Search papers, labs, and topics across Lattice.
This paper introduces REPAIR, a human-in-the-loop framework that integrates on-demand human assistance for physical action failures into LLM-based multi-robot coordination. When robots encounter irrecoverable physical failures, the LLM requests assistance from a human operator to enable task continuity through remote intervention. Experiments in a real-world multi-robot trash collection task demonstrate that REPAIR significantly improves task progress compared to fully autonomous methods, achieving performance comparable to full remote control for easily collectable items.
LLM-orchestrated multi-robot systems can overcome physical execution failures and achieve near-teleoperation performance by intelligently requesting human assistance only when needed.
Multi-robot coordination based on large language models (LLMs) has attracted growing attention, since LLMs enable the direct translation of natural language instructions into robot action plans by decomposing tasks and generating high-level plans. However, recovering from physical execution failures remains difficult, and tasks often stagnate due to the repetition of the same unsuccessful actions. While frameworks for remote robot operation using Mixed Reality were proposed, there have been few attempts to implement remote error resolution specifically for physical failures in multi-robot environments. In this study, we propose REPAIR (Robot Execution with Planned And Interactive Recovery), a human-in-the-loop framework that integrates remote error resolution into LLM-based multi-robot planning. In this method, robots execute tasks autonomously; however, when an irrecoverable failure occurs, the LLM requests assistance from an operator, enabling task continuity through remote intervention. Evaluations using a multi-robot trash collection task in a real-world environment confirmed that REPAIR significantly improves task progress (the number of items cleared within a time limit) compared to fully autonomous methods. Furthermore, for easily collectable items, it achieved task progress equivalent to full remote control. The results also suggested that the mental workload on the operator may differ in terms of physical demand and effort. The project website is https://emergentsystemlabstudent.github.io/REPAIR/.
