Universität zu Köln

Ahadi, Ramin ORCID: 0000-0002-8447-5008 (2025). Paving the Path toward Smart Sustainable Mobility; Data-Driven Operations Management for Emerging Mobility Systems: Sharing, Automation, and Electrification. PhD thesis, Universität zu Köln.

PDF PhD_Thesis_R_Ahadi.pdf Download (9MB)

Abstract

Despite significant advances in mobility systems, they continue to perform inefficiently, contributing to a range of global challenges. To address these issues, many researchers predict that the future of transportation will be characterized by shared, autonomous, and electric vehicles (SAEVs). However, this transformation poses several managerial and societal challenges that must be addressed to fully realize its potential benefits. This thesis highlights the potential of research at the intersection of information systems and operations management to improve the performance of next-generation mobility systems (Chapter 1). By integrating data-driven decision models four research projects underscore this potential. Chapter 2 studies the cooperative charging management of SAEVs to maximize profits and service quality while accommodating uncertain demand, limited charging infrastructure, and time-varying electricity prices. A distributed approach using cooperative multi-agent reinforcement learning is proposed that outperforms centralized static charging strategies and provides insights to improve the performance of SAEVs. Chapter 3 explores the impact of emerging technologies on the physical world and their interaction with user adoption behavior. Specifically, it evaluates a hybrid system that combines autonomous and human-operated ride-hailing services. The study first identifies mobility user preferences toward autonomous services, and using an agent-based model predict and analyze the future of such hybrid service platforms, and evaluate potential changes. The agent-based model enables analysis of the end-to-end impact of key factors, such as trust in the technology. Chapter 4 proposes a method for leveraging data-driven digital twin frameworks to design large-scale charging hubs. Such problem classes are difficult to solve with traditional mathematical programming optimization. This study shows how high-fidelity, data-driven simulation environments coupled with reinforcement learning can achieve arbitrary scalability and high modeling flexibility. The benchmark experiments show that the proposed model designs result in superior cost and service-level performance under real-world operating conditions. Chapter 5 supports the operational management of EVCHs through dynamic pricing. Drawing on cutting-edge deep reinforcement learning algorithms, a model-free solution is provided to find optimal pricing policies. The proposed pricing policy is a time-dependent function of the service rate, called dynamic capacity-based pricing. Benchmark analysis shows that the proposed model not only ensures high profits for EVCHs, but also successfully reshapes aggregated demand as desired, even in environments with high variability in supply and demand.

Item Type:	Thesis (PhD thesis)
Creators:	Creators Email ORCID ORCID Put Code Ahadi, Ramin ramin.ahadi@uni-koeln.de orcid.org/0000-0002-8447-5008 UNSPECIFIED
URN:	urn:nbn:de:hbz:38-784058
Date:	2025
Language:	English
Faculty:	Faculty of Management, Economy and Social Sciences
Divisions:	Faculty of Management, Economics and Social Sciences > Business Administration > Information Systems > Professorship for Informations Systems and Operations Research
Subjects:	Economics Technology (Applied sciences)
Uncontrolled Keywords:	Keywords Language Smart Mobility English Sustainability English Shared Autonomous Electric Vehicles English Charging Management English Agent-based Simulation English Deep Reinforcement Learning English Optimization English
Date of oral exam:	3 June 2025
Referee:	Name Academic Title Ketter, Wolfgang Prof. Dr. Valogianni, Konstantina Prof. Dr.
Refereed:	Yes
URI:	http://kups.ub.uni-koeln.de/id/eprint/78405