Event

MATE-BT: A Hierarchical Decentralized Task Framework for Heterogeneous Multi-Robot Systems

1 reports · 1 sources

Sunbeen Lee and co-authors propose MATE-BT, a hierarchical decentralized task autonomy framework for heterogeneous multi-robot teams, which decouples task execution, task allocation, and task execution into three layers, with an onboard active leader responsible for task advancement and team-level task allocation, and for transferring authority after leader failure. The authors state that in physical experiments all computation of MATE-BT runs onboard; hardware-in-the-loop resource measurements support its onboard execution on Jetson Orin Nano and Raspberry Pi 5. By updating only the affected layers rather than regenerating the full plan, the framework maintains a response time of approximately 130 ms with 30 robots and 120 targets.