NAYLI NAZURAH BINTI NORHALIM
A Research on the Redistribution Strategy of Bicycle Sharing System Through Historical Usage Pattern Analysis
Kazushi SANO
Bicycle Sharing Systems (BSS) offer a sustainable solution to urban mobility, addressing the last-mile problem and reducing car dependency.  Despite the benefits, BSS face operational challenges, particularly in maintaining bicycle availability across stations. “Empty stations” occurrence, where no bicycle is available arise from fluctuating rental and return throughout the day, influenced by factors such as station locations, user patterns, and weather. Failure to address these imbalances results in increased operational costs and reduced service quality.
To ensure the long-term sustainability of BSS, this research proposes a cost-effective static redistribution model that incorporates historical data to optimize bicycle allocation while minimize operational costs. The approach consists of two phases. Phase One analyzes historical usage patterns including bicycle GPS tracking, empty station logs, and weather conditions to determine demand fluctuations and set minimum bicycle per station. By categorizing demand patterns based on time of day and weather, this phase establishes the necessary minimum number of bicycle requirements for each station to prevent empty station. Phase Two develops a static redistribution model, which involves identifying surplus and deficit stations, optimizing vehicle routing, and prioritizing high-demand stations during low demand period. The model applies the Capacitated Vehicle Routing Problem (CVRP) to optimize redistribution while adhering to operational constraints, such as truck capacity and coverage level threshold. Additionally, different scenarios are considered to develop a cost-effective redistribution strategy that prioritizes demand, even when the number of surplus bicycles is insufficient to fully meet demand. Sensitivity analyses show the model's adaptability to changes in demand, ensuring optimal redistribution strategies that balance service levels and operational costs. 
The model was tested on Niigata City's BSS, successfully minimized unnecessary trips, optimized bicycle distribution, and reduces the empty station occurrence. These improvements led to enhanced user convenience and lower operational costs. The findings suggest that this model can be applied to other cities, supporting the development of sustainable urban transportation.