This talk has been automatically captioned. You can remove these by pressing CC on the video toolbar.

Name: Su-Fen Yang

Talk Title: The PCA-based Control Charts for Monitoring Multiple-stream Processes

Abstract: The cosmetic bottle cover weights are collected from a filling machine with eight identical filling heads every 4 hours. The cover weights are a critical quality variable of the bottles. The weights produced from every filling head should be identical, independent, and satisfy the requested specification limits of customers. However, the variation of the weights from each filling head is large, and the weights distributions from each filling machine are correlated but not normal and identical. Hence, the multiple processes of the bottle weights are the kind of multiple-stream processes. There are some in-control and out-of-control data of the bottle weights are collected from the multiple-stream processes, and the data are analyzed. To monitor whether the processes are in-control or out-of-control, the principal component analysis (PCA)-based EWMA mean and covariance control charts are proposed to demonstrate the in-control samples and monitor the out-of-control samples. We found that the proposed PCA-based EWMA mean and covariance control charts perform well.

This talk is a contributed talk at EVA 2021.