Event
Geometry-Aware Reinforcement Learning Enables Close-Fitting Wearable Insertion
1 reports · 1 sources
Takuma Tsukakoshi and co-authors model robot-assisted dressing as a geometry-aware insertion problem and propose a reinforcement learning framework combining a staged curriculum with per-finger insertion soft gating. In post-training evaluation on the deformable wristband task, the full method achieved a success rate of 84.7±1.2%; the authors state that insertion gating contributed the most, and that the score was highest when combined with the curriculum. On a mannequin hand, the same policy succeeded in 15 of 20 real-robot trials.
Report timeline
- Geometry-Aware Reinforcement Learning for Close-Fitting Wearable Insertion
The paper formulates robot-assisted dressing as a geometry-aware insertion problem and proposes a reinforcement learning framework combining a staged curriculum with a finger-wise insertion soft gate. In post-training evaluation on the deformable bracelet task, the full method achieved a success rate of 84.7±1.2%, with the insertion gate contributing the most and the highest score when combined with the curriculum. On a dummy hand, the same policy succeeded in 15 of 20 real-robot trials.
arXiv cs.RO ·