Presentation Profile

From Predictive to Prescriptive: Validating AI in Plant Operations

Main Author
David Cuthbert - Wasson-ECE Instrumentation

Abstract Number: 186
Abstract:

Last year’s GCC presentation, "The Data-Driven Plant," covered preparing analytical data for AI. A year later, predictive maintenance has moved into production, prescriptive and even autonomous actions are emerging, plant AI copilots are live, and the EU AI Act now classifies process-control AI as "high-risk." This paper shifts the question from "how do we prepare our data" to "how do we know we can still trust the model." Topics include model drift, synthetic training data, and the new governance obligations facing analytical and QA teams.   We will evaluate how our roles and responsibilities as analytical experts continue to evolve.

Back to main author bio