Autonomous and Modular Urban Mobility and Long-haul Electric Truck

An emerging Autonomous Modular Vehicle Technology (AMVT) enables the vehicle to adapt its size in response to demand variations. This thesis delves into the potential and viability of incorporating AMVT into urban mobility sectors, specifically public transit and on-demand ridesharing, while also examining the feasibility of transitioning long-haul trucks to advanced battery technologies. This thesis is comprehensively organized into five chapters, neatly divided into two main segments: Part I and Part II. The opening Chapter 1 lays out the foundational research background. Part I centers on urban mobility powered by AMVT, including public transit in Chapter 2 and on-demand ridesharing in Chapter 3. Chapter 2 unveils a proof-of-concept exploration of the Autonomous Modular-based Public Transit (AMPT) system by formulating stylized design models on a grid network considering both homogeneous and heterogeneous demand distributions and comparing it with the traditional fixed-route, fixed-capacity transit service in terms of total cost (both agency’s cost, and passenger time cost). Our numerical findings suggest that an appropriately designed AMPT could lead to cost savings, especially in scenarios of low demand. Venturing into the domain of on-demand ridesharing with AMVT, Chapter 3 puts forth a framework designed to allocate shared-ride requests to vehicles to minimize the total energy use. The results suggest that ridesharing service noticeably reduces energy consumption from single-ride schemes (up to about a third). On the other hand, the en-route pod joining/disjoining and the consolidation of passengers do not appear to bring much energy saving to the system and, on the contrary, may complicate real-world operation. Transitioning to Part II, the focus shifts to the electrification of long-haul freight. Chapter 4 aims to find a charging and siting strategy to determine when and where to charge along the route by synchronizing drivers’ work hours with the charging activities of the electric long-haul truck fleets in the corridor scale. The hypothesis-driven study demonstrates that building an electric vehicle (EV) ready corridor at selected rest areas and truck stops is feasible with careful schedule coordination and charging facility siting. Chapter 5 further investigates the problem on the network level. This chapter presents a bi-level optimization model for the placement and capacity of electric truck (ET) charging stations, considering temperature impacts on battery efficiency. The lower-level handles user equilibrium in travel and queueing times using a convex combinations method embedded with a multi-criterion label-correcting algorithm for finding network flows. In contrast, using a greedy heuristic algorithm, the upper level focuses on optimal site selection and station sizing within budget constraints. Numerical experiments validate the model’s effectiveness in enhancing network connectivity and reducing costs. The research fills a gap in the existing literature by integrating climate variability into infrastructure planning and provides strategic insights for policymakers.