Practical Framework for Advanced Quality-based Process Control in Interlinked Manufacturing Processes

As the economic manufacturing of high-quality products becomes an increasingly crucial competitive factor, corresponding quality assurance measures are gaining a growing interest. Even though research interest and industrial demand are both high, there is a large gap between methological approaches and practical applicability that needs to be closed. In this paper we therefore present a practical framework for advanced quality-based process control (AQPC) in interlinked manufacturing processes. Machine learning algorithms are used to predict the expected product quality based on recorded process parameters. That information then serves as an input for the derivation of optimal control decisions. Therefore, we formulate a mathematical optimization model including different options such as order reassignment and process parameter adaption to determine an optimal set of control decisions. We then break down the optimization into a gradual procedure that allows an application-specific integration into manufacturing.