Monitoring for changes in the nature of stochastic textured surfaces

Abstract

We propose an approach for monitoring general global changes in the nature of stochastic textured surfaces using streams of high-dimensional images or related profile data. Stochastic textured surfaces are fundamentally different than the profiles and images that are the focus of most prior profile monitoring works. We represent normal in-control behavior by using supervised learning algorithms to implicitly characterize the joint distribution of the stochastic textured surface pixels. Based on this characterization, we develop a control chart monitoring statistic using likelihood-ratio principles to quantify and detect changes in the stochastic nature of the surfaces, relative to the in-control surfaces. Unlike methods that look for changes in specific predefined features, our approach can detect very general changes in the nature of the textured surfaces. We demonstrate the implementation and effectiveness of the approach with a real textile example and a simulation example.

Keywords:

• anomaly detection
• fault detection
• generalized likelihood-ratio test (GLRT)
• Markov random field
• one-class classification
• statistical process control (SPC)
• supervised learning

Acknowledgments

The authors thank the two anonymous referees for their helpful comments.

Additional information

Funding

This work was supported in part by NSF Grant # CMMI-1265709 and AFOSR Grant # FA9550-14-1-0032, which the authors gratefully acknowledge. Anh Tuan Bui was also supported by the Vietnam Education Foundation.

Notes on contributors

Anh Tuan Bui

Anh Tuan Bui is a Ph.D. Candidate in the Department of Industrial Engineering and Management Sciences, Northwestern University. He obtained the B.S. degree in Electrical Engineering from Hanoi University of Science and Technology, and the M.S. degree in Industrial and Management Engineering from Pohang University of Science and Technology. His research interests lie in engineering statistics, statistical learning, and data analytics, with applications in manufacturing, materials, and enterprise engineering. He received the Mary G. and Joseph Natrella Scholarship award from the Quality and Productivity Section of the American Statistical Association in 2018.

Daniel W. Apley

Daniel W. Apley is a Professor of Industrial Engineering and Management Sciences at Northwestern University, Evanston, IL. His research interests are at the interface of engineering modeling, statistical analysis, and predictive analytics, with particular emphasis on enterprise process modeling and manufacturing variation reduction applications in which large amounts of data are available. He received the NSF CAREER award in 2001, the IIE Transactions Best Paper Award in 2003, and the Wilcoxon Prize for best practical application paper appearing in Technometrics in 2008. He is Editor-in-Chief of Technometrics and has served as Editor-in-Chief for the Journal of Quality Technology, Chair of the Quality, Statistics & Reliability Section of INFORMS, and Director of the Manufacturing and Design Engineering Program at Northwestern.