Conclusions

Abstract

This book has presented a set of intelligent diagnostic methods to reduce the dependence of board-level diagnosis on time-consuming and ineffective human effort. Multiple machine learning and statistical methods have been studied and adapted for diagnosis. Substantial improvement has been achieved over currently deployed diagnostic software. These solutions are not limited to a particular product; they are generic and can therefore be applied to various products. Although the goal of this book was to advance board-level diagnosis, the core techniques described in this book can also be leveraged for larger electronic systems.