Adaptive memory programming for a class of demand responsive transit systems

We discuss Demand Adaptive Systems, a class of hybrid transportation systems that integrate traditional bus transportation with on demand service. DAS lines serve a set of compulsory stops, each within a given time window. Optional stops are activated on demand and optimally scheduled in order to respect the time window constraints. We present an efficient meta-heuristic algorithm which exploits the mathematical structure of the problem. A lower bound based on Lagrangean Relaxation is provided. Computational results on realistic data confirm the efficacy of the approach.