Identifying Product and Process State Drivers in Manufacturing Systems Using Supervised Machine Learning

Mehr zum Buch

A novel approach for identifying state drivers in complex, multi-stage manufacturing systems is presented, utilizing diverse and high-dimensional data sets. The method integrates process interrelations and has been validated through scenarios in aviation, chemical, and semiconductor industries. By employing SVM-based feature ranking, it successfully identifies key process parameters and state characteristics, enhancing quality monitoring and advanced process control. Notably, this method is versatile, applicable across various manufacturing processes and quality concepts.