Data mining on sensor data of interlinked processes: Approach for autonomous automation of complex and interlinked processes by means of machine learning techniques

The principle of autonomous automation is a key method of lean production. It is used to reduce costs by eliminating waste in form of scrap and subsequent work. The application of automatic quality inspections is limited to processes of relatively low complexity. This paper presents an approach to realize autonomous automation in complex and interlinked processes referring to a milling process. This approach is based on machine learning techniques. The paper depicts challenges considering realtime data processing one the one hand and production-related constraints on the other hand. © Carl Hanser Verlag.