SSMSPC: self-supervised multivariate statistical in-process control in discrete manufacturing processes
نویسندگان
چکیده
Abstract Self-supervised learning has demonstrated state-of-the-art performance on various anomaly detection tasks. Learning effective representations by solving a supervised pretext task with pseudo-labels generated from unlabeled data provides promising concept for industrial downstream tasks such as process monitoring. In this paper, we present SSMSPC novel approach multivariate statistical in-process control (MSPC) based self-supervised learning. Our motivation is to leverage the potential of unsupervised representation incorporating into general (SPC) framework develop holistic and localization anomalous behavior in discrete manufacturing processes. We propose called Location + Transformation prediction, where objective classify both, type location randomly applied augmentation given time series input. task, follow one-class classification setting apply Hotelling’s $$T^2$$ T 2 statistic learned representations. further an extension chart view that combines metadata visualize steps which supports machine operator root cause analysis. evaluate effectiveness two real-world CNC-milling datasets show it outperforms approaches, achieving $$100\%$$ 100 % $$99.6\%$$ 99.6 AUROC, respectively. Lastly, deploy at demonstrate its practical applicability when used monitoring tool running process.
منابع مشابه
Statistical Process Control of Multivariate Processes
With process computers routinely collecting measurements on large numbers of process variables, multivariate statistical methods for the analysis, monitoring and diagnosis of process operating performance have received increasing attention. Extensions of traditional univariate Shewhart, CUSUM and EWMA control charts to multivariate quality control situations are based on Hotelling's T 2 statist...
متن کاملStatistical process control of the stochastic complexity of discrete processes
Changes in stochastic processes often affect their description length, and reflected by their stochastic complexity measures. Monitoring the stochastic complexity of a sequence (or, equivalently, its code length) can detect process changes that may be undetectable by traditional SPC methods. The context tree is proposed here as a universal compression algorithm for measuring the stochastic comp...
متن کاملStatistical quality control in micro-manufacturing through multivariate ?-EWMA chart
Micro-manufacturing processes are characterized by high process variability and an increased significance of measurement uncertainty in relation to tight tolerance specifications. Therefore, an approach that separates the superposition of measurement and manufacturing variation is demanded. A novel design for a quality control chart that makes it possible to monitor, control and extract measure...
متن کاملIntegration of Multivariate Statistical Process Control and Engineering Process Control
Two independent methods for improving quality are engineering process control (EPC) and statistical process control (SPC). The first method tries to minimize variability by handling process variables so as to keep the outputs of the process on target. While the latter method, SPC does the same basic task of minimizing variability by supervising and eradicating the assignable causes of variation...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Journal of Intelligent Manufacturing
سال: 2023
ISSN: ['1572-8145', '0956-5515']
DOI: https://doi.org/10.1007/s10845-023-02156-7