Back To Schedule
Saturday, October 20 • 2:30pm - 3:00pm
Control Automation with Reinforcement Learning and Engineering Simulation

Sign up or log in to save this to your schedule, view media, leave feedback and see who's attending!

Feedback form is now closed.
The complexity of controlling machineries with sophisticated rules in manufacturing and hard technologies makes them a good candidate for Reinforcement Learning (RL). This has the potential to generate billions of dollars in increased performance and efficiency. RL algorithms usually require a lot of trials and it is expensive and dangerous to run physical trials. Using RL and industrial simulation will not only optimize the rules of operating for more efficiency and time saving, it also can disrupt manufacturing by developing new methods that require less complicated machinery. The proposed solution is an RL as a service that is catered toward control engineers. We used a problem in rocket engines to develop a proof of concept. The goal was to develop a control loop for a system that mixes two fluids (oxygen and methane) and adapt to changing requirements for the output of the mixer. The platform will generate a rule to control inputs to the mixer in response to changes in the requirement for output.

avatar for Edward Mehr

Edward Mehr

CEO, Cyberpunk Lab
Edward Mehr is an entrepreneur and engineer with a background in smart manufacturing and artificial intelligence. He is the founder and CEO of Cyberpunk Lab, a company that provides AI solutions in manufacturing. He also runs a postdoctoral fellowship for Data Science where academics... Read More →

Saturday October 20, 2018 2:30pm - 3:00pm PDT
Ballroom # 406