Abstract
Nowadays, quality is essential in every field. Producing quality products or services can help to increase thecompetitiveness, reputation and customer loyalty of a company as this ensures that the products or servicesmeet customersβ expectations. However, for a product or service to have a satisfactory quality level, theunderlying process manufacturing the product or providing the service should be in statistical control. TheGroup Runs (GR) π chart is proposed as an improvement of the synthetic π chart for the detection ofprocess mean shifts. The GR π chart integrates the Shewhart π sub-chart and an extended version of theconforming run length (CRL) sub-chart. To have a better understanding of the behavior of the GR chart, inthis paper, we study the percentage points (or percentiles) of the run-length distribution of the GR chart,instead of depending entirely on the average run length (ARL) criterion. Before computing the percentiles ofthe GR π chart, we need to obtain the optimal parameters of the GR π chart. This can be achieved by writinga computer program to search for a set of optimal L and k values that minimize the out-of-control ARL,while maintaining the in-control ARL at a desired value. Then, the percentiles of the run-length distributionof the GR π chart can be investigated using this set of L and k values.
Original language | English |
---|---|
Pages (from-to) | 84-90 |
Number of pages | 7 |
Journal | CMS Bulletin |
Volume | 2019 |
Publication status | Published - 31 Dec 2019 |