Rising requirements for converting machines lead to more complex processes. Working with biogenic materials results in ever changing properties and so breakdowns are unavoidable and occur frequently. Many plant operators lack of a profound knowledge of these processes and remediate the impact of a breakdown and not the cause. To give operators hints on lasting elimination the cause we proclaim a self-learning assistance system providing information based on the current situation. To detect situations, unknown to the PLC, machine learning algorithms for pattern mining on field level sensors are used.
