Action-Consistent Decentralized Belief Space Planning with Inconsistent Beliefs and Limited Data Sharing: Framework and Simplification Algorithms with Formal Guarantees
Action-Consistent Decentralized Belief Space Planning with Inconsistent Beliefs and Limited Data Sharing: Framework and Simplification Algorithms with Formal Guarantees
This research introduces decentralized multi-robot belief space planning (MR-BSP) algorithms designed to ensure safe decision-making under inconsistent beliefs and limited data sharing. The authors present VerifyAC, which guarantees consistent joint action selection via a three-step verification process, triggering communication only when necessary. Extensions R-VerifyAC and R-VerifyAC-simp further reduce communication frequency and improve computation time by relaxing consistency criteria and verifying partial observations. Theoretical guarantees are supported by discrete simulations, while experimental results demonstrate applicability to continuous, high-dimensional state spaces using active multi-robot visual SLAM with real robots.
The paper proposes decentralized multi-robot belief space planning algorithms that handle inconsistent beliefs and limited data sharing, offering formal guarantees on joint action consistency and reduced communication overhead, validated through discrete simulations and active visual SLAM experiments with real robots.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- paper: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
Source：arXiv Robotics — research abstracts · arxiv.org