A multi-sensor approach for failure identification during production enabled by parallel data monitoring
Multi-Sensor-Ansatz mit parallelisierter Datenauswertung zur Fehleridentifikation unter Produktionsbedingungen
Modern production is becoming more complex and diverse, while requirements for reliability, consistency, sustainability and quality are significantly increasing. For automated inspection and control, sensors with ever-higher accuracy are required. Furthermore, data from multiple sensors have to be simultaneously collected and analyzed in order to identify the state of production and possible failures. Thus, the acquisition of relevant information and data from various sensors is a difficult task. A novel framework is proposed for parallelized multi-sensor monitoring which enables an automatic distribution of complex data processing tasks to multi-core computing hardware. The framework allows the subdivision of complex data processing problems into simple parallel entities. This is demonstrated by using a multi-camera system for failure monitoring of sheet metal parts. As a result, benefits could be achieved in terms of minimum detectable failure sizes and inspection speed, enabling 100% inline inspection of produced parts.