Optimised flexible transport mode without advanced reservations by a DES approach

Decentralized management of an autonomous demand responsive transport system (DRTS) is studied. Clients appear in the network in a continuous Poisson stream, without any prior reservation and aiming at an immediate departure search for an adequate vehicle. When no such vehicle is available, clients decide whether to abandon the system or wait, during a desired period, for another vehicle to come. Meanwhile, vehicles travel along the network, transporting passengers while looking for potential new customers. Dynamically optimised trips are constructed adapted to the current vehicle itinerary and client preferences. Station nodes and durations are decided for idle vehicles. Client-vehicle interaction forms an asynchronous dynamical spatio-temporal system for which an event-driven approach is adopted. Developped strategies are evaluated by a made-to-measure discrete event decision tool. Involving Optimisation and Operation Research methods the best suited configuration for any demand level and system dimensioning is provided.