Satish Bukkapatnam
Texas A&M University
Presenting in Track 3: Advanced Manufacturing
Presentation Title: AI in Manufacturing: Beyond Prediction, and towards Discovery
Abstract: During the past four decades, various AI and machine learning methods have been adapted for different manufacturing applications. The emergence of large streams of multimodal data has spurred diverse R&D efforts in academia and industry to harness AI/ML to improve decision-making and performance. Central to these efforts are the AI/ML models that can learn the underlying relationships from a variety of data sources to predict, organize, and control the outcomes. Harnessing these AI/ML models for knowledge discovery is evoking a growing interest in various disciplines.
This talk presents ongoing efforts in explainable AI ("XAI") for deducing plausible physics that underpin certain data and observations from manufacturing processes. The talk introduces an XAI method that can offer some guarantees on the consistency of the explanations and discusses its application to deduce interesting physical insights from the measured sensor data and image snapshots from a variety of manufacturing processes.
Biography: Satish Bukkapatnam is a Regents Professor and Rockwell International Professor of Industrial & Systems Engineering at Texas A&M University and serves as a program director at NSF's Advanced Manufacturing Program. He received his Ph.D. degree in Industrial and Manufacturing Engineering from Pennsylvania State University. His research interests are broadly in smart and precision manufacturing. Dr. Bukkapatnam is a Fellow of IISE and SME, an Associate member of CIRP, and a Fulbright-Tocqueville Distinguished Chair.