Approach to identify product and process state drivers in manufacturing systems using supervised machine learning

Abstract

The developed concept allows identifying relevant state drivers of complex, multi-stage manufacturing systems holistically. It is able to utilize complex, diverse and high-dimensional data sets which often occur in manufacturing applications and integrate the important process intra- and inter-relations. The evaluation was conducted by using three different scenarios from distinctive manufacturing domains (aviation, chemical and semiconductor). The evaluation confirmed that it is possible to incorporate implicit process intra- and inter-relations on process as well as programme level through applying SVM based feature ranking. The analysis outcome presents a direct benefit for practitioners in form of the most important process parameters and state characteristics, so-called state drivers, of a manufacturing system. Given the increasing availability of data and information, this selection support can be directly utilized in, e.g., quality monitoring and advanced process control.