Advances in Resilient and Sustainable Transport: Proceedings of the 6th Interdisciplinary Conference on Production, Logistics and Traffic 2023 3031282353, 9783031282355

Citation preview

Lecture Notes in Logistics Series Editors: Uwe Clausen · Michael ten Hompel · Robert de Souza

Uwe Clausen Marius Dellbrügge   Editors

Advances in Resilient and Sustainable Transport Proceedings of the 6th Interdisciplinary Conference on Production, Logistics and Traffic 2023

Lecture Notes in Logistics Series Editors Uwe Clausen, Institute of Transport Logistics, TU Dortmund University, Dortmund, Germany Michael ten Hompel, Fraunhofer Institute for Material Flow and Logistics IML, Dortmund, Germany Robert de Souza, The Logistics Institute - Asia Pacific, National University of Singapore, Singapore, Singapore

Lecture Notes in Logistics (LNL) is a book series that reports the latest research and developments in Logistics, comprising: • • • • • • • • • • • • • • • • • • •

supply chain management transportation logistics intralogistics production logistics distribution systems inventory management operations management logistics network design factory planning material flow systems physical internet warehouse management systems maritime logistics aviation logistics multimodal transport reverse logistics waste disposal logistics storage systems logistics IT

LNL publishes authored conference proceedings, contributed volumes and authored monographs that present cutting-edge research information as well as new perspectives on classical fields, while maintaining Springer's high standards of excellence, the content is peer reviewed. Also considered for publication are lecture notes and other related material of exceptionally high quality and interest. The subject matter should be original and timely, reporting the latest research and developments in all areas of logistics. The target audience of LNL consists of advanced level students, researchers, as well as industry professionals working at the forefront of their fields. Much like Springer's other Lecture Notes series, LNL will be distributed through Springer's print and electronic publishing channels. ** Indexed by Scopus (2021) **

Uwe Clausen Marius Dellbrügge •

Editors

Advances in Resilient and Sustainable Transport Proceedings of the 6th Interdisciplinary Conference on Production, Logistics and Traffic 2023

123

Editors Uwe Clausen Institute of Transport Logistics TU Dortmund University Dortmund, Germany

Marius Dellbrügge Institute of Transport Logistics TU Dortmund University Dortmund, Germany

ISSN 2194-8917 ISSN 2194-8925 (electronic) Lecture Notes in Logistics ISBN 978-3-031-28196-9 ISBN 978-3-031-28236-2 (eBook) https://doi.org/10.1007/978-3-031-28236-2 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 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 Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Preface

The series of Interdisciplinary Conferences on Production, Logistics and Traffic (ICPLT) address the research community as well as practitioners in these fields with special attention to links and interfaces between the three disciplines. The sixth ICPLT in particular deals with technology to plan and operate freight transport systems in a sustainable and resilient way. It took place on March 21 and 22, 2023, at TU Dortmund University, representing a joint effort by TU Dortmund University and TU Darmstadt University. Transport demand is driven by production and consumption demand, affected by manifold trends both in societies and by new technologies. The COVID-19 pandemic has caused global disruptions in production systems and supply chains. A new understanding of the importance of logistical systems and a stronger focus on their resilience can be observed. We have observed progress in relevant technologies such as machine learning and automation. We have seen the need for supply chains to be designed in more resilient and sustainable ways and the potential benefit of research in general and interdisciplinary research in particular. The simultaneous reduction of fuel demand by interrupted mobility and transport activities has, for the first time in decades, also led to a reduction of global oil production and, in consequence, to reduced greenhouse gas emissions. This should further encourage logistics experts to contribute to efficiency gains and demand management solutions as well as to promote alternative energy utilization to contribute to emission reduction goals as part of the global effort of climate action. International networks follow the intensified division of labor and exploit the capabilities of all transport modes. Allowing products to be distributed almost everywhere, they have helped global prosperity and thus also contributed to the United Nations’ No. 1 strategic development goal to eradicate extreme poverty. Sustainability goals and political tensions form a challenging environment for global production networks, logistics and traffic. The military invasion of Ukraine by the Russian army on February 24, 2022, as unprecedented escalation of an ongoing conflict since 2014 has led to largest refugee crisis since World War II and ongoing turbulence on energy markets in Europe. Against this background, the sixth ICPLT closely cooperates with its long-standing partners from academia from 42 institutions in 14 countries, this time with special sessions from the WCTRS SIG B3 “freight transport operations and intermodality” and FGSV AA 1.8 “freight transport” included in the conference program. To contribute to a high-level and beneficial exchange between authorities in politics and municipalities with researchers and practitioners in production and logistics management, the ICPLT has asked for contributions from the three disciplines to better understand innovative technologies, best practices and latest results. These contributions have been evaluated and selected based on a double-blind review process to become part of this book. It comprises 22 contributions examining

vi

Preface

trends and challenges, models and assessments, algorithms and knowledge gaps, traffic flows and logistic processes. Innovative technologies as well as human factors and strategies are presented and discussed to better understand the interdependencies, conflicts of interest and to develop feasible solutions in the field of production, logistics and traffic. The focus of this book is on the following core topics: • • • • • • • • • •

Freight Transport Public Transport Demand Modeling Cargo Bikes Maritime and Rail Transport Electrical and Hydrogen Vehicles Simulation and Optimization Production and Supply Chain Management Sustainable Logistics Intralogistics and Automation.

I am thankful for all these valuable contributions as well as for sponsoring from the logistics industry, this time especially DB Schenker, allowing us to strengthen research excellence in the scientific fields of production, logistics and traffic and promote exchange with planning and operations management practitioners. Furthermore, we were very pleased about three very interesting keynotes from business and research: Thank you, Dr. Maria Tsavachidis (CEO of EIT Urban Mobility) for presenting the topic “Mobility - lessons learned in European cities and the road ahead”, Dr. Joachim Weise (Senior Vice President Data Strategy and Analytics at DB Schenker) for sharing your insights on “Resilient Networks - Optimizing Global Logistics” from a corporate point of view and Prof. Dr. Hubertus Bardt (German Economic Institute in Cologne) for presenting latest results regarding “Supply Chains, Costs and Investments in Times of Multiple Crises” from a political economy point of view. As Chair of the sixth ICPLT, I would also like to thank all speakers for their interesting speeches and all authors for their contributions, my Co-Chair Ralf Elbert and all members of the Scientific Committee for their advice, everyone in our Organization Committee, led by Marius Dellbrügge from the Institute of Transport Logistics (ITL) at TU Dortmund University, for their efforts and commitment to the success of the conference and this book. March 2023

Uwe Clausen

Acknowledgements

The editors would like to thank all involved members of Conference Chair and the Scientific Committee of sixth ICPLT, who reviewed the contributions to assess their scientific and practical relevance, quality and originality. Without the untiring and voluntary effort of the Scientific Committee, this publication would not have been possible. In particular, the editors also would like to thank all authors for sharing their work and the Organization Committee for their effort for making sixth ICPLT possible. The editors would like to express its sincere thanks to DB Schenker and in particular to the Global Data Strategy & Analytics Department for their generous sponsorship of the sixth Interdisciplinary Conference on Production, Logistics and Traffic.

Organization

Scientific Committee of the 6th ICPLT Michael Bourlakis Michael Browne Laetitia Dablanc René De Koster

Gerard DeJong Gerald Ebel Verena Ehrler Alexander Eisenkopf Pietro Evangelista

Nathalie Fabbe-Costes Heike Flämig Hanno Friedrich Christoph Glock Michael Henke Carlos Jahn Eva Kaßens-Noor Danuta Kisperska-Moron Bernd Kuhlenkötter

Chair of Logistics, Procurement and Supply Chain Management, Cranfield University, UK Department of Business Administration, University of Gothenburg, Sweden Chair Logistics City, University Gustave Eiffel, France Department of Technology and Operations Management, Erasmus University Rotterdam, The Netherlands Institute for Transport Studies, University of Leeds, UK Lehrgebiet Logistik und BWL, FH Bielefeld University of Applied Sciences, Germany International Supply Chain Management and Logistics, IÉSEG School of Management, France Lehrstuhl für Wirtschafts- und Verkehrspolitik, Zeppelin University Friedrichshafen, Germany Research Institute on Innovation and Services for Development, National Research Council Naples, Italy Transportation and Logistics, Aix-Marseille Université, France Institut für Verkehrsplanung und Logistik, Technische Universität Hamburg-Harburg, Germany Freight Transportation-Modelling and Policy, Kühne Logistics University, Germany Production and Supply Chain Management, Technische Universität Darmstadt, Germany Lehrstuhl für Unternehmenslogistik, TU Dortmund University, Germany Institut für Maritime Logistik, Technische Universität Hamburg-Harburg, Germany Instituts für Verkehrsplanung und Verkehrstechnik, Technische Universität Darmstadt, Germany Department of Business Logistics, University of Economics Katowice, Poland Lehrstuhl für Produktionssysteme, Ruhr-University Bochum, Germany

x

Organization

Sebastian Kummer Bert Leerkamp

Barbara Lenz Tone Lerher Gernot Liedtke Alan McKinnon Herbert Meyr Stefan Minner Markus Muschkiet Hans-Christian Pfohl Markus Rabe Stefan Seuring Carsten Sommer Frank Straube Lóránt Tavasszy Michael ten Hompel Inge Vierth Jaroslaw Witkowski

Instituts für Transportwirtschaft und Logistik, Wirtschaftsuniversität Wien, Austria Lehr- und Forschungsgebiet für Güterverkehrsplanung und Transportlogistik, Bergische Universität Wuppertal, Germany Institut für Verkehrsforschung, German Aerospace Center (DLR), Germany Department of Technical Logistics, University of Maribor, Slovenia Institut für Verkehrsforschung, German Aerospace Center (DLR), Germany Logistics and Supply Chain Management, Kühne Logistics University, Germany Supply Chain Management, University of Hohenheim, Germany Logistics and Supply Chain Management, Technische Universität München, Germany Center Textillogistik, Hochschule Niederrhein University of Applied Sciences Chair of Supply Chain and Network Management, Technische Universität Darmstadt, Germany Chair of IT in Production and Logistics, TU Dortmund University, Germany Supply Chain Management, University of Kassel, Germany Fachgebiet Verkehrsplanung und Verkehrssysteme, University of Kassel, Germany Logistics, Technische Universität Berlin, Germany Faculty of Technology, Policy and Management, Delft University of Technology, The Netherlands Lehrstuhl für Förder- und Lagerwesen, TU Dortmund University, Germany Transport Economics, VTI Stockholm, Sweden Strategic Management and Logistics, Wroclaw University of Economics, Poland

Organization

Organization Committee of the 6th ICPLT 2023 Uwe Clausen Marius Dellbrügge Christiane Mieles Claas Langenbach Hannah Scheerer

Institute of Transport Logistics, University, Germany Institute of Transport Logistics, University, Germany Institute of Transport Logistics, University, Germany Institute of Transport Logistics, University, Germany Institute of Transport Logistics, University, Germany

TU Dortmund TU Dortmund TU Dortmund TU Dortmund TU Dortmund

xi

Conference Chairs, Editors and Keynote Speaker

Prof. Dr.-Ing. Uwe Clausen (Chair and Editor) Prof. Dr.-Ing. Uwe Clausen, born in 1964 in Düsseldorf, studied computer sciences at the University of Karlsruhe (TH) and graduated at TU Dortmund University. After having worked as Scientific Employee, he became Head of the Traffic Logistics department at Fraunhofer IML. Subsequently, he worked for Deutsche Post AG as Project Manager for logistics and later on as Managing Director of the subsidiary IPP Paketbeförderung GmbH in Austria. In July 1999, he joined Amazon.de in Bad Hersfeld, and, finally, he became European Operations Director at Amazon.com before he returned to Fraunhofer IML on February 1, 2001, as one of the directors of the Institute. At the same time, he started his work as Director of the Institute of Transport Logistics. From July 2002 until July 2005, Prof. Clausen was Dean of the faculty of mechanical engineering at TU Dortmund University, and from July 2005 until August 2008, he was Vice Dean of the very same faculty. Today, Prof. Dr.-Ing. Uwe Clausen is Director of the Institute of Transport Logistics at TU Dortmund University and Director of Fraunhofer-Institute for Material Flow and Logistics IML. He is also representing Fraunhofer in ECTRI European Conference of Transport Research Institutes, Member of the Advisory Council of the Association of German Transportation companies (VDV) and the scientific advisory board of the Bundesvereinigung Logistik (BVL) e.V.. He has been Member of the National Academy of Science and Engineering (acatech) since 2019. His research areas include green logistics, commercial traffic modeling, intermodal transportation, mathematical optimization, network optimization and distribution systems.

xiv

Conference Chairs, Editors and Keynote Speaker

Prof. Dr. Ralf Ebert (Co-chair) Prof. Dr. Ralf Elbert is Professor and Chair of Management and Logistics at TU Darmstadt since 2011. From 2009 to 2011, he held an assistant professorship at TU Berlin for Logistics Services and Transportation. Prof. Dr. Ralf Elbert represents his Chair of Management and Logistics in various national and international organizations and associations, including the scientific advisory board of the Bundesvereinigung Logistik (BVL) e. V., the Transportation Research Board (TRB) and the World Conference on Transport Research Society (WCTRS). His research focuses on the management and planning of transportation networks (especially intermodal transportation networks), specifically on the analysis of freight mode choice decisions, efficiency improvements by information sharing and measures for increasing utilization of transport capacities. Further research fields are warehouse management and the integration of human factors in intralogistics systems as well as the management of logistics and production networks. Simulation modeling is the preferred research method throughout most of his work. Marius Dellbrügge, M. Sc. (Editor) Marius Dellbrügge studied business administration with a specialization in logistics at FH Münster University of Applied Sciences. After completing his bachelor’s degree, he studied Supply Chain Management (M. Sc.) at the SRH University of Applied Sciences in North Rhine-Westphalia. During the time of his studies, he completed an apprenticeship as a forwarding agent and gained national and international practical experience in the field of food logistics, tender management and calculation in the corporate headquarters of the Nagel-Group in Versmold. Since February 2022, he has been working as Research Assistant at the Institute of Transport Logistics, TU Dortmund University, Germany. His research areas include transport modeling, process analysis and optimization, green logistics and urban freight transport. With this expertise, he took over the leadership of the group “transport modeling and process planning” in October 2022.

Conference Chairs, Editors and Keynote Speaker

xv

Dr. Maria Tsavachidis (Keynote Speaker) Dr. Maria Tsavachidis is the CEO of the EIT Urban Mobility, a European knowledge and innovation community to accelerate the transition toward sustainable urban mobility and liveable urban spaces. Before joining the European Institute for Innovation and Technology in 2018, she was responsible for Innovation at Siemens for more than 20 years in different positions. She started her career as Researcher in the field of Intelligent Transport Systems and holds a PhD in traffic engineering from the Technical University of Munich. EIT Urban Mobility is an initiative of the European Institute of Innovation and Technology (EIT). Since January 2019, we have been working to encourage positive changes in the way people move around cities in order to make them more liveable places. We aim to become the largest European initiative transforming urban mobility. Co-funding of up to € 400 million (2020–2026) from the EIT, a body of the European Union, will help make this happen. Dr. Joachim Weise (Keynote Speaker) Dr.-Ing. Joachim Weise is Senior Vice President, Data Strategy & Analytics at DB Schenker. The Global Data Strategy & Analytics department comprises teams of Business Consultants, Data Scientists, Operations Research specialists and Data Engineers who, in close cooperation with business units, develop and operate advanced analytics solutions and drive the shift to a data-driven culture. Prior to his current role, Joachim Weise has leveraged Business Analytics for 15+ years in top management consulting at The Boston Consulting Group and in corporate development roles at Deutsche Bahn AG. Joachim holds a PhD in Innovation Management from the Technical University of Berlin, an MSc in Production Management from Chalmers Tekniska Högskola Göteborg and a Diploma in Industrial Engineering from the Technical University of Berlin. With some 76,100 employees at over 1,850 locations in more than 130 countries, DB Schenker is one of the world’s leading logistics providers. The company operates land, air and ocean transport services and offers comprehensive solutions for logistics and global supply chain management from a single source.

xvi

Conference Chairs, Editors and Keynote Speaker

Prof. Dr. Hubertus Bardt (Keynote Speaker) Prof. Dr. rer. pol. Hubertus Bardt, born 1974 in Bonn, completed his studies of economics at the Philipps-University Marburg and of business administration at the FernUniversität in Hagen. He received his doctorate on the topic ““Labor” versus “Capital” - on the transformation of a classical conflict” from the Philipps-University Marburg in 2003. Since 2000, Prof. Dr. Bardt has been employed at the German Economic Institute in Cologne. Since 2014, he has been Managing Director of the German Economic Institute in Cologne and Head of Science. Since 2022, he has been Honorary Professor at the Heinrich Heine University Düsseldorf. His main research interests are economic policy, industrial policy and climate economics. The German Economic Institute (IW) is a private economic research institute in Germany, which is an advocate of a liberal economic and social order. We work to improve understanding of how business and society function and interact

Conference Chairs, Editors and Keynote Speaker

xvii

Societies and Associations The Road and Transportation Research Association (FGSV) prepares the technical regulations for the entire road and traffic sector in Germany. It is the independent German competence network for research and knowledge transfer in the entire road and traffic sector. The first part of this volume—Freight & Public Transport, Demand Modelling, Human Factors & Challenges - with contributions presented by the German Road and Transportation Research Association (FGSV)—was supported by the FGSV “AA 1.8 Güterverkehr”. The World Conference on Transport Research Society (WCTRS) provides a forum for interchange of ideas among transport researchers, managers, policy makers and educators from all over the world, from a multi-modal, multi-disciplinary and multi-sectoral perspective. Every three years, the society holds the World Conference, where leading transport professionals convene to learn from one another. The WCTRS brings together a worldwide network, with more than 1150 experts interested in transport research, representing over 70 countries. It is structured into nine topic areas (A-I) covering different transport-related topics: from general to finance to freight to planning and maintenance. The Society’s Topic Areas feature a number of Special Interest Groups (SIGs) which are each dedicated to a specific academic topic from mode-specific policy and modeling, to solutions for climate change and disaster resilience. The third part of this volume Maritime & Rail Transport – with contributions presented by the World Conference on Transport Research Society (WCTRS) - SIG B3: Freight Transport Operations and Intermodality was supported by the WCTRS SIG B3. SIG B3 is part of Topic Area B “Freight Transport and Logistics”. Topic Area B includes freight carried by all modes of transport, with a strong focus on the way in which freight modes are integrated, and in particular, the importance of intermodal transport and the relationship between freight transport and logistics management. SIG B3 itself focuses on freight transport operations and connecting logistics. As such, SIG B3 covers freight transported by all modes of transport with a strong focus on how freight transport modes are integrated—in particular, the importance of intermodal transport and the relationship between freight transport and logistics management and the effects of related ICT solutions.

Contents

Supply Chains, Costs and Investments in Times of Multiple Crises . . . . . . . . . Hubertus Bardt and Michael Grömling

1

Freight and Public Transport, Demand Modelling, Human Factors and Challenges - with Contributions Presented by the German Road and Transportation Research Association (FGSV) Design of a Forecasting Method for Occupancy Rates in Local Public Transport Based on Data from Automatic Passenger Counting Systems . . . . . . Stefan Saake and Carsten Sommer

13

Effects of Cognitive Biases and Their Fuzzy Measure During Freight Transportation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eszter Sós, Adrián Horváth, and Péter Földesi

29

Exploring Knowledge Gaps Amongst Key Actors in the Transition Towards an Electrified Freight Transport System in Sweden. . . . . . . . . . . . . . Petra Stelling and Sabrina Brunner

44

First Mile Challenges for Agricultural Logistics . . . . . . . . . . . . . . . . . . . . . . Taha Karasu, Shahid Hussain, and Pekka Leviäkangas Microscopic Agent-Based Parcel Demand Model for the Simulation of CEP-Based Urban Freight Movements to and from Companies . . . . . . . . . . Lukas Barthelmes, Mehmet Emre Görgülü, Jelle Kübler, Martin Kagerbauer, and Peter Vortisch

60

75

Cargo Bikes and Freight Transport Cargo Bikes in Transport Logistics Review of User Requirements and Related Bike Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Maximilian Mowe, Annchristin Weiß, and Uwe Clausen

95

Delivery and Shipping Behaviour of Commercial Freight Demand of Logistics Service Providers – An Empirical Study for Berlin . . . . . . . . . . . . . 108 Greta Hettich, Tilman Matteis, and Carina Thaller

xx

Contents

Maritime and Rail Transport – with Contributions Presented by the World Conference on Transport Research Society (WCTRS) SIG B3: Freight Transport Operations and Intermodality Dynamic Container Routing Problem on a Rail-Based Hub-and-Spoke Network. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Ralf Elbert and Hongjun Wu Reinforcement Learning at Container Terminals: A Literature Classification . . . 147 Michaela Grafelmann, Nicole Nellen, and Carlos Jahn Electric, Hydrogen Vehicles and Chemical Logistics Simulation-Based Impact Assessment of Electric and Hydrogen Vehicles in Urban Parcel Delivery Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 Ibraheem Oluwatosin Adeniran, Abdulrahmon Ghazal, and Carina Thaller Groundbreaking Challenges of Deploying Battery-Electric Terminal Trucks in Container Terminals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Andreas Mohr, Marvin Kastner, and Carlos Jahn Requirements Catalog for Intralogistics Processes in the Production of Specialty Chemicals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 Maximilian Kiefer, Alexander Stolte, Lasse Jurgeleit, and Uwe Clausen Simulation and Optimization in Production and Logistics Exploratory Analysis of Transportation System for Modular Vehicle Concept Operated as Passenger and Goods Transport Using Continuum Approximation Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 Elija Deineko, Gunnar Knitschky, and Daniela Rischke Approximation Algorithms for p-Hub Center Problems . . . . . . . . . . . . . . . . . 241 Niklas Jost and Uwe Clausen Partitioned vs. Integrated Planning of Hinterland Networks for LCL Transportation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257 Niklas Jost, Dorothee Henke, Ivo Hedtke, Oliver Bredtmann, Joachim Weise, Christoph Buchheim, and Uwe Clausen Cooling Technologies in Cooled Supply Chains. About the Suitability and Sustainability of Dry Ice as a Cooling Medium. An Exhaustive Review. . . . . . 274 Naeem Salim Bagwan, Roel Gevaers, and Wouter Dewulf

Contents

xxi

Production, SCM and Sustainability Resilience of Urban Factories and Their Supply Chains – Identification of Indicators and Recommendations for Increasing Resilience. . . . . . . . . . . . . . . 293 Marius Dellbrügge, Sina Rudolf, Felix Kreuz, Max Juraschek, Christoph Herrmann, and Uwe Clausen The Planning of Hyperloop-Based CargoTubes Routes for Sustainable Logistic Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306 Heiko Duin, Walter Neu, Thomas Schüning, Lukas Eschment, Thomas Nobel, and Stephan Wurst Occupational Health and Safety Practices and Supplier Selection in the South African Mining and Construction Industry. . . . . . . . . . . . . . . . . . . . . . 321 Jane Dolo and Chengedzai Mafini Intralogistics and Automation Literature Review on Current Approaches to Ergonomic Order Allocation in Order Picking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337 Linda Maria Wings, Christian Fahrenholz, and Aylin Uludag Enhancing Manual Order Picking Process Through Picking Robots in a Forward-Reserve Warehouse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355 Minqi Zhang

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367

Supply Chains, Costs and Investments in Times of Multiple Crises Hubertus Bardt(B) and Michael Grömling Institut der deutschen Wirtschaft, Konrad-Adenauer-Ufer 21, 50668 Köln, Germany [email protected] http://www.iwkoeln.de

Abstract. The Russian invasion of Ukraine has significantly clouded the economic outlook in Germany and worsened the investment climate. The investment gap created in the wake of the Covid pandemic will not be closed for the time being. This not only weighs on the pace of economic activity, but at the same time creates protracted deficits in the overall economic capital stock and the associated production and productivity potential. The business surveys conducted by the German Economic Institute in summer and autumn 2022 provide an empirical basis for explaining the current weakness in investment. On the one hand, this company data shows the influence of certain factors on general investment activity, which is unaffected by the Ukraine war. Digitalisation and decarbonisation have a positive impact on the level of investment. High labour costs, production problems due to a lack of supplies and, in particular, the prevailing global economic uncertainty, on the other hand, are weighing on the investment climate. Uncertainties regarding energy supply, a lack of employees and the level of energy costs are also dominant barriers to investment. Secondly, the companies were asked whether and how, in their view, these investment determinants are influenced by the war in Ukraine. While the few positive drivers of investment from the companies’ point of view have changed only little, there has been a massive deterioration especially in the already existing barriers to investment: the sharp rise in energy costs, uncertainties regarding energy supply, global uncertainties and disruptions. Keywords: Investment · Geopolitics · Covid pandemic · Business survey

1 Continuing Investment Gap Becomes a Brake on Prosperity The Russian invasion of Ukraine has significantly clouded the economic outlook in Germany. The production disruptions resulting from the Covid pandemic due to a lack of intermediate inputs continue to weigh on economic life. The high price increases as a result of the sharp rise in raw material and energy prices additionally reduce the hoped-for economic catch-up effects. In this environment characterised by high economic uncertainties, the necessary recovery of investment in Germany is also failing to materialise. According to the IW A previous version of this paper has been published as Bardt and Grömling (2022). © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 U. Clausen and M. Dellbrügge (Eds.): ICPLT 2023, LNLO, pp. 1–10, 2023. https://doi.org/10.1007/978-3-031-28236-2_1

2

H. Bardt and M. Grömling

business survey of summer and autumn 2022, the investment expectations of companies in Germany have declined significantly (Grömling 2022a): According to the autumn survey, although 24% of the companies surveyed expect higher investments for 2023 than in 2022, more than 35% of the companies anticipate lower investments than in the previous year. This means that the investment expectations of companies in Germany have weakened significantly since autumn 2021 and the current balance of positive and negative values is once again far away from those levels that were recorded in times of good investment activity (Fig. 1).

Fig. 1. Investment expectations in Germany. Balance between positive and negative expectations in percentage points. Weighted results of the IW business survey. Spring survey (F) and summer survey (S): expectations for the current year; autumn survey (H): expectations for the coming year. No comparable results are available for spring 2020. Spring 2022: Overall result for the three survey periods. Source: German Economic Institute

In comparison with the previous year, it is important to bear in mind that investments in 2021 and 2022 were significantly lower than before the Covid pandemic. Priceadjusted gross fixed capital formation (based on the national accounts publication of November 2022) was only 1.2% higher on average in 2021 than in the previous year. This did not make up for the 2.3% slump from the first Covid year and compared to 2019 there was still an overall investment gap of just over 1% before the outbreak of the war in Ukraine. When assessing investment activity, the significantly different developments in the individual investment categories are relevant: While construction investments actually increased in the first Covid year compared to the previous year, there was a slump of 11% in equipment investments in 2020. This was by no means compensated for by the 3.5% increase in 2021. Equipment investments primarily include machinery, business equipment and commercially used vehicles.

Supply Chains, Costs and Investments in Times of Multiple Crises

3

105 100 95 90 85 80 75 70 65 I 22

I 21

I 20

I 19

I 18

I 17

I 16

I 15

I 14

I 13

I 12

I 11

I 10

I 09

I 08

I 07

I 06

I 05

I 04

I 03

I 02

I 01

I 00

60

Fig. 2. Investment cycles in Germany. Price, seasonally and working-day adjusted equipment investment; index 4th quarter 2019 = 100. Sources: Federal Statistical Office; German Economic Institute

According to the results of the IW business survey of summer and autumn 2022, the investment gap that arose in Germany in the wake of the Covid pandemic will not be closed in 2023. Figure 2 shows the size of the investment gap in equipment goods still to be closed in 2022. Compared to the situation shortly before the outbreak of the Covid pandemic alone (fourth quarter of 2019), there was still a catch-up potential of 2.5% in autumn 2022. Figure 2 also shows that investment was already declining in 2019. In contrast to previous recovery phases, production and delivery delays slowed down the normalisation of investment activity. Since February 2022, the additional production problems in the capital goods industry and the global uncertainties caused by the war have been added. When assessing the investment activity after the pandemic and the war in Ukraine, it is not only the current investment gap that matters. Rather, considerable investment shortfalls have accumulated over the past two years, which in total have severely slowed down the development of the overall economic capital stock. These total losses in gross fixed capital formation in Germany in the years 2020 to 2022 can be estimated on the basis of a model calculation (Grömling 2022c). For this purpose, a counterfactual economic development was used: If it had not been for the pandemic with all its supplyand demand-side burdens and, since February 2022 the war in Ukraine, total investment in Germany would have been 125 billion euros higher in price-adjusted terms. These investments, which have not taken place in the last three years, not only weigh on the pace of economic activity, but at the same time create protracted deficits in the overall economic capital stock and the associated production and productivity potential. A recent analysis of productivity development in Germany (Grömling 2022b) shows that strongly declining impulses from capital accumulation had already slowed down productivity progress before the pandemic. For the development of the capital stock, it ultimately depends on whether and to what extent current investments exceed the

4

H. Bardt and M. Grömling

retirement of the capital stock. Even the higher gross fixed capital formation before the pandemic was not enough to bring about a noticeable increase in capital intensity and labour productivity. This is because modern capital goods in particular - which are crucial for the upcoming structural transformations as a result of demographic change, digitalisation and decarbonisation (see Demary et al. 2021) - are characterised by a high pace of innovation and correspondingly high retirements. For these reasons alone, the necessary investment activity will be demanding (Bardt et al. 2021) - in order to strengthen growth potential and to safeguard prosperity.

2 Drivers and Barriers of Investment in Germany In order to shed more light on the background to the current weakness of investment in Germany, we draw on the results of the IW Business Survey of summer 2022, in which almost 2,300 companies participated. The IW Business Survey (see in detail Grömling 2018) has been regularly surveying East German companies since 1992 and additionally West German companies since 2002 about their current business situation and economic prospects. The survey is conducted by the amsa Institute on behalf of the German Economic Institute in spring, autumn and, since 2021, also in summer. The IW Business Survey is not a panel survey with a constant group of participants. By and large, the group of regularly participating companies is dominant and largely stable. The companies surveyed are distributed among industry (with three sub-sectors), construction and the service sector (with four groups). The banking/insurance sectors and the public sector are not included. The results are available in an unweighted version as well as in a weighted version according to establishments, employees and regions. This regular survey on business cycle indicators is supplemented by additional questions that can be used to address special business cycle issues. As part of the Summer Survey 2022, the participating companies were first asked what influence a total of 11 factors from a given list have on the general investment activity in their own company, i.e. unaffected by the war in Ukraine. This describes both investment drivers and investment barriers that influence investment projects and the business models behind them. The companies surveyed could each answer on a five-point scale: no influence; positive and negative values could be differentiated into “strongly positive/negative” and “(rather) positive/negative”. The difference of positive and negative assessments of the respective determinants provides an assessment of which factors currently support general investment activity in Germany on the one hand and what can be seen as an obstacle to investment on the other. In Fig. 3 the weighted results are presented in order to better capture the macroeconomic weight of the individual determinants.

Supply Chains, Costs and Investments in Times of Multiple Crises

5

Fig. 3. Determinants of general investment activity in Germany. Balance of positive and negative assessments of the respective factors on investment activity in percentage points. Underlying question: What influence do the following factors have on your company’s general investment activity? Weighted results of the IW Business Survey of June 2022 among 2,282 companies. Source: German Economic Institute

On the investment driver side, two factors in particular stand out. The two disruptors digitalisation and decarbonisation (see Demary et al. 2021) have a positive and the largest surplus of companies that see positive effects on the level of investment here. Digitisation is seen positively by 49% of firms and strongly positively by another 11% for their own investment activity, while only 8% see a decline in investment due to digitisation. The vast majority of the companies thus seem to be preparing for the profound trends of digitalisation by higher investments. This is particularly stressed by companies in the information and communication technology (ICT) sector, by media companies and by capital goods manufacturers. This argument has the weakest effect in the construction industry. It is striking that larger companies with over 500 employees see stronger positive effects here, while small companies with under 50 employees show a clearly positive but weaker tendency. Unsurprisingly, the planned investments are striking among companies with decidedly digital business models, whereas 21% expect very strong positive effects of digitalisation on their own investment activities. Decarbonisation as an investment driver also has a positive effect, albeit much weaker. The balance of positive and negative values here is 24% points. The difference to digitalisation is mainly due to the fact that there are significantly fewer companies (35%) that expect a positive investment effect from decarbonisation. More than half of the companies surveyed see no impact. The proportion of businesses with negative investment expectations from decarbonisation is, at 12%, only slightly higher than it is the case with digitalisation. Overall, the reduction of greenhouse gas emissions and the associated conversion of production processes will therefore be associated with additional investments. Companies producing intermediate goods and the logistics sector are

6

H. Bardt and M. Grömling

particularly noteworthy here, while only a moderate effect is assumed by the companies surveyed in the construction industry and ICT and media companies. There is also a clear size effect for this determinant. For large enterprises, the balance of positive and negative effects, at 39% points, is considerably higher than in the economy as a whole. Among small enterprises, positive and negative effects almost balance each, with a particularly large number of enterprises with neutral expectations. While the two disruptive trends of digitalisation and decarbonisation require and are expected to lead to additional investment overall, numerous current influencing factors stand in the way of dynamic investment activity. The negative influence exceeds the positive effect by more than 60% points with regard to labour costs, current production problems due to a lack of supplies and, in particular, the prevailing global economic uncertainty. Energy supply uncertainties, a skilled labour shortage and high energy costs are also dominant barriers to investment in Germany and result in a negative balance of more than 50% points. This finding also confirms the results from previous studies on the determinants of investment activity in Germany (Bardt et al. 2015). In recent years, for these reasons, there has been insufficient investment, particularly in the energyintensive sectors, to keep the capital stock there stable (Bardt 2021). A negative balance of almost 40% points or more was measured for tax burdens, reliable economic conditions and Covid risks. Infrastructure deficiencies also come at the expense of private sector investment. Measured against the mostly supply-side influencing variables, the demandside factors have only a moderate investment effect. With the demographic development, which companies are increasingly experiencing due to the lack of employees, and the uncertainties resulting from a possible deglobalisation, two of the disruptors with long-term impact are associated with declining investment expectations. In contrast, digitalisation and decarbonisation lead to expecting higher investment. Here, however, other factors can counteract an expansion of investment. For example, the already high energy costs in Germany for some time, exacerbated by the current energy price explosion and the prospect of permanently higher gas prices, contradict the identified investment needs caused by decarbonisation. Infrastructure deficiencies, insofar as they relate to digital infrastructures, hinder digitisation investments. It is to be feared that the high investments that are necessary for future production in Germany will not be sufficiently realised because the current situation and, in particular, structural investment conditions are not sufficiently supply-oriented.

3 Influence of the Ukraine War on Investments The economic consequences of the Covid pandemic continue to be felt in the form of persistent production and investment gaps. Based on weighted survey data, 56% of the companies surveyed in the IW Business Survey continue to see negative influences of the pandemic on their own investments. In the midst of this ongoing crisis situation, Russia’s invasion of Ukraine led to further uncertainties and adjustment burdens. In the first few weeks, it were in particular increased energy costs, intensified problems in supply chains and concerns about energy supply that placed a massive burden on companies in Germany (Grömling and Bardt 2022a).

Supply Chains, Costs and Investments in Times of Multiple Crises

7

As part of the supplementary question to the IW Business Survey of summer 2022, the companies were also asked whether and how, in their view, the factors relevant to investment (see Fig. 3) are influenced by the war in Ukraine. Figure 4 represents, on the basis of weighted values, the proportion of companies that assess the development of the respective factors influencing their own investments negatively or positively. Positive investment effects can result, for example, as additional investments have to be made in energy saving or cybersecurity as a result of the war. The survey results nevertheless clearly indicate that the negative effects dominate throughout.

Fig. 4. Change in investment conditions due to the war. Proportion of companies that expect the war in Ukraine to have a positive or negative impact on the respective investment factors in Germany, in per cent. Weighted results of the IW business survey of June 2022 among 2,282 companies. Remainder to 100: No influence. Source: German Economic Institute

Five investment determinants stand out in this question: For many companies, the dramatic increase in energy costs in particular represents a massive deterioration in competition with clearly negative effects on their own investment activities. On the basis of weighted values, 69% of the companies see a deterioration in this investment-relevant criterion due to the war in Ukraine. A good 7% of the companies surveyed are motivated to invest more. Around 60% of the respondents see a deterioration in the investmentrelated framework conditions with regard to the security of energy supply, the already high global economic uncertainty and the disruptions in the international supply chains. For 40%, the perceived lower reliability of the economic policy framework represents a further increase in investment obstacles. The level of labour costs in Germany, which has a negative impact on investment activity irrespective of the war in Ukraine, has become even more relevant as a result of the geopolitical problems. Especially in the manufacturing sector, Germany has a long-term labour cost disadvantage, even in comparison with advanced economies (Schröder 2021).

8

H. Bardt and M. Grömling 70

Deterioration due to the Ukraine war

Energy costs Energy supply global economic uncertainty

60 50

Supply chain disrupons

40 Reliable framework condions 30 20 10

-80

-60

-40

0 -20 0 20 General influence on investments

40

60

Fig. 5. Existing barriers to investment exacerbated by the war. Balance of positive and negative assessments of the influence of the respective factors on investment activity in Germany in percentage points as well as share of companies that expect a deterioration of the respective investment factors due to the war in Ukraine, in per cent. Weighted results of the IW business survey of June 2022 among 2,282 companies. Source: German Economic Institute

The combination of the answers to the two sub-questions (see here Fig. 3 and Fig. 4) illustrates the significance of this negative dynamic. While the few positive drivers of investment changed little as a result of the war, there have been massive deteriorations especially regarding the existing strong barriers to investment (Fig. 5): Energy costs and energy supply risks as well as global economic uncertainties and supply chain difficulties are particularly problematic in this respect. Thus, those factors continue to be particularly critical for investment activity that were already causing high burdens for companies before the war began.

4 Options for Economic Policy In terms of policy options, these four factors - energy costs, energy supply risks, global economic uncertainties, supply chain difficulties - are quite different. While the global economic uncertainties can hardly be reduced unilaterally, the supply chain problems can be defused over time if logistics chains are stabilised and - where necessary - alternative sources of supply are developed. These are primarily entrepreneurial tasks. The greatest leverage for political action to defuse this critical situation is to be seen in government cost factors and public security of energy supply. The Russian war in Ukraine is causing considerable additional cost shocks for companies - especially through the increase in the price of energy and raw materials. Very high price levels are to be expected for energy costs in the short and medium term. In the longer term, too, a changeover of gas supply to internationally tradable liquefied gas is likely to result in a higher price level than in the past, although hardly at the extreme

Supply Chains, Costs and Investments in Times of Multiple Crises

9

prices of 2022. According to the economic survey conducted by the German Economic Institute in spring 2022, more than 90% of companies expect strong and medium effects of higher energy and raw material costs on their own prices by the end of 2022 (Grömling and Bardt 2022b). The companies also fear higher producer prices in the medium term as a result of rising labour costs. For more than 80% of the companies surveyed, this applies to a strong and medium extent. The IW survey from spring 2022 shows on the one hand that around three out of five companies are able to pass on the currently higher costs to their customers - other companies or end consumers - to a high and medium extent. The remaining companies, however, are not able to pass on the higher costs to any significant extent - and accordingly they are left with the high production costs. This can also have a negative impact on economic dynamics, the development of growth potential and the transformation of the economic structure through investment losses. In terms of energy policy, the cost explosion of energy prices must be addressed insofar as it endangers the short- and long-term competitiveness of companies. A shortterm reduction of production can lead to unsustainable costs and threaten the existence of the companies concerned, partial state compensation is necessary to safeguard the companies. A permanently higher price level also calls into question the long-term competitiveness of certain energy-intensive companies. These companies are deeply integrated into industrial networks. To ensure the necessary investments, appropriate investment conditions must also be created with a view to the important investment drivers. For digitalisation, this concerns modern and efficient infrastructures, legal clarifications, training and research and development, but also a start-up-friendly environment. With regard to decarbonisation, there are also infrastructural challenges; in addition, price signals and support systems must be brought into an investment-friendly relationship. Improving investment conditions is a basis for economic prosperity, so that the current crisis does not turn into a permanent weakness.

References Bardt, H.: Industrie, Klimaschutz und Handel. Ausgleich unterschiedlicher Kosten und Preise für industriellen Klimaschutz. IW-Report, No. 41, Cologne (2021) Bardt, H., Grömling, M.: Krieg in der Ukraine verschärft bestehende Investitionsschwäche. IWReport, No. 44, Cologne (2022) Bardt, H., Grömling, M., Hüther, M.: Schwache Unternehmensinvestitionen in Deutschland? Diagnose und Therapie. Z. Wirtsch. 64(2), 224–250 (2015) Bardt, H., Hüther, M., Klös, H.-P.: Modernisierung durch Investition. IW-Report, No. 22, Cologne (2021) Demary, V., Matthes, J., Plünnecke, A., Schaefer, T. (eds.): Gleichzeitig: Wie vier Disruptionen die deutsche Wirtschaft verändern. Herausforderungen und Lösungen, IW-Studie, Cologne (2021) Grömling, M.: Methods and applications of the IW business survey. IW-Report, No. 5, Cologne (2018) Grömling, M.: Die Zuversicht der Unternehmen schwindet. IW-Konjunkturumfrage Sommer 2022. IW-Report, No. 39, Cologne (2022a) Grömling, M.: Ökonomische Verluste in Deutschland durch Pandemie und Krieg. IW-Kurzbericht, No. 91, Cologne (2022b)

10

H. Bardt and M. Grömling

Grömling, M.: Kapitalarmes Produktivitätswachstum in Deutschland – eine schwere Ausgangslage. Zeitschrift für das gesamte Kreditwesen 75(9), 32–37 (2022c) Grömling, M., Bardt, H.: Betriebliche Belastungen durch Ukrainekrieg. Wirtschaftsdienst 102(4), 283–287 (2022a) Grömling, M., Bardt, H.: Bleiben Unternehmen auf den hohen Kosten sitzen? Ergebnisse der IW-Konjunkturumfrage zur Preisentwicklung in Deutschland. IW-Report, No. 36, Cologne (2022b) Schröder, C.: Lohnstückkosten im internationalen Vergleich. Starke Belastung der deutschen Industrie seit 2018. IW-Trends 48(2), 85–104 (2021)

Freight and Public Transport, Demand Modelling, Human Factors and Challenges - with Contributions Presented by the German Road and Transportation Research Association (FGSV)

Design of a Forecasting Method for Occupancy Rates in Local Public Transport Based on Data from Automatic Passenger Counting Systems Stefan Saake(B)

and Carsten Sommer

Transportation Planning and Traffic Systems, University of Kassel, 34125 Kassel, Germany [email protected]

Abstract. The predominant focus on individual motorized transport is neither sustainable nor socially just. One goal of a more sustainable design of the transport sector is to encourage people to use public transport. One barrier for passengers to use public transport are heavily occupied vehicles and the uncertainty about whether an empty seat is available on the desired connection. In this paper, a model is presented that is able to forecast the occupancy of vehicles in public transport. This information can be provided to passengers to increase customer satisfaction. Different sub models are presented, which differ according to their forecast horizon and the data sources used. The most important data source is data from automatic passenger counting systems collected in vehicles in the region of Northern Hesse during the project period of the research project U-hoch-3. After linking further data sources such as weather and timetable data, stratification characteristics are developed based on which occupancy states can be derived for future journeys. By linking the data with real-time data, the forecast quality can be significantly improved. It is shown which influences the Covid-19 pandemic and the introduction of the 9 e ticket in Germany had on the model development and by which functions these changes in demand can be correctly represented by the model. The results presented in this paper show that it is possible to reliably predict occupancy rates for vehicles in public transport. Keywords: Public transport · Occupancy · Forecasting · Data analysis · Automatic passenger counting

1 Introduction In order to achieve the climate targets, set by the German government for 2030, a drastic reduction of greenhouse gas emissions is indispensable (target for the transport sector: 85 million metric tons of CO2 equivalents in 2030) [1]. While reductions have already been achieved in some sectors, there is no trend reversal visible yet in the transport sector [2]. Besides meeting climate targets, the still predominant focus on individual transport creates a multitude of further problems and challenges, especially in urban © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 U. Clausen and M. Dellbrügge (Eds.): ICPLT 2023, LNLO, pp. 13–28, 2023. https://doi.org/10.1007/978-3-031-28236-2_2

14

S. Saake and C. Sommer

areas: Transport systems regularly reach their capacity limits, the limited availability of space leads to distribution conflicts between users of different means of transport, traffic noise, air pollution and accidents cause considerable damage to health. Moreover, in some areas, social participation requires access to a private car, which excludes certain groups (e.g., women, children, the elderly, marginalized groups, people with disabilities). Considering all these problems, it becomes clear that the current transport system is neither sustainable nor socially just. A necessary component of a sustainable transport sector is public transport. Public transport can transport large numbers of people efficiently and can save significant amounts of CO2 [3]. Push and pull measures are available to encourage the use of public transport. Providing passengers with information about the expected occupancy of vehicles is an example for a digital pull measure that increases comfort and thus customer satisfaction. Especially against the background of the Covid 19 pandemic and the experiences with the 9 e ticket in Germany, reliable occupancy information is gaining in importance. This article presents a forecasting method for occupancy levels in vehicles of local public transport. 1.1 Project Context and Data Collection Study Area The model presented in this paper is being developed in the context of the research project U-hoch-3 funded by the German Federal Ministry of Education and Research. Within the framework of this project, an assistance system for passengers of public transport is developed, that provides three services. Passenger occupancy forecasting is one of these three developed services. Further information about the research project and the other project components can be found in a separate publication [4]. The study and pilot area of the project and thus also of the forecasting models is the region Northern Hesse with the city of Kassel (approx. 200,000 inhabitants). The public transport in this area is organized by the Nordhessischer Verkehrsverbund (NVV), the most important transport company is the Kasseler Verkehrsgesellschaft (KVG), which is responsible for the local public transport in the urban area of Kassel. The NVV thus generally covers regional public transport and public transport in smaller towns and communities, while the KVG serves the urban area of Kassel with buses and trams. The interconnected area covers an area of approximately 7,000 km2 and includes 5,608 stop positions. The transport performance amounts to approx. 480 million passenger kilometres per year generated by around 80 million passengers [5]. Data Collection The vehicle fleet ordered by the NVV for regional public transport is almost completely equipped with automatic passenger counting (apc) systems. These systems are installed above the vehicle doors and count boarding and alighting passengers at each stop by means of infrared rays (see Figs. 1 and 2). The so-called time-of-flight method measures the travel time of the infrared rays until they hit an object. This information can be used to generate a three-dimensional image, which then can be assigned to an object type in the next processing step.

Design of a Forecasting Method for Occupancy Rates

Fig. 1. Apc system in a tram in Kassel

15

Fig. 2. Apc system close up

The apc data is automatically read out after each day of operation and stored in a database. Compensation calculations are performed, for example, to minimize the effects of measurement errors and sensor failures. The apc data is linked with timetable data (planned timetable and actual timetable), position data, vehicle data and weather data. A comprehensive data set is thus available for the development of the forecast models, which will be continuously updated starting in January 2019. As part of the U-hoch-3 research project, some of KVG’s trams are being equipped with a new generation of apc systems. Thanks to a significantly higher resolution, it is possible to perform a finer object recognition. This makes it possible to distinguish between wheelchairs, bicycles, and baby carriages, among other things. In the context of the presented model development, it is also relevant that the counting data can be transmitted in real time. This is a requirement for the short-term forecasting model.

2 State of Research The provision of information on the occupancy of a vehicle has a major influence on customer satisfaction and thus also on usage behaviour [6]. With an increasing number of accessible data sources, larger storage capacities, higher transmission rates and available computer-aided algorithmic methods, a broad research field has consequently developed. The present research can be described by the field of application (e.g. bus or train) and by the data and methods used. Vandewiele et al. [7] use crowd-generated data to predict the occupancy of trains in three classes. App users were asked to provide trip origins, destinations, and perceived occupancy status. An XGBoost algorithm selects the most important features from the available input features. The train departure time is identified as the most influential feature. Pasini et al. [8] also investigate forecasting models to predict train occupancy. Different data sources are used: Besides calendar data (day, day type, vacation type, minute in the day), routing data of the train are used. In addition, there are input features that are available at short notice: Delay of the train and occupancy on the last 6 sections before the section to be forecasted. The applicability of 5 different methods is tested. In the simplest model, the occupancy of the last section is given as a forecast for the next

16

S. Saake and C. Sommer

section. The second approach gives the average occupancy of the same day type and time shift. The third approach is a gradient boosting regression model, and the fourth is a random forest model. The fifth approach is a neural network capable of representing temporal dependencies (LSTM). Wang et al. [9] also review different forecasting methods (LightGBM, XGBoost, Random-Forest, ARIMA). Calendar week, day of the week, and average temperature are identified as the most important input features for predicting the occupancy of a train. Heydenrijk-Ottens et al. [10] investigate the prediction of bus and tram occupancy in transportation network of The Hague. Three data sources are used: GTFS data (data on geographic structure, stops, routes, and schedules), vehicle location data (automatic Vehicle Location; AVL), and automated fare collection (AFC) data. Four different classification methods are applied (Random-Forest, Gradient Boosting, Multi-layer Perceptron, and K-Nearest Neighbors), with Random-Forest classification yielding the best results. Arabghalizi and Labrinidis [11] test four different approaches for forecasting bus occupancy in the Pittsburgh region. They consider forecasting as a regression task (Negative Binomial Regression Model) in one approach, and as a classification task (Logistic Regression, Artificial Neural Network, Random Forest) in the other three approaches. Calendar data, weather data, historical count data, and timetable data are used as input features. Vial and Gazeau [12] use a random forest algorithm to calculate forecasts. In their method for forecasting bus occupancy, they rely primarily on apc data. The following input features are used in their model: Temporal information (hour, day, month, bridge day, weekend), historical count data of the stop being forecast, historical count data of other stops, and delays between vehicles. Jenelius [13] also uses apc data in a forecasting model. He additionally considers different forecast horizons. In addition to historical count data, real-time position data and real-time count data are used. He notes that historical count data tends to underestimate high occupancy rates and overestimate low occupancy rates. Real-time count data greatly improves forecasts, the more recent the data, the greater the improvements. In Germany, there are several forecasting tools used in practice [14–16] and different tools and models that are being developed in research projects [17]. Depending on the context different input data and models are used.

3 Model Development 3.1 Introduction of the Model Goal The general purpose of the model development presented in this paper is to forecast the occupancy of public transport vehicles in the study area. The question is to be answered how many people are in a vehicle on a certain section at a certain time in the future. This absolute number of passengers is then coded to discrete occupancy categories and can be communicated to passengers via different communication channels.

Design of a Forecasting Method for Occupancy Rates

17

Fundamental Functioning of the Model The forecasts are generally calculated based on historical data. The model distinguishes between two forecast horizons: the long-term forecast and the short-term forecast. The long-term forecast can calculate forecasts for any trip in the future. The short-term forecast, on the other hand, is used if the trip for which forecasts are to be calculated has already begun. Depending on the forecast horizon, different data sources are available for the calculation. Table 1 gives an overview of the two forecast models, the forecast horizons, and the data sources. Table 1. Forecast horizons and data sources Forecast model

Forecast horizon

Data sources

Long-term forecast

Weeks to days until start of the trip

Historical apc data Timetable data Weather data

Short-term forecast

From start of trip

Historical apc data Timetable data Weather data Real-time apc data

The most important data source for both forecast models is apc data. Information on the (historical) structure of passenger demand is available by linking the apc data with timetable data (planned timetables and actual timetables) of the vehicles used. In addition, historical and current weather data are linked to the data basis for the forecast model. In the case of short-term forecasting, real-time data is also available, which is transmitted from the vehicle in transit. 3.2 Algorithm The following chapter explains how the two forecast models work. Since the algorithm of the short-term forecast is based on that of the long-term forecast, the model of the long-term forecast is explained first. Long-Term Forecast The long-term forecast model can calculate forecasts for any trip in the future. A trip is broken down to the individual edges. An edge is a section between two stopping points of a trip (see Fig. 3). For the forecast calculation, the previously described data set of historical data is divided into homogenous strata. The aim of the stratification is to find data points in the past that have similar characteristics to the situation to be forecasted and whose mean value of the passenger occupancy is initially taken as the forecast value.

18

S. Saake and C. Sommer

Fig. 3. Visualization of edges. All model predictions refer to the passenger load on an edge between two stops of the network.

Variance analyses were performed to determine these stratification characteristics. The initial consideration is that those variables should be used as stratification characteristics that define homogeneous strata in relation to the mean values of the occupancy figures. The following variables were examined (Table 2): Table 2. Determination of the stratification characteristics Variable

Description

Value range

Significance level

Effect strength

Edge

Section between two stops on which the count was conducted

ID oft the edge