arXiv Robotics — research abstracts· Bowen Li, Mayank Mishra, Y. Isabel Liu, Stone Tao, Nishanth Kumar, Alexander G. Gray, Ruwan Wickramarachchi, Jonathan Francis, Sebastian Scherer, Tom Silver·2026-10-12 04:00 UTC· 15 hours agoEditorial score65

Recover, Discover, Plan: Learning Skills and Concepts from Robot Failures

Summary

This research proposes ReSYNC, a method that allows robots to progressively discover and refine state abstractions from failure-recovery experiences, enabling abstract planning. The approach combines skill learning via reinforcement learning with concept refinement through dual-learning, achieving over 50% performance improvements in simulated domains and demonstrating sim-to-real transfer for real-world manipulation tasks.

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.

Environment: SimulationOpen source S4

Source excerpts and review record

Automatically extracted; no manual editorial approval recorded.

obot discovers new relational predicates and refines its abstract planning model to explain and generalize the learned recovery behaviors. This interaction enables ReSYNC to convert local recoveries seen during training into global failure avoidance at test time. Across four simulated domains, we show that ReSYNC's ability to continually expand and refine its abstraction library allows it to solve long-horizon, previ
Open source S4

Source：arXiv Robotics — research abstracts · arxiv.org

#Research #Reinforcement learning #World models #Sim-to-real #Robot policies