Skip to content
SIG//79112026.09.07 · 20:33 UTCARXIV

Enhancing Failure Detection in Vision-Language-Action Policies for Robotics

A new framework improves failure detection in robotic manipulation tasks.

Vision-language-action (VLA) policies are promising for robotic manipulation but struggle with failure detection during long tasks. Traditional methods often detect failures post-action or rely on inadequate supervision, leading to inaccuracies.

This study introduces a data-efficient framework that utilizes unlabeled VLA action segments to create weak supervision signals, identifying abnormal behaviors. By employing active learning, the approach selectively annotates uncertain trajectories, enhancing both timestamp-level and trajectory-level failure detection.

More from Research & Papers

Enhancing Failure Detection in Vision-Language-Action Policies for Robotics — RoboSignal