Fully Dynamic Rebalancing in Dockless Bike-Sharing Systems via Deep Reinforcement Learning

2026-05-14

AI Summary

This work proposes a fully dynamic, real-time rebalancing approach to address supply-demand imbalances, vehicle unavailability, and mobility deserts in dockless bike-sharing systems caused by static scheduling. The rebalancing problem is formulated as a Markov decision process, and a deep reinforcement learning agent is trained within a graph-based, realistic simulation environment. This agent leverages graph neural networks combined with a spatiotemporal criticality scoring mechanism to guide a single truck in performing localized pickup, drop-off, and charging operations. By abandoning conventional periodic global interventions, the method achieves substantial reductions in vehicle unavailability using only a minimal fleet size, thereby significantly enhancing system reliability and spatial equity.

Abstract

This paper proposes a fully dynamic Deep Reinforcement Learning (DRL) method for rebalancing dockless bike-sharing systems, overcoming the limitations of periodic, system-wide interventions. We model the service through a graph-based simulator and cast rebalancing as a Markov decision process. A DRL agent routes a single truck in real time, executing localized pick-up, drop-off, and charging actions guided by spatiotemporal criticality scores. Experiments on real-world data show significant reductions in availability failures with a minimal fleet size, while limiting spatial inequality and mobility deserts. Our approach demonstrates the value of learning-based rebalancing for efficient and reliable shared micromobility.

Problem

• dockless bike-sharing
• rebalancing
• availability failures
• spatial inequality
• mobility deserts

Innovation

• Deep Reinforcement Learning
• Dynamic Rebalancing
• Dockless Bike-Sharing
• Graph-based Simulation
• Spatiotemporal Criticality