Dynamic Car-Passenger Matching based on Tabu Search using Global Optimization with Time Windows

On-Demand Mobility is an increasingly popular concept especially in urban areas, which has the potential to reduce congestion and the space needed by privately owned vehicles due to shared car fleets. To avoid a decline of flexibility and convenience for the customers and to minimize the costs for the service provider, a fleet management algorithm matches the requests and the vehicles in order to quickly find a reliable and time efficient solution for the whole system. The focus of this work is to introduce a new approach to find solutions periodically using a Tabu Search metaheuristic, called Global Optimization with Time Windows. It is shown that this method allows significantly better solutions compared to those found by the Nearest Neighbor Policy, without losing the ability to quickly inform customers about their pick-up time.