Autonomous Vehicles in Ride-Hailing and the Threat of Spatial Inequalities

Abstract

Problem definition: This paper studies how the potential introduction of autonomous vehicles (AVs) in a fleet of human drivers (HVs) may impact the quality of service and the equality of access to transportation on ride-hailing platforms. Methodology: We formulate a game-theoretical queueing model in which a platform aims to maximize its profit while HVs make strategic joining decisions. Then, using New York City data, we build a highly detailed simulation of the operations of a ride-hailing platform to confirm that our theoretical results from the queueing model still hold and demonstrate additional effects on service levels. Results: Our results demonstrate that the introduction of AVs may deteriorate the service level. As the platform incorporates AVs, it may prioritize them, which affects the earnings of HVs and drives them out of the market. We then reveal that the reduction of service level is not homogeneous across areas in a city: while the more profitable high-demand areas, such as downtown areas, may see a high concentration of vehicles and reasonable service levels, remote locations may suffer from a drop in service level. Managerial implications: This paper provides insights on how decision-makers should maintain service quality when introducing AVs in a ride-hailing market with heterogeneous demand areas. A profit-maximizing platform may be willing to increase the utilization of AVs at the price of sacrificing service levels. We should carefully balance profitability and service quality, and incentivize human drivers to remain in the market, especially in low-demand areas.

Keywords: Autonomous vehicle, Ride-hailing, Queueing model, Simulation, Service level