SHORT-RUN STATISTICAL PROCESS CONTROL FOR REAL-TIME MANUFACTURING-DESIGN, IMPLEMENTATION AND PERFORMANCE STUDY

Abstract

The goal of this dissertation is to design and implement a real-time short-run Statistical Process Control (SPC) method in low volume manufacturing to improve the manufactured product quality. There are many ways to achieve quality improvement. SPC is one of them which can be applied to any area within a company or organization, including manufacturing, process development, finance and accounting, marketing, and field service of products. SPC is a tool that applies basic statistics to control a process in terms of both precision and accuracy, Conventional SPC methods only address the process control problems when the target is known or the SPC parameters have been calculated in a high-volume manufacturing environment. The short-run SPC technology has been applied to adapt the conventional SPC methods to quickly and accurately detect the unknown short-run SPC parameters in a low-volume manufacturing environment. It is still a major challenge to researches and engineers. This dissertation reviews the basic SPC theory and algorithms and identifies the limitations of conventional SPC techniques. Based on these studies, the conventional SPC method is adapted to short-run SPC. In order to improve the performance of the short-run SPC, the idea of using dynamic SPC parameters to replace the static SPC parameters is implemented. Another idea to introduce a dynamic sampling group size to replace the static sampling group size in short-run SPC is academically proved in this dissertation. The performance of the short-run SPC in Motorola's Modular Cell Controller implementation is studied, This dissertation also studies the problems related to the realization of short-run SPC and proposes solutions for these problems.