This paper presents a two-step approach to managing service level requirements (SLRs) in real-life bicycle sharing systems (BSSs). SLR is a general concept that broadly describes how to effectively manage a BSS to improve user satisfaction while minimizing system operation costs. The two steps consist of two proposed problems: First, the target-level problem computes target bicycle quantities for stations, maximizing trip satisfaction. Second, the bicycle rebalancing problem designs vehicle routes to adjust bicycle quantities. The SLR literature includes several variants of these problems, but to our knowledge, very few exact approaches such as the one we propose can successfully handle real-life BSS, which comprise thousands of stations, tens of thousands of bicycles, and nearly 100,000 daily trips. We gather data from real-life BSSs from Boston, Chicago, Madrid, Mexico City, Montreal, New York, San Francisco, Toronto, and Washington, DC. From a managerial perspective, our numerical results provide the decision makers of BSSs with several insights related to bicycle and station usage throughout the network of stations. Funding: This work was supported by the Canadian Natural Sciences and Engineering Research Council (NSERC) under Grant 2021-04037.
Parada et al. (Mon,) studied this question.