Locating and pricing park-and-ride service in the era of autonomous vehicles

This paper studies the optimal location and pricing scheme of Park-and-Ride (P&R) services specifically designed for Autonomous Vehicles (AVs). The unique features of self-cruising and autonomous parking allow AV users to drive directly to the transit stations to access transit service while letting AVs self-cruise and park at P&R stations. This will cause AV users to make different choices regarding P&R stations compared to those driving traditional human-driven vehicles (HVs). Consequently, the layout of P&R stations and service charges designed for AVs may deviate significantly from the existing P&R service design for HVs. Standing from the perspective of P&R service operators, we formulate a bilevel model that captures the intricate interplay between service design and travelers’ choices, which aims to maximize the derived profit for operating P&R services by optimizing the location and pricing scheme of P&R stations. To solve the proposed bilevel programming effectively, we present two exact solution approaches, i.e., the mixed-integer linear programming reformulation approach and value-function-based exact solution approach. Numerical experiments are conducted to evaluate the proposed model and solution methods. Based on the results, we find that the P&R service designs for HVs and AVs are considerably different, and the P&R service will be more advantageous in the era of AVs. Through sensitivity analysis, we analyze the impacts of multiple parameters on the model solutions. The results of this study will provide guidance and insights for the deployment of P&R service in the future mobility system with AVs.