Title:

Distributionally Robust Multi-Agent Reinforcement Learning for Dynamic Chute Mapping  at Amazon Sortation Centers

Abstract:

In Amazon robotic warehouses, the destination-to-chute mapping problem is crucial for efficient package sorting. Often, however, this problem is complicated by uncertain and dynamic package induction rates, which can lead to increased package recirculation. To tackle this challenge, we introduce a Distributionally Robust Multi-Agent Reinforcement Learning (DRMARL) framework that learns a destination-to-chute mapping policy that is resilient to adversarial variations in induction rates. Specifically, DRMARL relies on group distributionally robust optimization (DRO) to learn a policy that performs well not only on average but also on each individual subpopulation of induction rates within the group that capture, for example, different seasonality or operation modes of the system. This approach is then combined with a novel contextual bandit-based estimator of the worst-case induction distribution for each state-action pair, significantly reducing the cost of exploration and thereby increasing the learning efficiency and scalability of our framework. Extensive simulations demonstrate that DRMARL achieves robust chute mapping in the presence of varying induction distributions, reducing package recirculation by an average of 80% in the simulation scenario.

Bio:

Guangyi Liu is a Postdoctoral Research Scientist at Amazon Robotics, where he develops reinforcement learning and optimization algorithms for large-scale warehouse automation. He received his Ph.D. in Mechanical Engineering from Lehigh University, advised by Prof. Nader Motee. His research focuses on risk-aware decision-making and control for networked autonomous systems, with recent work on distributionally robust multi-agent reinforcement learning, cascading failure analysis in complex networks, and safety under perception and communication uncertainty. Since 2023, Guangyi has also served as the lead organizer for invited sessions and workshops at the American Control Conference.