Bridging the Gap between Prediction and Action: Information Demands and Requirements for a Data-Based Decision Support System for Root Cause Analysis in Production

Root cause analysis (RCA) is essential for continuously improving manufacturing processes. Increasing product complexity and rising demands on product quality require advanced approaches to RCA in production. Predictive quality models can detect failures in manufacturing processes, but can only determine their causes to a limited extent. This publication therefore investigates information demands for a data-based decision support system for RCA that builds on and extends predictive quality models. For this purpose, a semi-structured expert survey was conducted comprising eleven hypotheses and two open-ended questions derived from established failure analysis methods and the theory of causal reasoning. The results indicate that additional information beyond prediction outcomes is necessary to identify failure causes effectively. Foremost, it is important to understand process parameters and their influence on the product. Additionally, the study identifies requirements for the decision support system, including model independence, stability and consideration of use case and target group. The results of the study provide valuable insights for improving RCA processes through enhanced information provision and system usability in production environments.