Revealing the extent of route-level accessibility under online bus-hailing routing strategies

The online bus-hailing service integrates the advantages of both ride-hailing platforms and conventional public transit, offering a flexible solution to meet residents' individualized travel demands that are often underserved by traditional public transportation systems. This study develops a bi-level optimization model for online bus-hailing service routing, with the dual objectives of maximizing accessibility and minimizing travel costs. The model incorporates key operational constraints, including time windows, vehicle capacity utilization, and maximum route distance. An extended greedy randomized adaptive search procedure (GRASP) is proposed to solve a bus sharing problem with time windows. Then, the proposed model and algorithm are empirically validated using case data from Beijing. The accessibility evaluation was compared between two different scenarios: weekday and non-weekday. Results demonstrate that the spatial distribution of route-level accessibility for online bus-hailing is fairer on weekdays compared to non-weekdays. These findings are important for the sustainable development of public transit systems.