A Bayesian approach to breaking things: efficiently predicting and repairing failure modes via sampling

Charles Dawson, Chuchu Fan
Proceedings of The 7th Conference on Robot Learning, PMLR 229:1706-1722, 2023.

Abstract

Before autonomous systems can be deployed in safety-critical applications, we must be able to understand and verify the safety of these systems. For cases where the risk or cost of real-world testing is prohibitive, we propose a simulation-based framework for a) predicting ways in which an autonomous system is likely to fail and b) automatically adjusting the system’s design to preemptively mitigate those failures. We frame this problem through the lens of approximate Bayesian inference and use differentiable simulation for efficient failure case prediction and repair. We apply our approach on a range of robotics and control problems, including optimizing search patterns for robot swarms and reducing the severity of outages in power transmission networks. Compared to optimization-based falsification techniques, our method predicts a more diverse, representative set of failure modes, and we also find that our use of differentiable simulation yields solutions that have up to 10x lower cost and requires up to 2x fewer iterations to converge relative to gradient-free techniques.

Cite this Paper

BibTeX
@InProceedings{pmlr-v229-dawson23a, title = {A Bayesian approach to breaking things: efficiently predicting and repairing failure modes via sampling}, author = {Dawson, Charles and Fan, Chuchu}, booktitle = {Proceedings of The 7th Conference on Robot Learning}, pages = {1706--1722}, year = {2023}, editor = {Tan, Jie and Toussaint, Marc and Darvish, Kourosh}, volume = {229}, series = {Proceedings of Machine Learning Research}, month = {06--09 Nov}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v229/dawson23a/dawson23a.pdf}, url = {https://proceedings.mlr.press/v229/dawson23a.html}, abstract = {Before autonomous systems can be deployed in safety-critical applications, we must be able to understand and verify the safety of these systems. For cases where the risk or cost of real-world testing is prohibitive, we propose a simulation-based framework for a) predicting ways in which an autonomous system is likely to fail and b) automatically adjusting the system’s design to preemptively mitigate those failures. We frame this problem through the lens of approximate Bayesian inference and use differentiable simulation for efficient failure case prediction and repair. We apply our approach on a range of robotics and control problems, including optimizing search patterns for robot swarms and reducing the severity of outages in power transmission networks. Compared to optimization-based falsification techniques, our method predicts a more diverse, representative set of failure modes, and we also find that our use of differentiable simulation yields solutions that have up to 10x lower cost and requires up to 2x fewer iterations to converge relative to gradient-free techniques.} }
Endnote
%0 Conference Paper %T A Bayesian approach to breaking things: efficiently predicting and repairing failure modes via sampling %A Charles Dawson %A Chuchu Fan %B Proceedings of The 7th Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2023 %E Jie Tan %E Marc Toussaint %E Kourosh Darvish %F pmlr-v229-dawson23a %I PMLR %P 1706--1722 %U https://proceedings.mlr.press/v229/dawson23a.html %V 229 %X Before autonomous systems can be deployed in safety-critical applications, we must be able to understand and verify the safety of these systems. For cases where the risk or cost of real-world testing is prohibitive, we propose a simulation-based framework for a) predicting ways in which an autonomous system is likely to fail and b) automatically adjusting the system’s design to preemptively mitigate those failures. We frame this problem through the lens of approximate Bayesian inference and use differentiable simulation for efficient failure case prediction and repair. We apply our approach on a range of robotics and control problems, including optimizing search patterns for robot swarms and reducing the severity of outages in power transmission networks. Compared to optimization-based falsification techniques, our method predicts a more diverse, representative set of failure modes, and we also find that our use of differentiable simulation yields solutions that have up to 10x lower cost and requires up to 2x fewer iterations to converge relative to gradient-free techniques.
APA
Dawson, C. & Fan, C.. (2023). A Bayesian approach to breaking things: efficiently predicting and repairing failure modes via sampling. Proceedings of The 7th Conference on Robot Learning, in Proceedings of Machine Learning Research 229:1706-1722 Available from https://proceedings.mlr.press/v229/dawson23a.html.

Related Material