Global Rewards in Multi-Agent Deep Reinforcement Learning for Autonomous Mobility on Demand Systems
This addresses performance issues in scalable vehicle dispatching for mobility services, though it appears incremental by modifying reward structures in existing multi-agent frameworks.
The paper tackles vehicle dispatching in autonomous mobility on demand systems by proposing a global-rewards-based multi-agent deep reinforcement learning algorithm, which shows statistically significant improvements over state-of-the-art methods with local rewards on real-world data.
We study vehicle dispatching in autonomous mobility on demand (AMoD) systems, where a central operator assigns vehicles to customer requests or rejects these with the aim of maximizing its total profit. Recent approaches use multi-agent deep reinforcement learning (MADRL) to realize scalable yet performant algorithms, but train agents based on local rewards, which distorts the reward signal with respect to the system-wide profit, leading to lower performance. We therefore propose a novel global-rewards-based MADRL algorithm for vehicle dispatching in AMoD systems, which resolves so far existing goal conflicts between the trained agents and the operator by assigning rewards to agents leveraging a counterfactual baseline. Our algorithm shows statistically significant improvements across various settings on real-world data compared to state-of-the-art MADRL algorithms with local rewards. We further provide a structural analysis which shows that the utilization of global rewards can improve implicit vehicle balancing and demand forecasting abilities. Our code is available at https://github.com/tumBAIS/GR-MADRL-AMoD.