Failure Prediction with Statistical Guarantees for Vision-Based Robot Control
This addresses safety concerns in robotics by providing statistical guarantees for failure prediction, which is incremental as it builds on PAC-Bayes theory with novel bounds.
The paper tackles the problem of predicting failures in safety-critical robotic systems using high-dimensional vision data, by synthesizing a failure predictor with guaranteed bounds on false-positive and false-negative errors, demonstrated through simulations and hardware experiments with drones and manipulators.
We are motivated by the problem of performing failure prediction for safety-critical robotic systems with high-dimensional sensor observations (e.g., vision). Given access to a black-box control policy (e.g., in the form of a neural network) and a dataset of training environments, we present an approach for synthesizing a failure predictor with guaranteed bounds on false-positive and false-negative errors. In order to achieve this, we utilize techniques from Probably Approximately Correct (PAC)-Bayes generalization theory. In addition, we present novel class-conditional bounds that allow us to trade-off the relative rates of false-positive vs. false-negative errors. We propose algorithms that train failure predictors (that take as input the history of sensor observations) by minimizing our theoretical error bounds. We demonstrate the resulting approach using extensive simulation and hardware experiments for vision-based navigation with a drone and grasping objects with a robotic manipulator equipped with a wrist-mounted RGB-D camera. These experiments illustrate the ability of our approach to (1) provide strong bounds on failure prediction error rates (that closely match empirical error rates), and (2) improve safety by predicting failures.