This video addresses a few challenges that occur when using reinforcement learning for production systems and provides some ways to mitigate them. Even if there aren’t straightforward ways to address some of the challenges that you’ll face, at the very least it’ll get you thinking about them.This video will cover the difficulties of verifying the learned solution, pinpointing the location of failures, and manually fixing the policy. It will show some workarounds that make the policy more robust and the overall system safer. Lastly, it will show how you can use reinforcement learning to learn the parameters in a traditional control system and why, at the moment, that can be the best of both worlds.