Introspective Visuomotor Control: Exploiting Uncertainty in Deep Visuomotor Control for Failure Recovery

Abstract

End-to-end visuomotor control is emerging as a compelling solution for robot manipulation tasks. However, imitation learning-based visuomotor control approaches tend to suffer from a common limitation, lacking the ability to recover from an out-of-distribution state caused by compounding errors. In this paper, instead of using tactile feedback or explicitly detecting the failure through vision, we investigate using the uncertainty of a policy neural network. We propose a novel uncertainty-based approach to detect and recover from failure cases. Our hypothesis is that policy uncertainties can implicitly indicate the potential failures in the visuomotor control task and that robot states with minimum uncertainty are more likely to lead to task success. To recover from high uncertainty cases, the robot monitors its uncertainty along a trajectory and explores possible actions in the state-action space to bring itself to a more certain state. Our experiments verify this hypothesis and show a significant improvement on task success rate: 12% in pushing, 15% in pick-and-reach and 22% in pick-and-place.

Rapid prototyping for series of tasks in atypical environment: robotic system with reliable program-based and flexible learning-based approaches

“…One possible solution is to improve the accuracy of the operation by increasing the number of its learning patterns. Future tasks include real-time failure determination and estimation of recovery operations [11,12,25] and improvements to stability through hybridization with conventional control [36].…”

“…However, the learning approach is a black box because it acquires behavioral rules on the basis of data, and there are still problems with reliability. Some methods have been proposed to use a model-based controller as prior knowledge for deep reinforcement learning [8,9] or to monitor the state of a system in real time during operation generation and execute a predetermined recovery action when an anomaly is detected [10][11][12]. However, none of these methods focus on the reliability of behavior.…”