Impact aware dexterous catching improved by outcome sensitive motion search
Outcome-Sensitive Motion Search for Impact-Aware Dexterous Catching
Robotics
Summary
Catching fast objects gently requires carefully timed movements to match speed and reduce impact. The authors found that teaching robots to catch this way with trial-and-error learning is hard because small differences in approach can cause big changes in success. They created a way to identify key motion windows that affect catching outcomes and used this to improve robot training by refining actions where the robot's initial learning failed. Their method helped robots catch objects more reliably and with less impact than the original trained system.
What this means in practice
- •For robotics developers: Develop robot controllers that catch fast objects more reliably and with reduced damage from impact.
- •For automation engineers: Improve automated systems handling delicate, fast-moving parts by learning better contact timing and motion adjustment.
Authors
Guorui Pei, Jinsong Wu, Songyuan Su, Jiaming Qi, Sichao Liu, David Navarro-Alarcon, Bin Liu, Peng Zhou
Abstract
Skilled humans can catch fast-moving objects softly by coordinating interception, velocity matching, and follow-through to mitigate impact. Learning such impact-aware catching with reinforcement learning (RL), however, is challenging, as the policy must achieve reliable interception and grasping while regulating the sensitive transition into contact. Moreover, even a capable privileged-state RL teacher may not provide ideal demonstrations for a deployable imitation-learning (IL) student: teacher failures limit task coverage, while small variations in pre-contact motion can produce substantially different impact and grasping outcomes. We characterize this phenomenon through interventional outcome sensitivity and introduce the outcome-sensitive window (OSW) to guide targeted demonstration construction. Building on this formulation, we propose Outcome-Sensitive Motion Search, which learns a task-conditioned manifold of successful OSW motions and performs local geodesic search to refine successful teacher rollouts and repair task conditions where the teacher fails. We then validate candidate motions through complete rollouts under a calibrated IL-student action-error model and retain only successful executions as demonstrations. Extensive simulation experiments demonstrate that our method effectively repairs task conditions where the teacher fails and enables the resulting IL policy to outperform the privileged RL teacher in both catching success and impact mitigation.