Trip planning for a mobility-as-a-service system: Integrating metros and shared autonomous vehicles

Abstract

Mobility as a service (MaaS) integrates various transport modes into an on-demand and real-time platform, providing door-to-door service, and has received extensive attention. For MaaS, personalized trip planning is important but intractable. In this paper, we present a two-phase decision-support optimization framework for the problem of a MaaS system incorporating metros and shared autonomous vehicles (SAVs). First, a mixed integer programming model is proposed to optimize the routes of heterogeneous travelers considering five transport mode combinations, in which SAVs are regarded as not only a first- and last-mile connector to the metro but also a competitor. Next, the scheduling of SAVs and departure time of each traveler is determined with the purpose of minimizing the SAV operation cost. To apply the proposed framework to scenarios with real-time requests, we adopt the rolling horizon solution method, which includes four sub-modules. The method is evaluated on the Sioux Falls network, and the experimental results show that travelers become more sensitive to the mode choice as the additional time of the metro increases. In addition, the connectivity of the metro network has a considerable influence on the relationship between the metro and SAVs. The methodology can be useful for the trip planning of other transport mode combinations.