Digitalization forces improved maintenance in shop-floor systems. Companies have begun to upgrade their existing production lines by equipping them with new machinery or sensors. This enables intelligent tracking and control of manufacturing activities. Simultaneously, the advancement of computing power enables complex analyses including the adaptation of machine learning algorithms to gain new knowledge. However, previous research has revealed that intelligent decision support systems are only applied successfully if they are comprehensible for employees within the factory. Therefore, we have developed a prototype based on a comprehensible set of rules for automated anomaly identification in real-time. We include employee’s expert knowledge from the very beginning to establish a sense of participation. This is improved and enhanced by techniques from the fields of process mining and machine learning. Thus, the prototype presents previously unknown error correlations in an understandable and descriptive way combining intelligent anomaly detection by a linked knowledge database system.