Heterogeneous impacts of the built environment on survival of customized bus services across the downtown and suburbs

Customized bus services (CBS) lines have been increasingly added in cities to better serve diversified travel needs. The notable variations in the built environment across the downtown and suburbs could result in wide differences in local customers' behaviors and preferences. The allocation of CBS stops thus needs to be considered separately in urban centers and peripheries to ensure the best performance, but such spatial heterogeneity is not specifically studied and justified in the literature. This study takes empirical data containing continuous CBS stops in Shanghai, China, as an example. Using a latent class binomial logit model, it investigates the effects of bus characteristics, geographic, and built environment elements on the survival of CBS stops in downtown and suburbs. The findings demonstrate that the performance of CBS systems is significantly influenced by factors including road density, population density, POI, sidewalk width, greenery area, ticket fare, and the type of CBS line (commute or general) across urban centers and peripheries. Specifically, two latent classes of CBS stops are found to have different effects depending on the stop's location or type of the linked areas, i.e. downtown or suburbs, on a CBS line.