Bike share programs emerged as crucial components of urban transportation systems, offering an eco-friendly and accessible option for short-distance travel. In Boston, the Bluebike system relied on strategically placed docking stations where users could rent and return bikes. However, the dynamic nature of user demand posed a significant challenge known as the re-balancing problem. Maintaining an optimal distribution of bikes and docking spaces across the network was essential to ensure user satisfaction and system efficiency. Without effective re-balancing, certain stations faced shortages or surpluses, leading to frustration among users and operational inefficiencies. To tackle the re-balancing challenge in Boston's Bluebike system, this project proposed a data-driven approach leveraging machine learning algorithms. By analyzing historical data from April 1, 2022, to April 30, 2022, the project aimed to predict future demand patterns at each docking station. The predictive models considered various factors such as time of day, day of the week, weather conditions, and spatial context. To enhance the model's accuracy, additional features like proximity to educational institutions and socioeconomic characteristics of the surrounding areas were explored. The ultimate goal was to develop an intelligent re-balancing strategy that optimized bike distribution, minimized user inconvenience, and maximized system efficiency. By proactively addressing the re-balancing problem, this project sought to contribute to the sustainable growth and success of Boston's bike share program.
