LLM powered robots limit false claims to prevent fleet disruption
Containing Behavioral Cascades from Manipulated Claims in LLM-Powered Multi-Robot Systems
Robotics
Summary
When groups of robots use large language models to communicate, a false statement about their surroundings can cause many robots to change plans unnecessarily, leading to delays and difficulties. The authors developed a way to catch and contain these false statements quickly by having some robots verify the claim before others change their behavior. This approach helps keep most robots on their correct tasks while only a few adjust temporarily, reducing chaos and extra costs. Their tests show that treating verification as a team problem rather than just trusting or distrusting fixes works better to avoid big disruptions.
What this means in practice
- •For robotics system designers: Incorporate active verification strategies to maintain task efficiency when false environment data arises in multi-robot fleets.
- •For factory automation teams: Improve reliability of warehouse robots by limiting incorrect obstacle reports from causing widespread plan changes and mission delays.
Authors
Waleed Bin Khalid, Byung-Cheol Min
Abstract
Large language model (LLM)-powered multi-robot systems are vulnerable to semantic manipulation: an accepted false world-state claim can trigger a fleet-wide behavioral cascade, causing unnecessary replanning, increased path costs, congestion, or apparent mission infeasibility. Conditioning on a successful manipulation, we propose an active verification framework that contains its downstream effects before they propagate across the fleet. A dedicated verification module generates a structured Verify-Adapt-Hold plan: selected robots inspect consequential regions, a limited subset provisionally adapts when necessary, and the remaining robots retain their trusted plans. We evaluate the framework in a multi-robot transportation environment using injected false obstacle claims across different impacts and team sizes. Evaluation measures cascade containment, Sum-of-Costs, makespan, and coverage ratio. Results show that treating post-compromise verification as a team-level planning problem, rather than a binary trust decision, effectively limits the cascading physical consequences of semantic manipulation.