Event
New Method Proposed for Humanoid Robot Recoverability Monitoring
1 reports · 1 sources
Zeyuan Feng and other authors propose an active reachability learning framework for learning the stochastic finite-horizon recoverable envelope of a given fallback policy, with the goal of triggering fallback before an unrecoverable fall occurs. The method reframes data exploration as a reach-avoid problem, in which an uncertainty-guided explorer discovers unseen states and avoids definitely unrecoverable regions, and combines temporal difference with Monte Carlo targets to improve value prediction and uncertainty estimation. The authors evaluate on stochastic benchmarks and on a 29-DoF Unitree G1 humanoid robot, reporting markedly improved monitoring coverage and accuracy, and zero-shot transfer to real hardware.
Report timeline
- Active Reachability Learning for Humanoid Recoverability Monitoring
The paper proposes an active reachability learning framework for learning the stochastic finite-horizon recoverability envelope of a given fallback policy, so that fallback can be triggered before an unrecoverable fall occurs. The method reframes data exploration as a reach-avoid problem, in which an uncertainty-guided explorer discovers unseen states while avoiding confidently unrecoverable regions, and combines temporal-difference and Monte Carlo objectives to improve in-distribution value prediction and uncertainty estimation. The authors evaluate on stochastic benchmarks and a 29-DoF Unitree G1 humanoid robot, reporting clear improvements in monitoring coverage and accuracy, and zero-shot transfer to real hardware.
arXiv cs.RO ·