Phase I Monitoring With Applications In Manufacturing And Healthcare

Abstract

This research develops statistical methods for quality monitoring in complex systems. Quality monitoring typically consists of two phases called Phase I analysis (or offline monitoring) and Phase II analysis (or online monitoring). This research is focused on Phase I monitoring. Two application areas are considered, complex manufacturing processes and healthcare delivery processes. In the first application, a robust strategy for Phase I analysis of optical profiles in low-E glass manufacturing is developed. The proposed approach aims to solve the problems such as violation of normality, high dimensionality, detection of multiple change points, etc. It will provide a convenient process monitoring tool for practitioners in the low-E glass industry. In the second application, a systematic methodology for Phase I monitoring of patient readmission is developed. Patient readmission is a critical contributor to the rising health care costs and has become an important performance indicator for assessing and monitoring quality of care. This work consists of two parts: construction of readmission model and change detection based on the constructed model. The proposed approach is demonstrated using real data from chronic obstructive pulmonary diseases (COPD) patients.