Modeling and managing the shared and automated mobility services

Abstract

The sharing economy and autonomous vehicles (AVs) have emerged as promising solutions for addressing transportation issues due to their cost-saving, eco-friendly, and sustainable characteristics. Shared and automated transportation services are believed to enhance system efficiency and provide benefits that contribute to more efficient, sustainable, and convenient transportation options in the near future.

Firstly, this thesis aims to investigate the operation strategies and potential impacts of introducing the sharing economy into the current transportation system. Specifically, this dissertation examines two typical shared services, namely ride-pooling service and crowdshipping service. By analyzing the operation strategies and considering the reactions of involved stakeholders, the results demonstrate that properly implemented shared transportation services can benefit stakeholders and enhance operational efficiency. Furthermore, the thesis delves into the different sharing formats/paradigm that can be applied in various market scenarios, considering different administrative objectives. In particular, the thesis explores the concept of differentiated pooling sizes in ride-pooling services and varying degrees of asset utilization (light, medium, and heavy formats) in crowdshipping systems. The results show that by effectively utilizing spare resources, such as vacant car positions and available delivery capacities of travelers, it is possible to reduce deadweight loss and provide additional benefits to the transportation system.

In addition, the advent of AVs presents new prospects and potentials for advancing the implementation of the sharing economy. To further investigate the effects of shared and automated services in the near future, this dissertation examines the effects of introducing a vehicle-sharing scheme and a shared autonomous vehicle (SAV) ride service. Within the proposed vehicle-sharing scheme, we formulate the choice behaviors of travelers of either parking or sharing their vehicles under reselling and commissioning formats. The results analytically demonstrate that a vehicle-sharing scheme can lead to a Pareto-improving outcome for the stakeholders involved, and the introduction the AVs might bring additional benefits to the existing system. Furthermore, this thesis investigates the potential operation paradigm of the SAV ride service. We consider that travelers using private autonomous vehicles (PAVs) must make decisions regarding both the route they take and where they park. The SAV ride service enables multiple travelers with a common origin and destination (OD) pair to share the same vehicle, while the routing of SAVs is determined by the service operator. We demonstrate the uniqueness/non-uniqueness properties of the joint network equilibrium, and further examine the SAV service operator's optimal operation decisions subject to the network equilibrium.

The insights derived from this thesis have the potential to generate modeling and optimization tools, managerial insights, and policy implications. These findings can support decision-making processes related to shared and automated services and can be applied to real-world cases for practical assessment and implementation.