A Modular Soft Computing based Method for Vehicles Repositioning in Bike-sharing Systems

A crucial issue in bike-sharing systems (BSS) is the unbalanced distribution in space and time of the bikes among the stations. Literature shows several methods, to solve the vehicle reallocation problem and most of them are based on rigid control thresholds and refer to car-sharing systems. In this paper a more flexible fuzzy decision support system for redistribution process in BSS is presented. The aim of the proposed method is to minimize the redistribution costs for bike-sharing companies, determining the optimal bikes repositioning flows, distribution patterns and time intervals between relocation operations, with the objective of a high level for users satisfaction. The proposed method allows to define the best bikes repositioning jointly to the best route for the carrier vehicles. The optimization method has been applied to a simulated BSS that can be considered as a module of a wider real BSS thanks to the scalable architecture of the decision support system. The results of this first tests are interesting even if further investigation are in progress.