Industrial Machine Learning: Cost-Effectiveness, Scalability, and Generalizability

Chenhui Shao, PhD
University of Michigan

Abstract: Machine learning plays a pivotal role in smart manufacturing by powering a wide array of decision-making tasks. Since machine learning heavily relies on data, the “cost of data” now accounts for an ever-increasing proportion of the “cost of decision-making” in modern manufacturing. Moreover, while traditional machine learning methods have proven to be highly effective in stable, well-defined settings, they often struggle to scale or adapt to complicated, rapidly evolving production environments. The cost-prohibitive nature of manufacturing data together with the lack of scalability and generalizability have become critical barriers to the widespread adoption of industrial machine learning. To address these challenges, we have developed a suite of methodologies, including sampling design (or active learning), transfer learning, domain adaptation and generalization, and federated learning, which span the full life cycle of data-driven decision-making. Real-world case studies from diverse manufacturing applications (e.g., additive manufacturing, micro- and nano-scale two-photon lithography, ultrasonic metal welding, high-precision machining) will be discussed to illustrate these methodologies and demonstrate their effectiveness in various decision-making tasks such as part qualification, online monitoring, maintenance, fault diagnosis, and quality improvement.

Bio: Dr. Chenhui Shao is an Associate Professor in the Department of Mechanical Engineering at the University of Michigan, Ann Arbor. Before joining Michigan, he served on the faculty at the University of Illinois at Urbana-Champaign between 2016 and 2023, where he advanced to the rank of Associate Professor in the Department of Mechanical Science and Engineering. Chenhui received his B.E. degree in Automation from the University of Science and Technology of China; M.S.E. degree in Industrial and Operations Engineering, M.A. degree in Statistics, and Ph.D. degree in Mechanical Engineering, all from the University of Michigan, Ann Arbor. His research focuses on smart manufacturing, machine learning, statistics, materials joining, and manufacturing systems control and automation. Chenhui’s honors and awards include NSF CAREER Award, ASME Chao and Trigger Young Manufacturing Engineer Award, IISE Manufacturing and Design Division Outstanding Young Investigator Award, SME Barbara M. Fossum Outstanding Young Manufacturing Engineer Award, SME 30 Under 30 Honoree, and Dean’s Award for Excellence in Research at Illinois. His research contributions have been recognized by multiple best paper awards and highlighted in the Manufacturing USA Report. In 2021, Chenhui was selected as one of 30 nationwide to attend the ASEE DELTA Junior Faculty Institute. He currently serves as an associate editor for the Journal of Manufacturing Processes and the ASME Journal of Dynamic Systems, Measurement, and Control.