Leave It Better Than Found: Value-Guided Cooperation in Sequential Tasks
Leave It Better Than Found: Value-Guided Cooperation in Sequential Tasks
This research addresses sequential task continuation where one robot operates in a workspace modified by another, proposing a method for ad hoc cooperation between robots that have never operated together. By sharing value functions to communicate environmental preferences and using a k-nearest neighbors (KNN) uncertainty filter to prevent overconfidence in out-of-distribution states, the approach achieves up to 48.5 percentage point improvement in task success rates compared to naive or selfish agents across simulated and real-world experiments.
The paper proposes a method for ad hoc multi-robot cooperation where agents share value functions to communicate environmental preferences without prior joint training, supported by a KNN-based uncertainty filter to handle out-of-distribution states; results show up to 48.5 percentage point improvement in task success rates in both simulation and real-world experiments.
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.
r even selfish agents, in task success rates across simulated and real-world experiments, demonstrating that value function transmission between agents is useful not merely as an internal learning tool, but as an effective basis for ad hoc coordination. Website: https://proroklab.github.io/PAVE/
Open source S5
Source：arXiv Robotics — research abstracts · arxiv.org