Safe Reinforcement Learning With Model Uncertainty Estimates
Name
1810.08700.pdf
Description
Submitted version
Size
2.08 MB
Format
Adobe PDF
Checksum (MD5)
c50185f1df5cdde50f214688d704cd3a
Author(s)
Lutjens, Bjorn
Everett, Michael F
How, Jonathan P
Date Issued
August 2019
Journal
2019 International Conference on Robotics and Automation
Publisher
IEEE
Citation
Lutjens, Bjorn, Everett, Michael, and How, Jonathan P., "Safe Reinforcement Learning With Model Uncertainty Estimates." 2019 International Conference on Robotics and Automation (ICRA), May 2019, Montreal, Canada, IEEE, August 2019.
Version
Original manuscript
Abstract
Many current autonomous systems are being designed with a strong reliance on black box predictions from deep neural networks (DNNs). However, DNNs tend to be overconfident in predictions on unseen data and can give unpredictable results for far-from-distribution test data. The importance of predictions that are robust to this distributional shift is evident for safety-critical applications, such as collision avoidance around pedestrians. Measures of model uncertainty can be used to identify unseen data, but the state-of-the-art extraction methods such as Bayesian neural networks are mostly intractable to compute. This paper uses MC-Dropout and Bootstrapping to give computationally tractable and parallelizable uncertainty estimates. The methods are embedded in a Safe Reinforcement Learning framework to form uncertainty-aware navigation around pedestrians. The result is a collision avoidance policy that knows what it does not know and cautiously avoids pedestrians that exhibit unseen behavior. The policy is demonstrated in simulation to be more robust to novel observations and take safer actions than an uncertainty-unaware baseline. Keywords: Uncertainty; Collision avoidance; Neural networks; Computational modeling; Training; Data models; Reinforcement learning
MIT Department
Massachusetts Institute of Technology. Aerospace Controls Laboratory
Massachusetts Institute of Technology. Department of Aeronautics and Astronautics
Terms of Use
Creative Commons Attribution-Noncommercial-Share Alike
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1109/icra.2019.8793611
