Leveraging ride-hailing services for social good: Fleet optimal routing and system optimal pricing

Abstract

With the penetration of ride-hailing services, the impacts of transportation network companies (TNC) on network performance grow. TNC comes with a ubiquitous sensing and pricing system that may be leveraged by public agencies to improve transportation system performance. This study first formulates and solves a mixed equilibrium (ME) of personal driving vehicles and ride-hailing vehicles, where TNCs centrally assign routes to ride-hailing vehicles to achieve fleet-wide optimum. We propose a novel fleet behavior desired by TNCs, named fleet-optimal behavior with service constraint (FOSC), which provides a good compromise between total fleet cost minimization and fairness among riders. However, we show that the system state of FOSC can be far from the system optimum state. To this end, we propose a novel Optimal Ride-hailing Pricing (ORHP) scheme for public agencies as an efficient manner to intervene ride-hailing online platforms. The essential idea of ORHP is to regulate and subsidize TNCs, in exchange for guaranteed network performance improvement. Under ORHP, public agencies set the value of a subsidy for each link for any TNC rider using this link. The subsidies are provided to TNCs, not directly to riders. TNC receives subsidies, and determines the best way to provide compensations for each rider who deviates from his/her ”shortest” route, determined by FOSC. TNC’s ultimate goal is to reduce their total cost, including total fleet vehicle travel time and total compensations, subtracted by the total subsidy received from public agencies. The ORHP is likely less controversial than other pricing schemes, since it calls for voluntary participation of travelers who are provided with multiple route options. Because it is built into the TNCs’ fare system, it is cost effective to implement and hard to game. This would be a win-win: a win for public agencies to cost-effectively improve system performance leveraging TNC’s platform without building physical infrastructure or services; and a win for TNC to profit from subsidies and improve service quality. ORHP is formulated as a bi-level optimization problem, solved with a sensitivity analysis based algorithm and tested on two networks implying the ORHP scheme can be effective: a small total subsidy provided to TNC can lead to significant improvement in system performance.