Implementation of a spontaneous matching algorithm for on-demand shuttle systems in microsimulation

Abstract

On-demand shuttle systems have gained popularity in the last years, recently more and more applications have started worldwide. The core of such an on-demand shuttle system is an intelligent matching algorithm, where vehicles and passenger requests are matched to each other in an optimized way. An appropriate platform to model and simulate matching algorithms is indeed a microscopic traffic simulation, as well-known for other modes of transportation. However, on-demand shuttle systems were not studied sufficiently in microsimulation environments up to now, most probably due to special characteristics of dynamic routing. This paper aims to close this gap and model a platform for on-demand shuttle systems in a microscopic traffic simulation, allowing a holistic microsimulation considering all transport modes. A spontaneous matching algorithm is introduced and its implementation in the microsimulation is elaborated. The results of a case study in the University of the Bundeswehr Munich campus are presented, showing the impacts of operation type and fleet size on essential metrics. The potential of such a platform as well as further research insights are discussed at the end of the paper.