November Meeting Plans
The next meeting of the ISA Birmingham Section will be on Tuesday, November 12, 2024, at 5:30 PM in the offices of Revere Control Systems, 5201 Princeton Way, Hoover, AL. This will be a technical meeting with Professional Development Hour (PDH) offered for those needing credit for professional or company record. The session will begin at 5:00 PM with networking and social activity, along with the business meeting announcements. The speaker will begin at 5:30 PM and conclude with a Q & A session prior to 6:30 PM. In Person and on Zoom – Connection Link Below.
Speaker: Thomas Casey – TriNova Director of AI
Topic: “The future of AI in process controls systems”
Thomas Casey spent more than 10 years in academic and industrial positions using Magnetic Resonance Spectroscopy in biophysics applications while developing data analysis processes and corresponding software and applications. In 2021 Thomas transitioned to Evonik where, as a Senior Digital Process Engineer, he developed digital twins and AI methods for process optimization. In 2024 Thomas joined TriNova where he now supports their partner, Ai-OPs, in their mission to expand the use of AI for industrial controls strategies. Thomas holds a B.S. in Biochemistry, a Ph.D. in Physical Chemistry, has taught courses in Physical Chemistry at two universities, co-authored several peer reviewed publications, and enjoys developing AI and machine learning solutions in python.
Abstract:
For AI to become the basis of next-generation controls strategies it must root itself in real domain expertise and learn intelligence through as close an approximation of real learning as possible. Leveraging their extensive domain expertise, the Ai-OPs team has figured out how to balance complexity and usability in AI based controls strategies to most effectively apply the unique capabilities of AI. Of the many methods that can be classified as AI, the one that most closely approximates real intelligence is called Deep Reinforcement Learning (DRL). DRL builds artificial intelligence much like process engineers build real intelligence, through trial and error, repetition, and memory. Trained DRL models can observe process data and make controls decisions by retrieving memorized intelligence that was built through trial and error and stored in neural networks.
CLICK ON THE ZOOM CONNECTION LINK: https://us02web.zoom.us/j/87030705904
Wait to be admitted from the waiting room. YOU DO NOT WANT TO MISS THIS MEETING!!