Demand-Adaptive Route Planning and Scheduling for Urban Hub-based High-Capacity Mobility-on-Demand Services
In this study, we propose a three-stage framework for the planning and scheduling of high-capacity mobility-on-demand services (e.g., micro transit and flexible transit) at urban activity hubs. The proposed framework consists of (1) the route generation step to and from the activity hub with connectivity to existing transit systems, and (2) the robust route scheduling step which determines the vehicle assignment and route headway under demand uncertainty. Efficient exact and heuristic algorithms are developed for identifying the minimum number of routes that maximize passenger coverage, and a matching scheme is proposed to combine routes to and from the hub into roundtrips optimally. With the generated routes, the robust route scheduling problem is formulated as a two-stage robust optimization problem. Model reformulations are introduced to solve the robust optimization problem into the global optimum. In this regard, the proposed framework presents both algorithmic and analytic solutions for developing the hub-based transit services in response to the varying passenger demand over a short-time period. To validate the effectiveness of the proposed framework, comprehensive numerical experiments are conducted for planning the HHMoD services at the JFK airport in New York City (NYC). The results show the superior performance of the proposed route generation algorithm to maximize the citywide coverage more efficiently. The results also demonstrate the cost-effectiveness of the robust route schedules under normal demand conditions and against worst-case-oriented realizations of passenger demand.