Quality improvement and control based on defect reduction

Abstract

This thesis addresses the quality improvement in a printing process at a food packaging company now experiencing hundreds of printing defects. Methodologies of Define, Measure, Analyze, Improve, and Control (DMAIC), and Response Surface Model were introduced to reduce the defect rate and control the process. As a result, critical inputs were identified, and a statistical regression model was constructed to predict the flaw size by knowing the critical inputs of the process. The mathematical optimal settings were determined to minimize the flaw size. Moreover, advanced control charts were developed to monitor and control the process.