Warehouse robot teams keep working despite communication loss
Hierarchical Multi-agent Reinforcement Learning for Warehouse Robot Coordination under Communication Loss
Robotics
Summary
Robots working together in warehouses usually need to talk to each other all the time to get their jobs done. This paper shows how groups of robots can keep working well even when they lose some communication. The authors split robots into smaller groups that coordinate inside the group using memory, while a bigger policy manages coordination between groups. They also have a safety check to stop unsafe moves. Their tests in simulations showed the robots finished tasks better with less communication and stayed safe.
What this means in practice
- •For warehouse automation teams: Improve coordination of robot fleets that face communication dropouts while moving goods efficiently and safely.
- •For industrial robotics engineers: Design control strategies that handle partial communication failures in multi-robot systems with layered decision-making and safety checks.
Authors
Weihao Sun, Gehui Xu, Andreas A. Malikopoulos
Abstract
In this paper, we propose a hierarchical multi-agent reinforcement learning framework for coordinating robot teams in warehouse environments under communication loss. We partition the robot team into groups, with centralized coordination within each group and distributed coordination across groups. Each group uses a recurrent predictor to estimate unavailable interaction information due to communication loss. A higher-level policy then generates a compact coordination reference that conditions the local control policy within each group. A predictive safety filter evaluates and modifies the proposed controls when they violate safety constraints. Simulation results show improved task completion under communication loss, reduced communication growth as the team size increases, and safe operation in the tested scenarios.