From Real-Time Planning to Reliable Execution:Scalable Coordination for Heterogeneous Multi-Robot Fleets in Industrial Environments
2026-07-01 • Robotics
RoboticsMultiagent Systems
AI summaryⓘ
The authors address the challenge of coordinating many different robots working together in busy industrial settings, where robots have different movement abilities and face delays or disturbances. They created SCALE, a system that plans robot paths in real-time and adapts to unexpected changes to avoid traffic jams and waiting. Their approach includes a way to reduce conflicts caused by robot movements and a smart method to reorder robot priorities when disturbances happen. They tested SCALE extensively, including a three-day real-world use in a warehouse, showing it works well.
heterogeneous robot fleetsreal-time path planningmotion-induced conflict reductionConjugate Action-Precedence Hypergraph (CAPH)robot coordinationexecution uncertaintiesindustrial robotsonline conflict resolution
Authors
Bo Cao, Zhe Liu, Hesheng Wang
Abstract
With the increasing deployment of heterogeneous robot fleets in industrial environments, efficient coordination remains a critical challenge. Real-time path planning must simultaneously accommodate high robot densities and heterogeneous motion capabilities, while communication delays, execution uncertainties, and other disturbances may cause robots to deviate from the temporal assumptions underlying planned paths. Such deviations can lead to excessive waiting and congestion propagation across the fleet. This paper presents SCALE, a reactive online coordination framework that enables real-time planning while maintaining robust execution. Within this framework, we introduce a motion-induced conflict reduction mechanism to support the online generation of feasible paths for online conflict resolution. To mitigate the effects of disturbances, we further design a generalized Conjugate Action-Precedence Hypergraph (CAPH) that adaptively adjusts precedence relations among robots. Extensive validation experiments, together with a three-day deployment in a warehouse, demonstrate the