Electric Autonomous Mobility-on-Demand: Jointly Optimal Vehicle Design and Fleet Operation
Fabio Paparella, Théo Hofman, Mauro Salazar
Abstract
The advent of autonomous driving and electrification is enabling the deployment of Electric Autonomous Mobility-on-Demand (E-AMoD) systems, whereby electric autonomous vehicles provide on-demand mobility. Crucially, the design of the individual vehicles and the fleet, and the operation of the system are strongly coupled. Hence, to maximize the system-level performance, they must be optimized in a joint fashion. To this end, this paper presents a framework to jointly optimize the fleet design in terms of battery capacity and number of vehicles, and the operational strategies of the E-AMoD system, with the aim of maximizing the operator’s total profit. Specifically, we first formulate this joint optimization problem using directed acyclic graphs as a mixed integer linear program, which can be solved using commercial solvers with optimality guarantees. Second, to solve large instances of the problem, we propose a solution algorithm that solves for randomly sampled sub-problems, providing a more conservative solution of the full problem, and devise a heuristic approach to tackle larger individual sub-problem instances. Finally, we showcase our framework on a real-world case study in Manhattan, where we demonstrate the interdependence between the number of vehicles, their battery size, and operational and fixed costs. Our results indicate that to maximize a mobility operator’s profit, a fleet of small and light vehicles with battery capacity of 20kWh only can strike the best trade-off in terms of battery degradation, fixed costs and operational efficiency.