Car sharing is expected to reduce traffic congestion and pollution in cities while at the same time improving accessibility to public transport. However, the most popular form of car sharing, one-way car sharing, still suffers from the vehicle unbalance problem. Innovative solutions to this issue rely on custom vehicles with stackable capabilities: customers or operators can drive a train of vehicles if necessary, thus efficiently bringing several cars from an area with few requests to an area with many requests. However, how to model a car sharing system with stackable vehicles is an open problem in the related literature. In this paper, we propose a queueing theoretical model to fill this gap, and we use this model to derive an upper-bound on user-based relocation capabilities. We also validate, for the first time in the related literature, legacy queueing theoretical models against a trace of real car sharing data. Finally, we present preliminary results about the impact, on car availability, of simple user-based relocation heuristics with stackable vehicles. Our results indicate that user-based relocation schemes that exploit vehicle stackability can significantly improve car availability at stations.
Relocation in Car Sharing Systems with Shared Stackable Vehicles: Modelling Challenges and Outlook
Chiara Boldrini;Riccardo Incaini;Raffaele Bruno
2017
Abstract
Car sharing is expected to reduce traffic congestion and pollution in cities while at the same time improving accessibility to public transport. However, the most popular form of car sharing, one-way car sharing, still suffers from the vehicle unbalance problem. Innovative solutions to this issue rely on custom vehicles with stackable capabilities: customers or operators can drive a train of vehicles if necessary, thus efficiently bringing several cars from an area with few requests to an area with many requests. However, how to model a car sharing system with stackable vehicles is an open problem in the related literature. In this paper, we propose a queueing theoretical model to fill this gap, and we use this model to derive an upper-bound on user-based relocation capabilities. We also validate, for the first time in the related literature, legacy queueing theoretical models against a trace of real car sharing data. Finally, we present preliminary results about the impact, on car availability, of simple user-based relocation heuristics with stackable vehicles. Our results indicate that user-based relocation schemes that exploit vehicle stackability can significantly improve car availability at stations.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.