Dynamic Incentive Schemes for Managing Dockless Bike-Sharing Systems

Balancing supply and demand for a dockless bike-sharing system is difficult as users are allowed to pick up and return bikes anywhere within a large service area. One strategy to help manage the amount of dockless bikes scattered within a service area is to offer monetary rewards for users who would be willing to pick up bikes scattered around some central location or to return their bikes at this central location. We develop a modeling framework to analyze the effectiveness of using dynamic incentive schemes for balancing the amount of bikes in the system to minimize the expected operating cost. Using an extensive set of numerical experiments, we illustrate specific operation environments under which these dynamic incentive schemes would be most effective for reducing the operating cost of the system. We find that the use of dynamic pickup and return rewards can generate very substantial cost reductions in an operating environment with a high traffic intensity of bike return outside the central location and a high overall traffic intensity of bike returns relative to bike pickups in the system. We also find that return rewards are generally more cost-effective than pickup rewards