A simulation-optimization model for analyzing a demand responsive transit system for last-mile transportation: A case study in São Paulo, Brazil

Ride-hailing services, in particular Demand Responsive Transit (DRT) systems, have emerged in many cities worldwide as a potential solution to bridge the public transport supply gap. In such services, passengers whose itineraries and times are coincident to a certain degree can share vehicles that follow flexible routes to accommodate distinct boarding and alighting locations with minor disturbances to their convenience. In this context, this paper aims to evaluate the potential of a DRT as an alternative for a feeder to a trunk transport system, such as BRT (Bus Rapid Transit), or a metro or rail system as well. This is accomplished by a simulation–optimization model that allows determining, in very short time, the optimized way to serve each new transport request that arises dynamically over time. The model was applied to a first/last mile transportation system in an area adjacent to a subway station in the city of Sao Paulo, Brazil. We analyze and measure the impacts of different operational characteristics of supply and demand, measured in terms of service level to riders, number of vehicles, average occupation and total kilometer traveled per vehicle. The results evidence the potential of these DRT services as an alternative to complement the existing bus lines, as well as to attract users of individual transport, as they can offer adequate service level at a competitive cost.