Toward a better monitoring statistic for profile monitoring via variational autoencoders. Issue 5 (20th October 2021)

• Record Type: Journal Article
• Authors: Sergin, Nurettin Dorukhan; Yan, Hao
• Is Part Of: Journal of quality technology. Volume 53:Issue 5(2021)
• Page Start: 454
• Page End: 473
• Publication Date: 2021-10-20
• Subjects: Deep learning -- high-dimensional nonlinear profile -- latent variable model -- profile monitoring -- variational autoencoder; Quality control -- Periodicals; Qualité -- Contrôle -- Périodiques; Quality control; Quality control; Periodicals; 620.0045
• DOI: 10.1080/00224065.2021.1903821
• Languages: English
• ISSNs: 0022-4065

Abstract

Abstract: Variational autoencoders have been recently proposed for the problem of process monitoring. While these works show impressive results over classical methods, the proposed monitoring statistics often ignore the inconsistencies in learned lower-dimensional representations and computational limitations in high-dimensional approximations. In this work, we first manifest these issues and then overcome them with a novel statistic formulation that increases out-of-control detection accuracy without compromising computational efficiency. We demonstrate our results on a simulation study with explicit control over latent variations, and a real-life example of image profiles obtained from a hot steel rolling process.