Approach to Introducing a Statistical Quality Control

This paper describes some examples and points about implementing a statistical quality control in developing business systems in Sumitomo Electric Industries, Ltd. and Sumitomo Electric Information Systems Co., Ltd. Although the X-R chart is often used in statistical quality control, it is recommended to introduce the u-chart if defects are to be controlled in future. If a defect detection process such as a review or a test does not reach a statistical steady state that indicates a process is stable on the control chart, an appropriate control could be conducted by not only further standardizing the process, but also reviewing the definition of size indicators. Quality prediction can be made possible by accumulating defect data collected to create a control chart and analyzing the distributions of introduced defect densities at an organization level. If the accuracy of quality prediction is low, it can be enhanced by carrying out improvements to narrow the widths of distributions of introduced defect densities.