This document provides a summary report of a research study combining multivariate control charts and neural networks to estimate the starting time of process disturbances. The study was conducted by graduate students at Fu Jen Catholic University under the guidance of Dr. Shao Yuehjen. The research aims to address the limitations of traditional univariate and multivariate control charts in identifying the starting point of process faults by leveraging the computational abilities of neural networks. The study develops theoretical frameworks for combining multivariate control charts with neural networks and conducts simulation experiments to demonstrate the advantages of the proposed approach over using control charts alone. The results show the combined method can more accurately and earlier detect the starting fault point, helping reduce costs and resume normal operations.