The classifier judged this item not relevant to AI security, so it is left out of lists, feeds, the API and dataset releases. If that is wrong, report a correction on this page.
InfoResearchPreprint
Safe Multi-Robot Collaboration With Uniform Exclusion Using Trust
- Published
- Record updated
Summary
The paper presents a Safe Wrapper for robot swarms that filters incoming communication links so that compromised links carrying malicious data are excluded. It selects trusted links by testing a stream of trust values from an anomaly detector with Safe Anytime-Valid Inference, requiring only that the detector perform better than chance. Hardware experiments on 3 LoCoBots and 8 GoPiGos for neural implicit mapping and heading consensus, plus simulations, indicate the approach scales to swarms of hundreds of robots.