From Data to Better Steel: AI for Process Control, Quality and Materials

4 November 2026 • 10–11:45am

▶Register Now

Four production deployments trace AI from the plant network down to the grain structure of the steel itself. Suhas Mehta of Falkonry opens with ironmaking operations generating up to 100 million data points a day, introducing the Average Monitoring Reduction Ratio, a measure of how much manual signal review AI removes without sacrificing decision quality. Bryan DeBois of RoviSys follows with a hot rolling deployment combining image-based measurement, historian data, and predictive models to detect and forecast roughing mill turn-up events. Anastasiia Glebova of o-hive.ai turns to surface quality, covering automated defect detection, real-time pass/fail decisions and the root-cause analysis that vision data makes possible. Saeid Kamalpour closes below the surface with transparent grain structure analysis that generates ASTM E112 grain-size numbers and predicts mechanical behavior before material ships. Every speaker focuses on implementation: what worked, what it took and what others should expect.”

Organized by: AIST Digitalization Applications Technology Committee