Trip Route Planning for Bicycle-Sharing Systems

Bicycle-sharing systems (BSSs) offering shared bike usages to the public are becoming more and more popular nowadays. In bicycle-sharing systems, people can borrow and return bikes at any bike stations in the service region, but the free-ride time is usually limited. Therefore, for long-distance bike trips, individuals need to pre-schedule the bike trip route in advance and change the bike within the free-ride time so no over-time fees will be charged. In this paper, we will study the trip route planning problem for individuals when using the bicycle-sharing systems. Given the trip origin and destination, we aim at identifying the optimal trip route from the origin to the destination through the bike stations. To address the problem, we conduct a thorough analysis about an existing BSS, Divvy, launched in Chicago. Based on the analysis results, a novel bike route planning framework " BSSs based Trip rOute Planning" (STOP) is proposed in this paper, which can identify the optimal trip route by mapping the problem into a minimum-cost network flow problem. Extensive experiments conducted on real-world bicycle-sharing system datasets demonstrate the effectiveness of STOP.