Smith (2011) Electric vehicle charge optimization including effects of lithium-ion battery degradation. IEEE vehicle power and propulsion conference, pp 1–8. IEEE 14. Shen, Ming., Qing, Gao (2019) A review on battery management system from the modeling efforts to its multi application and integration. Int J Energy Res 43, 5042–5075

Electronic Voting Machine as a Service on the Cloud—Azure for EVM (A4EVM) Mohammad Equebal Hussain , Mukesh Kumar Gupta, and Rashid Hussain

Abstract Electronic voting machine (EVM) is a well-known device used to record votes. In India, world largest democracy, EVM is extensively used in all the elections. This mode of election is preferred over ballot paper due to many reasons. The main reason is to discourage the use of paper to save the environment. Another reason is to reduce electoral fraud and abuse. An ideal EVM is expected to be accurate and secure. An EVM should not have any networking capability, like Bluetooth connectivity, or any type of wireless connectivity, to make sure that the data (saved votes) stored within EVM cannot be accessed, altered, or modified. In recent years, there are various questions arises on its credibility, because data is recorded locally within the EVM, it can be manipulated easily. However, the M3 version of the EVMs includes the Voter Verifiable Paper Audit Trail (VVPAT) system for cross-verification. In recent elections, it has been noticed that election is delayed due to faulty EVMs. To solve all the above issues, we have proposed an advanced solution, based on Microsoft Azure cloud service. In this paper, we are proposing EVM as a service on Azure-based client–server architecture, where the machine will act as a client, the data will be recorded on Azure server. The proposed solution is secure, cost-effective, and robust due to the security guaranteed from Microsoft. Other advantages include, the counting can be done automatically using Azure services, result can be published very next day, data can be dumped permanently in Azure data lake (ADLS) for future use, various analyses can be done using machine learning and analytical services like power BI. Last but not least, EVM is free immediately after the election.

M. E. Hussain (B) · M. K. Gupta Suresh Gyan Vihar University, Jaipur, India e-mail: [email protected] M. K. Gupta e-mail: [email protected] R. Hussain Moti Babu Institute of Technology, Bihar, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. Misra et al. (eds.), Internet of Things (IoT): Key Digital Trends Shaping the Future, Lecture Notes in Networks and Systems 616, https://doi.org/10.1007/978-981-19-9719-8_29

353

354

M. E. Hussain et al.

Keywords Electronic voting machine as a service (EVMaaS) · Microsoft Azure · Cloud technology · Observability metrics · Business intelligence · Virtual machine · Docker container as virtual EVM · Power BI

1 Introduction Electronic voting machine has replaced the ballot paper-based traditional system. The new systems consist of direct read electronic system (DRE) and computercounted system; however, this system has some drawbacks like confusing to voters and difficult to count. DRE-based systems provide the interface or buttons to enter the choice. The vote is recorded in machine internal storage. However, DRE systems don’t satisfy security demands. This is covered in detail by Chris Dahlberg in his paper [1]. In case of fraudulent transactions, there is no way to determine the point of failure, unlike banking transactions where user and bank know each other. A valid election must meet certain criteria like privacy, fraud resistance, ease of use, scalable, and speed. A preliminary result must be available within hours of election completion. Lengthy counting time is not acceptable. Many problems like duplicate voters, and forged votes, are also identified by the author in [2]. Designing secure and correct system is a big challenge and difficult too. A single error in EVM can have a much larger effect on the fair election [3]. A report published in [4] mentioned about software error and possibility of fraud. It provides some data like “4,438 electronic ballots to be lost and never recovered in North Carolina, November 2004. Similarly, 134 electronic ballots were blank in a one-race election held on DRE voting machines in which the margin of victory was 12 votes” [4]. In this paper, we are going to discuss the existing design of electronic voting machine, followed by the next-generation “EVM with VVPAT”, followed by our proposed design. We will also go through some of the literature reviews. Since not much work is done in this area, therefore, we will discuss the virtualization and cloud technology, to simulate our proposed design as proof of concept. The analysis of result will be done using power BI, using data available on the election commission website. All the diagrams are prepared using online tool draw.io [5].

1.1 EVM and VVPAT EVM in India is adopted in 2004. It is designed and developed by Bharat Electronics Limited (BEL) and Electronics Corporation of India Limited (ECIL) [6]. The system is a set of two devices, voting unit and control unit, running on a 6 V battery. Voting unit is used by voter, whereas control unit is operated by electoral officer. Voting unit can accommodate between 16 and 64 candidates where each button represents one candidate. The control unit contains three buttons. One for release of single vote, one for total number of votes, and one to close the election process. The result button

Electronic Voting Machine as a Service on the Cloud—Azure for EVM …

355

Fig. 1 Indian voting machine and VVPAT-M3

remains sealed. Figure 1 shows the existing Indian electronic voting machine and Voter Verifiable Paper Audit Trail (VVPAT) [7]. It helps by providing feedbacks to voters as an independent verification system in order to ensure the accuracy, transparency, and confidence. However serious allegations have been made from time to time on the possible hacking of EVM as well as tampering wirelessly using a particular network [8]. The use of EVM to collect the votes seems to be logical choice to accomplish the desired goals. But unfortunately, this system is marked by a series of problems which worries everyone. A number of studies have shown that most of the EVM being used are flawed due to poor quality. Most of the problems are related to the reliability of voting machine [9].

1.2 Cloud Computing and Microsoft Azure Cloud Service Modern applications are designed using cloud computing. It allows application to be broken into microservices, which communicates through API using message queue, possibly Kafka. In cloud-based architecture, the required resources can be scaled UP or DOWN as well as it can be scaled IN or OUT based on the demand. Scale up/ down refers to increase/decrease of computing power, respectively, whereas scale out refers to integrating more resources into existing units. Microsoft Azure [10, 11] provides end-to-end support to build enterprise solution using cloud computing services, i.e., Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS). Microsoft Azure is quick, flexible, and most affordable cloud platform. Few examples include web and mobile application development, testing and hosting, VM creation and business intelligence, etc. It also supports power BI tools to capture important metrics for application insights and its activities.

356

M. E. Hussain et al.

1.3 High-Level Proposed Design A high-level architecture of the proposed design is shown in Fig. 2, where EVM is simply a client, connected to a dedicated IP address on the cloud, possibly preconfigured endpoint, read-only destination. This is to make sure that data is stored in the correct location on the Azure cloud.

2 Detailed Design—Individual Components The secure connection between EVM and Microsoft Azure cloud is shown in Fig. 3, where EVM will behave as a client machine, containing a driver software. All the existing functionality and software of EVM will be taken out and replaced with the Azure driver software, which will use to connect to Azure instance directly with the predefined end point. The endpoints are agreed upon in order to assure security where the data will be saved on Azure as read only. The Azure side end point will be further protected through firewall, multi-factor authentication (MFA) [12], Azure subscription-level security, and access control measure to assure the safety of the data during and after the transfer. In Fig. 4, we have explained the Azure data factory [13] in detail, which uses linked service, datasets, and pipeline instead of filename directly, on either end. This ensures security of the data stored within the file. The link service and dataset assure

Fig. 2 High-level architecture—proposed

Electronic Voting Machine as a Service on the Cloud—Azure for EVM …

357

Fig. 3 Connectivity between EVM-client to MS Azure ecosystem

Fig. 4 Azure data factory (data transfer between Azure storage)

that data is secure during transfer. Most of the connectivity is zero code, configuration only.

3 Low-Level Design Detail 3.1 Hardware to Software Mapping—EVM as a Docker Container The proposed design consists of 3 components, i.e., EVM, pipeline, and Azure ecosystem, which is explained separately in detail, in Sects. 3.2, 3.3, and 3.4, respectively. The mapping needs virtualization of EVM. In order to achieve that, we have defined a mapping between physical EVM to docker container in Azure ecosystem. The end point (EP: evm-id ↔ 0.0.0.0: ) is used to connect between physical EVM and docker registry in Azure side. The data will be transferred periodically (say every hour) to the Azure storage. The connection will be secure and encrypted using https (TLSv3) from source to destination. We can also use the Azure key-vault service for security purpose.

358

M. E. Hussain et al.

3.2 Docker Configuration # Base evm image FROM evm:1 # Create app directory WORKDIR /usr/src/evmapp # Install app dependencies ensure package.json and package-lock.json are copied COPY package*.json./ RUN evm install # Bundle evmapp source COPY.. EXPOSE 8080 CMD [ "evm", "server"]

3.3 Pipeline Between EVM and MS Azure Pipeline contains the following components for secure transfer of data from source to destination as shown in Fig. 5. It is achieved using Azure data factory, a cloudbased ETL (Extract, Load, and Transform) service for secure data movement. The linked service or the connection string is used to define the endpoints. The dataset represents the actual file to be transferred and the copy utility to copy the data from source dataset to destination dataset before moving to database table. The driver program sitting inside the modified EVM will connect to the linked service as shown in Fig. 5.

Fig. 5 Design of pipeline [5]

Electronic Voting Machine as a Service on the Cloud—Azure for EVM …

359

Fig. 6 Azure ecosystem

3.4 Azure Ecosystem The third component in the proposed design is the Azure cloud service, where multiple containers will be deployed within the cluster. Every container represents one virtual EVM. There will be one-to-one mapping between physical EVM machine and virtual docker container as a vEVM. This is shown in Fig. 6.

4 Result and Analysis The data once loaded onto the Azure storage, we can connect through power BI to analyze and visualize the result. The power BI-DAX code to achieve the same is below.

360

M. E. Hussain et al.

Winner Party = var maxVote = MAX ('ElectionResult'[Number of Votes]) var winnerCandidate = CALCULATE ( SELECTEDVALUE (ElectionResult [Party]), ElectionResult [Number of Votes] = maxVote ) return IF ( HASONEVALUE (ElectionResult [State]) && HASONEVALUE(ElectionResult [Year] ), winnerCandidate )

The data will be available in the database within Azure storage. The data is shown in Fig. 5 as a CSV file and the analysis of the result can be published as a dashboard shown in Table 1 [14]. The final report can be generated using power BI into a sample dashboard for visualization as shown in Fig. 7. Actual analysis may contain many such dashboards, not just one dashboard.

5 Conclusion and Future Work In this paper, we tried to solve the real issues related to electronic voting machine in the election system. In earlier days when cloud service was not available, there were no alternatives available, except using ballot paper. It is assumed that election done through ballot paper are fair and transparent, with no option to alter the result. Though it has other issues which are beyond the scope of this paper. The existing EVM saves data locally, i.e., within EVM itself. Therefore, it needs to be protected until counting is done. Now with the evolution of cloud service, such as Microsoft Azure as cloud service provider, a robust and foolproof system is available. Therefore, EVM can now be transformed into virtual electronic voting machine (vEVM). With few changes, as proposed, the existing machine can be transformed into a dummy, non-intelligent client machine, which can connect to the Azure ecosystem using a driver program and securely transfer data in batches. Once the election is over, the result can be published immediately, without waiting for several days (as happens today). The data can be saved forever, that helps analyzing the pattern in multivariate dimension, using powerful tools such as power BI and Azure Synapse service. The security is guaranteed by the cloud service provider. As part of this exercise, we have proposed

AC no

87

89

90

88

385

281

206

279

AC name

Agra Cantt

Agra North

Agra Rural

Agra South

Ajagara

Akbarpur

Akbarpur - Raniya

Alapur

SC

GEN

GEN

SC

GEN

SC

GEN

SC

Type

Ambedkar Nagar

Kanpur Dehat

Ambedkar Nagar

Varanasi

Agra

Agra

Agra

Agra

District

Party

Bharatiya Janta Party

Bharatiya Janta Party

Tribhuwan Dutt

Pratibha Shukla

Ram Achal Rajbhar

Tribhuwan Ram

Yogendra Upadhyaya

Samajwadi Party

Bharatiya Janta Party

Samajwadi Party

Bharatiya Janta Party

Bharatiya Janta Party

Baby Rani Maurya Bharatiya Janta Party

Purushottam Khandelwal

Dr. G S Dharmesh

Winning candidate

Table 1 Election result in CSV format saved to Azure [14] Total electors

341,484

326,545

334,855

374,215

368,236

427,669

437,969

469,018

Total votes

207,109

210,639

213,187

243,004

207,620

259,710

239,144

250,396

Poll%

60.60

64.50

63.70

64.90

56.40

60.70

54.60

53.40

Margin

9383

13,417

12,336

9160

56,640

76,608

112,370

48,697

Margin %

(continued)

4.50

6.40

5.80

3.80

27.30

29.50

47.00

19.40

Electronic Voting Machine as a Service on the Cloud—Azure for EVM … 361

AC no

103

76

262

263

261

AC name

Aliganj

Aligarh

Allahabad North

Allahabad South

Allahabad West

Table 1 (continued)

GEN

GEN

GEN

GEN

GEN

Type

Allahabad

Allahabad

Allahabad

Aligarh

Etah

District

Bharatiya Janta Party

Bharatiya Janta Party

Bharatiya Janta Party

Party

Sidharth Nath Singh

Bharatiya Janta Party

Nand Gopal Gupta Bharatiya Janta “nandi” Party

Harshvardhan Bajpai

Mukta Raja

Satyapal Singh Rathore

Winning candidate

460,837

411,157

447,231

396,401

342,369

Total electors

221,343

179,914

175,022

251,531

225,112

Total votes

48.00

43.80

39.10

63.50

65.80

Poll%

29,933

26,182

54,883

12,786

3810

Margin

13.50

14.60

31.40

5.10

1.70

Margin %

362 M. E. Hussain et al.

Electronic Voting Machine as a Service on the Cloud—Azure for EVM …

363

Fig. 7 Dashboard (result) using power BI tool

Fig. 8 Future expansion of the proposed work—ML and AI using Azure synapse

the complete end-to-end design, as well as testing methodology. The proposed design suggests that a drastic change is required to change the way electronic voting machine works today. Researchers, policymakers, and government need to work together to win the confidence of the public in democratic system. Implementing secure voting process will require hardware design, software implementation, and willpower. As a future extension of the proposed model, which could be used for any kind of high-level analysis using machine learning techniques. Microsoft Azure provides a Synapse analytics service tool, which brings together, the ingest, prepare, transform, and manage data for ready-to-do BI (Business Intelligence) and machine learning needs. The design of the same is shown in Fig. 8. Acknowledgements I would like to thank Dr. Mukesh Kumar Gupta and Dr. Rashid Hussain for their valuable feedback and encouragement, without their courageous approach these studies would not have been possible.

364

M. E. Hussain et al.

References 1. Dahlberg C (2008) Challenges in designing an electronic voting system. Citeseer 2. Bederson BB, Lee B, Sherman RM, Herrnson PS, Niemi RG (2003) Electronic voting system usability issues. In: Proceedings of the SIGCHI conference on human factors in computing systems, pp 145–152 3. Keller AM, Mertz D, Hall JL, Urken A (2006) Privacy issues in an electronic voting machine BT—privacy and technologies of identity: a cross-disciplinary conversation. In: Strandburg KJ, Raicu DS (eds). Springer US, Boston, MA, pp 313–334 4. Verified Voting. Summary of the problem with electronic voting. [Online]. Available: https:// verifiedvoting.org/publication/summary-of-the-problem-with-electronic-voting/ 5. Free online diagram software. https://seibert-media.com/, [Online]. Available: https://app.dia grams.net/ 6. Kumar DA, Begum TUS (2012) Electronic voting machine—a review. In: International conference on pattern recognition, informatics and medical engineering (PRIME-2012), pp 41–48. https://doi.org/10.1109/ICPRIME.2012.6208285 7. Banerjee S, Philipose P, Dayal J, Burra S, Devasahayam MG (2021) Citizens’ commission on elections’ report on EVMs and VVPAT 8. I. T. 21 J. 2019. R. 22 J. 2019 (2019) Motivated slugfest: election commission slams man claiming EVMs can be hacked 9. Balzarotti D et al (2009) An experience in testing the security of real-world electronic voting systems. IEEE Trans Softw Eng 36(4):453–473 10. Copeland M, Soh J, Puca A, Manning M, Gollob D (2015) Microsoft azure. New York, NY, USA Apress, pp 3–26 11. Collier M, Shahan R (2015) Microsoft azure essentials-fundamentals of azure. Microsoft Press 12. Ometov A, Bezzateev S, Mäkitalo N, Andreev S, Mikkonen T, Koucheryavy Y (2018) Multifactor authentication: a survey. Cryptography 2(1):1 13. Klein S (2017) Azure data factory. In: IoT solutions in Microsoft’s azure IoT suite. Springer, pp 105–122 14. 2022 Vidhan Sabha/Assembly election results Uttar Pradesh. [Online]. Available: https://www. indiavotes.com/vidhan-sabha/2022/uttar-pradesh [2000 onwards]/289/60

Internet of Bio-nano Things for Diabetes Telemedicine System with Secured Access Lokavya Gabrani, Rajeev Kumar Singh, Sonali Vyas, Sunil Gupta, and Goldie Gabrani

Abstract Internet of Bio-Nano Things (IoBNT) is the network of bio-nano devices and bio-nano nodes that are ingested in the human body to sense and transmit data from the human body for further processing. IoBNT deals with various issues such as whether the designated cells would be able to interact with smart devices or not and also aid in processing of the data collected. This requires studying in detail the IoBNT and also developing healthcare applications that can fully utilize the benefits of IoBNT—also referred to as nanomedicine. Further, with the advent of Telemedicine, it becomes imperative to combine the benefits of IoBNT and Telemedicine. The usage of IoBNT begins with the injectable design of the BNT implants that have the ability to recognize biochemical data from the bodies of humans and subsequently transmit the collected data to the wearable port outside the human body. This paper presents a study of IoBNT architecture with its components. It also discusses the utilization of bio-nano devices and bio-nano nodes for collecting real-time data when ingested in the human body. In this paper, a model has been proposed for remotely monitoring Diabetes in patients through IoBNT. The proposed Diabetic Telemedicine system will sense patients’ vitals via IoBNT and will transmit them to internet-based nanomedicine servers where they will be accessible to different stakeholders through

L. Gabrani (B) · R. K. Singh Department of Computer Science and Engineering, Shiv Nadar University, Noida, India e-mail: [email protected] R. K. Singh e-mail: [email protected] S. Vyas · S. Gupta School of Computer Science, UPES, Dehradun, Uttarakhand, India e-mail: [email protected] S. Gupta e-mail: [email protected] G. Gabrani Vivekananda Institute of Professional Studies, Technical Campus, New Delhi, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. Misra et al. (eds.), Internet of Things (IoT): Key Digital Trends Shaping the Future, Lecture Notes in Networks and Systems 616, https://doi.org/10.1007/978-981-19-9719-8_30

365

366

L. Gabrani et al.

a secured access. This proposed model promotes remote intrusive ways to investigate Diabetes through IoBNT. Keywords Internet of bio-nano things (IoBNT) · Biochemical data · Bio-nano devices · Bio-nano nodes · Diabetes · Telemedicine

1 Introduction The advancements in the area of nanotechnology, biotechnology and materials science have resulted in improvements in Internet of Things (IoT) applications. The interaction of Nano biosensors, Nano materials and Nano implants with present IoT networks has promoted the idea of Internet of Nano Things (IoNT), Internet of Ingestible Things. (IoIT), Internet of Bio-Nano Things (IoBNT) and Internet of Biodegradable Things (IoBDT) [1]. In order to improvise prevailing technologies and to launch completely novel applications, there has been a growing interest in these upcoming years related to applications utilizing nanotechnology ideas and tools. The basic idea of nanotechnology is the deceptive power of things at the cellular level. The use of bio-nanotechnology is considered in the current research trends as IoBNT that encompasses IoNT by means of embedded biological substances. IoBNT is seen as a varied network of Nano bio-devices, identified as Bio-Nano Things (BNTs), linked through unconventional means like molecular communications (MC), in progressive systems, such as the internal body of humans. The goal of this upcoming networking framework is to permit direct and continuous communication with biological systems for exact measurement and control of their dynamics which will be done in realtime. This strong connection amid the bio domain and cyber world with the highest unprecedented localization is assumed to unlock many opportunities to establish new programs, mainly in the area of health care, like ongoing physical health monitoring [2]. Advanced methods of treatment and diagnosis of diseases with nanotechnology guarantees to develop a new era of healthcare. Applying IoT concepts to the nanoscale grant efficient health monitoring which results in suitable and prompt actions [3]. Current diagnostic techniques are most of the times based on the culture of microorganisms that cause infections found in samples that are collected from patients or making use of polymerase chain reaction (PCR) which requires the use of heating as well as cooling of large samples along with the reagents to obtain enzymatic reactions in order to recognize micro-cell formation. The IoBNT architecture is different from prevailing technologies by allowing regular disease monitoring by implanting nanoscale sensors or nodes that detect an interaction between contagious microorganisms inside the human body. After that, such sensors report to a wearable port that transmits the collected data to health care staff. Therefore, the patient need not go to the medical lab for testing and diseases can be diagnosed timely, and informing patients to gain medical counseling. By this, the premature death risk of patients can be minimized [4].

Internet of Bio-nano Things for Diabetes Telemedicine System …

367

The architecture of IoBNT includes the following components: • Bio-Nano nodes are the simplest and smallest part of the IoBNT configuration which can carry out activities like transmitting data, broadcasting, and performing basic calculations. They are responsible for all the sensing required for this setup. Their size limits their data transmitting capabilities and have an internal memory that is small in size. They are able to transfer the data collected to a nano router and these nodes are placed inside the human body itself. • Bio-cyber port is a cross tool which changes the biochemical signals taken from internal body nanonetworks into electric signals in order to get computed by the exterior network. • Nano routers are more progressive structures as compared to nano nodes when it comes to storage and computing. They collect data from nano nodes and monitor the nano node with basic control instructions. Due to their bigger size, they have a larger computational power and act as the aggregator for all surrounding nano nodes. They send their data to the nano micro interface. • Nano micro interfaces are combination devices which are able to connect all the data received from nano routers and then send it across to the gateways using nano communicating techniques as well as the classical network protocol. • Gateways are responsible for controlling the full setup and they allow the data collected to be accessed anywhere using the internet. • Application definite servers may be used in some applications for the examination, storage and instantaneous controlling of data from nano networks [5]. In Sect. 2, the related work supporting our ideology is presented. In Sect. 3, the proposed model of IoBNT in Diabetic Telemedicine System is briefed followed by an explanation of various phases of the projected model and authentication procedures for accessing the patient’s data. Then, in Sect. 4, the paper is concluded and some further work is suggested along with the benefits of using the proposed model.

2 Related Work Many of the world’s current problems can be resolved through manufacturing using nanotechnology. A more plausible assumption is that nanotechnology will affect every element of existence. Nanomedicine has several uses, including the creation of innovative imaging and diagnostic tools, stronger medications, and drug delivery systems [6]. Nanotechnology offers a wide range of uses in the medical field, including the creation of innovative diagnostic and imaging tools, stronger medications, and innovative implantable devices and drug delivery systems [7]. Nanomedicine are being used in drug delivery systems that use nanotechnology. Nano sensors are being employed in the healthcare implementation of nanotechnology monitoring applications for healthcare, where they can track most conditions, including blood lipids, pressure, sugar, and temperature [8]. As part of the ecosystem

368

L. Gabrani et al.

Fig. 1 IoNT architecture components

for timed medication delivery, the updated nano scale sensors are now capable of monitoring a number of physical and biological variables, including wound healing and the level of hormones present in the human body [9]. A fresh and contemporary component of IOT is the Internet of Nano Things (IoNT). Healthcare has been consistently impacted by IoNT and has undergone significant change as a result, which has helped to improve outcomes [3]. It has numerous applications in a variety of industries, including agriculture and food, environmental conservation, energy and healthcare, and these are already relying heavily on IoNT [3]. The technologies of utmost importance which are used in Nano communication networks have been recognized, along with the structure of the IoNT as well as the advantages of each component. The Internet of Nanoscale Things (IoNT) has two subfields: the Multimedia Internet of Nanoscale Things (IoMNT) and the Internet of Nanoscale Bio-Things (IoBNT). Both entails establishing connections between nanodevices and existing communication networks. The architectures of the IoNT network are influenced by the application domain and its distinctive features. The IoNT network’s design is made up of four key elements (as shown in Fig. 1): • • • •

Nano nodes Nano routers Nano Micro Interface Devices Gateway

The drug delivery systems incorporating nanotechnology are designed to offer non-invasive drug administration methods while also increasing the bioavailability and pharmacokinetics of medications. Beyond the delivery of drugs, it is crucial for a swarm of nano or microdevices to work together in unison which is implemented in the heavy industry and defense applications of the future to accomplish precise embedded sensing and actuation [10]. Liposomes, nanosuspensions, carbon nanotubes and inorganic nanoparticles are some of the specific areas of drug delivery methods using nanomaterials that are currently being developed. Realizing essential

Internet of Bio-nano Things for Diabetes Telemedicine System …

369

elements of the IoNT and IoBNT paradigms and enabling nanoscale communications will usher in a new era of nanomedicine [11]. A Targeted Drug Delivery (TDD) is a specific type of delivery system where the drug is only given to the site where the action or the absorption would take place and not to the target organs, cells or even the tissues [12]. An existing biological system or biochemical molecules serve as the basis for a nanodevice’s design. These nanodevices can be utilized for illness detection, monitoring, and treatment, as well as analytical and imaging tools. Other than this some of the other areas of utilization include biochemical sensors which are used in monitoring and tissue engineering which is used in tissue repair as well as reengineering [13]. A heterogeneous network of Bio-Nano Things (BNTs), or nanoscale and biological devices, is IoBNT as previously mentioned [7]. IoBNT would be able to elevate connectivity as well as control over unconventional domains (like the human body) with unheard of spatio-temporal resolution. This would enable a new revolution of applications, particularly in the healthcare sector which involves intrabody health monitoring that is continuous in nature and on top of that theragnostic systems which involve single molecular precision [2]. The nervous system which is made up of a network at a large scale comprising of neurons connected with the help of neuro-spike and synapse communication channels, is the most sophisticated and complicated human body network. The cardiovascular system and the endocrine system are two networks that are present throughout the body but carry information at a time scale which is slower as compared to the electrochemical pulses of nervous nanonetworks. Both of these systems are made up of vessels that transport molecules that carry information in the form of blood as well as the lymph [14]. One of the biggest issues based on IoBNT technology is the issue of security and privacy during the usage of this technology, which has many sensitive applications as well as its work outside the human body. As a result, there is a greater requirement for safety to protect people and property due to the sensitivity of the information transmitted by this technology. Because IoBNT operates inside terahertz bounds, additional security measures that are compatible with this technology are required to avoid data theft and user damage. Attempts to prevent therapeutic injection techniques, data theft, and changing linkages present in nano communication levels or a BAN gateway are some of the biggest security risks [15].

3 Proposed Model for IoBNT The IoBNT defines Bio-Nano Things as basic structural and functional elements which interact and perform in a biological environment. It is seen that BNT would perform tasks and actions similar to those of Internet of Things integrated with computer devices, including sensing, processing, and communication with other Bio-Nano things. These Bio-Nano things are created using synthetic biology and nanotechnology and are generated from biological cells. These bio-nano devices are based on living cells which are part of the particular sub-structure of bio-nano objects because biochemicals, molecules, or existing biological systems must be

370

L. Gabrani et al.

used to create them. In the proposed model, we provide a system for diabetes remote monitoring. Diabetes is an amalgamation of metabolic illnesses that is characterized by high levels of sugar or glucose in the blood. The medical community refers to this persistent and common disease as diabetes mellitus. According to one of the many definitions for this condition that have been established, it has been described as “a metabolic illness with many etiologists that is defined by chronic hyperglycaemia and abnormalities in the metabolism of proteins, lipids, and carbohydrates because of errors in insulin secretion and/or action.” Fig. 2 shows the proposed model of Internet of Bio-Nano Things for Diabetes Telemedicine System. It illustrates the interconnection process that takes place between the different parts of a network of this type. The Nano nodes are joined via Nanonetworks to enable the collection of vital data in challenging locations. Nano nodes present in the human body can collect and send signals to one another using the bio-cyber port that converts biochemical signals to electric signals that carry important information for the continuous monitoring of blood glucose levels. To do their tasks, these components analyze blood glucose levels, and process and transmit data from the bio-cyber port to the nano router which acts as an aggregator for all nano nodes. The nano router sends all the information to the nano micro interfaces which in turn acts as an aggregator for all the nano routers. This information gets passed onto conventional-sized devices, using the gateway and conveys alerts to the users about potential problems or changes in the situation. Nanomedicine, one of the branches of nanotechnology, offers the possibility of identifying and monitoring illnesses at the cellular and molecular levels. Providing diagnostic, preventative, and therapeutic tools for diseases that are difficult to control, diabetes, kidney failure, cancer and the development of HIV contagion, is the main objective in this scenario. Diabetes Telemedicine System is a communication tool between patients, internet-based servers for nanomedicine, and healthcare providers. In this system, a nano biosensor aids patients in gathering their capillary glucose, blood pressure, and other physiological data. The sensor continuously monitors the blood glucose levels and the information is sent to the server. The model requires security since data from a body sensor contains sensitive and private information. We also implement an authentication process so that the server and sensor may establish mutual trust.

Fig. 2 Internet of bio-nano things for diabetes telemedicine system with secured access

Internet of Bio-nano Things for Diabetes Telemedicine System …

371

The goal of the authentication approach is to enable patients to access medical services from home. In this approach, patients’ illnesses are monitored remotely and prescriptions and suggestions are given to patients based on symptom intensity, age, and location. This model has a central cloud-based telemedicine server, telemedicine service providers, and patients (or users). Patients can easily access medical services using their internet-connected mobile devices (phones, laptops, etc.). Telemedicine service providers are doctors, medical teams, clinics, or hospitals anywhere.

3.1 Registration Phase • Patient registration Patients can use mobile devices. Registration requires entering an ID, password, name, address, date of birth and phone number. This data is masked for security reasons before being sent to the server. The server uses the OTP to verify the user, checks the database and registers the user if the ID received does not already exist in the server database. • Healthcare provider registration Telemedicine service provider registration is done through a secure channel, verified and provided with an ID and password. After registration, telemedicine service providers can use the ID and password provided to log in, see patients, view data shared by patients, and communicate with patients. Secure Smart Card Generation for Cloud Server Access to User The User can be both the healthcare provider as well as the patient. The user ‘U’ sends a request to Cloud Server (CS) for registration and executes the following steps. Figure 3 explains the steps performed in the registration phase. Step 1: The user Uj selects its identity IUj, random number Rj and the password PUj. The user Uj completes the random identity of itself as Fig. 3 Smart card generation for user to access cloud server

372

L. Gabrani et al.

PUj = H(IUj||Rj) The PUj is sent to a cloud server for the registration of the user Uj using a secure communication link. Step 2: The CS, after receiving PU j from user Uj, then compute a message Mi = H(PUj||d), where H is the one-way hash function and d is the private key of the gateway. Step 3: The CS generates a smart card SCj = {Mi, H (.), a, b}. and sends it to user Uj using a secure communication link.

3.2 Login and Authentication Phase The patient enters their ID and password, time-stamps it, and sends it to the server. The server authenticates the user using ID, password, timestamp and OTP. After the patient is successfully authenticated by the server, the patient also authenticates with the server, and upon successful mutual authentication, a session key is generated and validated for further communication. Data is shared using security protocols to ensure secure transmission and access to data. For ease of use, we do not use biometrics. Elderly patients and rural patients may not have fingerprint scanners. Upon successful login, the patient can view his or her report and share the data with the telemedicine service provider.

4 Conclusion Nowadays, the use of nano technology is gradually increasing in the field of healthcare and medicine. The technological advancements in the field of IoT are aiding in the increased use of this technology. The amalgamation of IoT and Bio-nano technology is known as IoBNT and it is making a significant impact on the way medicines are delivered and the way they work. Targeted Drug Delivery along with IoBNT is going to be used in the utmost quantities in the near future. In this paper, we have put forward a model for monitoring blood glucose levels using IoBNT. This process involves the use of IoBNT for collecting the blood glucose levels, analyzing it and then transmitting this data outside of the body. All this procedure is made secure by adding an authentication protocol as well because the data that is being transmitted is sensitive. The authentication protocol contains two phases, namely the Registration phase and the Login and Authentication phase. With the implementation of this model, the treatment for diabetes can be done in an efficient and convenient manner along with it being secure as well. In the future, a model like this can even be used

Internet of Bio-nano Things for Diabetes Telemedicine System …

373

for maintaining the glucose levels whenever it is not in the advised range with the help of nanomedicines.

References 1. Senturk S, Kok I, Senturk F (2022) Internet of nano, bio-nano, biodegradable and ingestible things: a survey. arXiv:2202.12409 2. Kuscu M, Unluturk BD (2021) Internet of bio-nano things: a review of applications, enabling technologies and key challenges. arXiv:2112.09249 3. Maksimovi´c M (2017) The roles of Nanotechnology and Internet of Nano things in healthcare transformation. TecnoLógicas 20(40):139–153 4. Akyildiz IF, Ghovanloo M, Guler U, Ozkaya-Ahmadov T, Sarioglu AF, Unluturk BD (2020) PANACEA: an internet of bio-nanothings application for early detection and mitigation of infectious diseases. IEEE Access 8:140512–140523 5. Mohamed S, Dong J, El-Atty SMA, Eissa MA (2022) Bio-cyber interface parameter estimation with neural network for the internet of bio-nano things. Wireless Pers Commun 123(2):1245– 1263 6. Omanovi´c-Mikliˇcanin E, Maksimovi´c M, Vujovi´c V (2015) The future of healthcare: nanomedicine and internet of nano things. Folia Medica Facultatis Medicinae Universitatis Saraeviensis 50(1) 7. Miller G, Kearnes M (2012) Nanotechnology, ubiquitous computing and the internet of things. Council of Europe Report 8. Naser HA, Lateef AT, Bida FA, Zorah M (2021) Systematic review of internet of nano things (IoNT) technology: taxonomy, architecture, open challenges, motivation and recommendations. Iraqi J Nanotechnol 2:7–19 9. Youan BBC (2010) Chronopharmaceutical drug delivery systems: hurdles, hype or hope? Adv Drug Deliv Rev 62(9–10):898–903 10. Guo W, Wei Z, Li B (2020) Secure internet-of-nano things for targeted drug delivery: distancebased molecular cipher keys. In: 2020 IEEE 5th Middle East and Africa conference on biomedical engineering (MECBME). IEEE, pp 1–6 11. Akyildiz IF, Pierobon M, Balasubramaniam S, Koucheryavy Y (2015) The internet of bio-nano things. IEEE Commun Mag 53(3):32–40 12. El-Fatyany A, Wang H, El-atty A, Saied M (2021) Efficient framework analysis for targeted drug delivery based on internet of bio-nanothings. Arab J Sci Eng 46(10):9965–9980 13. Cai P, Zhang X, Wang M, Wu YL, Chen X (2018) Combinatorial nano–bio interfaces. ACS Nano 12(6):5078–5084 14. Malak D, Akan OB (2012) Molecular communication nanonetworks inside human body. Nano Commun Netw 3(1):19–35 15. Dressler F, Fischer S (2015) Connecting in-body nano communication with body area networks: challenges and opportunities of the internet of nano things. Nano Commun Netw 6(2):29–38

Author Index

A Aaryan Bhatia, 49 Aaryan Rastogi, 113 Aditi Bhole, 103 Aditya Khazanchi, 49 Akanksha Akanksha, 283 Akshay Tanpure, 89 Ananya Tyagi, 49 Anchal Koshta, 63 Aneerudh, M., 305 Anshika Jain, 27 Anshuta Kakuste, 103 Anup Kumar Das, 243 Arohi Singhal, 27 Ashim Mondal, 151 Atharva Suryavanshi, 113

B Balasubramanian, G., 165 Bhumika Jain, 27

C Chandan Koner, 243 Chandrani Singh, 63, 77, 89

D Daya Sagar Gupta, 315 Deepali Virmani, 1 Deivalakshmi, S., 197 Deo Prakash Vidyarthi, 327 Dey, B., 221 Dhanvardini, R., 233

Dheeraj Kallakuri, 123 Dushyant Bodkhey, 77

E Ekta Verma, 327

G Garvita Ahuja, 11 Gerard Deepak, 233 Girish Mogalgiddikar, 77 Goldie Gabrani, 365

H Harish Chandra, 315

I Ishan Sharma, 49

J Jyotsna Yadav, 327

K Kavitha Chekuri, 211 Ketan Parikh, 1 Kumudini Manwar, 77

L Lacchita Soni, 315

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. Misra et al. (eds.), Internet of Things (IoT): Key Digital Trends Shaping the Future, Lecture Notes in Networks and Systems 616, https://doi.org/10.1007/978-981-19-9719-8

375

376 Lokavya Gabrani, 365

M Maanik Sharma, 11 Madhab Paul Choudhury, 179 Manpreet Kaur, 255 Milind Godase, 89 Mohammad Equebal Hussain, 353 Mukesh Kumar Gupta, 353

N Nazneen Alam, 39 Nidhi Shrivastav, 113 Nikhil Londhe, 123

P Palak Wadhwa, 113 Pallav Dutta, 139, 151 Paul Choudhury, J., 179 Pratiksha Mahamine, 77 Preeti Jawla, 39 Priyal, 49

R Rachit Khandelwal, 197 Radhika Chawla, 49 Rahul Johari, 327 Rajdeep Ray, 243 Raj Desai, 269 Rajeev Kumar Singh, 365 Rakhi Kalantri, 103, 113, 123 Ramya, T., 165 Rashi Aggarwal, 255 Rashid Hussain, 353 Rita Roy, 39, 211 Routhu Shanmukh, 211 Rowthu Lakshmana Rao, 211 Rumpa Saha, 139, 151 Rutu Parekh, 269

Author Index S Santhanavijayan, A., 233 Shagufta, R., 113 Shagufta Rajguru, 103 Shane Rex, S., 305 Sharon Laurance, 123 Sheeba Priyadarshini, J., 233 Shivansh Sharma, 11 Simran Khanna, 339 Sivaji Bandyopadhyay, 221 Sonali Vyas, 365 Sri Sankar, S., 165 Srishti Singh, 11, 27, 49, 339 Subhodeep Mukherjee, 39, 211 Sudiksha Mullick, 103 Suhas Harbola, 327 Sukumar Nandi, 221 Sunil Gupta, 365 Sunil Khilari, 63 Surya Kant Pal, 39

T Tanmay Mishra, 339 Tripathi, K. C., 255

U Uday Kumar Banerjee, 243 Udit Meena, 269 Utkarsh Asari, 269

V Vansh Bhandari, 339 Vijayalakshmi, M., 305 Vinayak Kurup, 123 Vishalakshi Prabhu, H., 293 Vishal, M., 293 Viswesh, S., 165

Y Yash Shrivastav Yashas Bajaj, 339 Yeswanth, P. V., 197