Best Practice Story 4 – Anomaly Detection

Brief description:

The fourth best practice story focuses on anomaly detection for the optimization of automated manufacturing processes for plastics processing. The focus is divided into three main groups, the data collection and data analysis, the processing of the obtained data and the implementation of the anomaly detection. For this purpose, individual manufacturing processes were recorded and documented in event logs. Due to the high variance of the components to be manufactured, the data to be collected was defined in detail. Modern machines already have digital interfaces for connection to networks. Here, it was determined how process data acquisition can be carried out. Furthermore, it was defined which data should be recorded (tool life, tool maintenance intervals, tool loading, machining time, etc.). Finally, competence development requirements could be estimated.

Expected results:

• Improvement of the clarity of the process landscape
• Potential IoT applications (customer portals, order status)
• Optimization of order processing (planning of design, mechanical production, assembly and run-in tests)
• Automated anomaly detection in welding processes