Risk analyses lose acceptance among employees due to a lack of objectivity. However, the revision of the quality management standard ISO 9001 requires a risk-based decision-making approach in quality management. Data-based approaches are currently used reactively in networked productions, and data from different IT systems is rarely merged. As a result, valuable synergy effects are lost, redundant data sets are available and reliable, predictive data use is prevented.
The aim of "quadrika" is to determine a data-based key figure for predictive risk management. For this purpose, a system-independent software tool is being developed which combines the data from all production IT systems (e.g. CAQ, MES) and analyses them in accordance with standards. This is followed by the implementation and validation of the software.
The risk management model developed in "quadrika" consists of three parts: Process analysis, scenario-based FMEA as well as risk analysis and management. The selection and use of machine learning algorithms allows pattern recognition in process analysis and scenario-based FMEA. The software tool "Quality Data Module" is validated using an injection molding process. The increase in production quality is achieved by using merged data from different software systems.
German Aerospace Center (Deutsches Zentrum für Luft- und Raumfahrt e. V.)