Small Fleet, Big Impact: Enhancing Shared Micromobility Efficiency through Minimal Autonomous Vehicle Deployment

2025-10-05

AI Summary

Shared micromobility systems suffer from spatiotemporal demand fluctuations that cause persistent supply–demand mismatches; conventional static rebalancing strategies—typically executed once or twice daily—fail to adapt to sudden demand surges. To address this, we propose SMART, a novel hierarchical reinforcement learning framework that jointly optimizes high-level initial deployment and low-level real-time rebalancing of Autonomous Shared Micromobility Vehicles (ASMVs), while enabling seamless integration with existing dispatch systems. SMART integrates real-time demand forecasting with an adaptive rebalancing policy. Evaluated on real-world electric scooter data from Chicago, it demonstrates that a small fleet of ASMVs significantly improves service coverage and vehicle utilization across diverse scenarios—including peak hours, special events, and weather anomalies—exhibiting strong generalization and scalability. SMART thus provides an efficient, scalable, and dynamically responsive enhancement for micromobility operations.

Technology Category

• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Intelligent Robots: Multi-Robot Systems
• Multiagent Systems: Multiagent Learning

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Shared micromobility systems, such as electric scooters and bikes, have gained widespread popularity as sustainable alternatives to traditional transportation modes. However, these systems face persistent challenges due to spatio-temporal demand fluctuations, often resulting in a mismatch between vehicle supply and user demand. Existing shared micromobility vehicle scheduling methods typically redistribute vehicles once or twice per day, which makes them vulnerable to performance degradation under atypical conditions. In this work, we design to augment existing micromobility scheduling methods by integrating a small number of autonomous shared micromobility vehicles (ASMVs), which possess self-rebalancing capabilities to dynamically adapt to real-time demand. Specifically, we introduce SMART, a hierarchical reinforcement learning framework that jointly optimizes high-level initial deployment and low-level real-time rebalancing for ASMVs. We evaluate our framework based on real-world e-scooter usage data from Chicago. Our experiment results show that our framework is highly effective and possesses strong generalization capability, allowing it to seamlessly integrate with existing vehicle scheduling methods and significantly enhance overall micromobility service performance.

Problem

• Addressing spatio-temporal demand fluctuations in shared micromobility systems
• Enhancing vehicle rebalancing with autonomous self-rebalancing capabilities
• Improving service performance through hierarchical reinforcement learning optimization

Innovation

• Autonomous vehicles enable dynamic real-time rebalancing
• Hierarchical reinforcement learning optimizes deployment and rebalancing
• Small autonomous fleet integrates with existing scheduling methods