Comparative analysis of inline quality control system process capability with offline lab QC against product specification by MSPC method

Abstract

The similarity of values shown by the two methods of reading paper specifications, namely Quality Control System (QCS) and QC Lab (controlling offline in Quality Control lab) to the attention of PT Bukit Muria Jaya (BMJ). The QCS is an online control process while QC Lab is offline control process. The company wants to change the offline control process online, but there is no analysis to show whether the output given by both methods is the same. The study involved Grade A paper measured at once by five paper specification parameters, i.e. paper weight, porosity, CaCo3 content, opacity, and paper strength. However, after normality testing was found, the data is not multivariate normal distribution. Therefore, this study aims to answer the problem with a multivariate free distribution procedure. The study was conducted in two stages, firstly by looking at the median equality of both outputs of the paper specification reading method, and the second is by looking at the similarity of characteristic capability. A preliminary study was conducted to see whether the two methods of reading the Grade A paper specifications had the same median. This test is done by Kruskal Wallis Multivariate method. The results of the analysis received a null hypothesis which means the median of the two outputs generated QCS and QC Lab for each Grade A paper are equal. Furthermore, based on a capability comparison analysis where the value of this capability is derived from the sample without observation out of control, out of control process control is assisted in the range of T2 Hotelling Control Free Distribution. It can be concluded that Grade A paper capability of QCS does not show the same result as QC Lab, i.e. Porosity variable, where Porosity is capable in QC Lab but not QCS, so manual process control for Grade A paper cannot yet be replaced online.