Developing and Qualifying an ML Application for MRO Assistance

Abstract This study presents a framework for integrating and qualifying Machine Learning (ML) in Maintenance, Repair, and Overhaul (MRO) processes for gas turbines. Using neural networks for damage detection and decision trees for repair estimation, it emphasizes continuous qualification aligned with ISO/IEC standards and responsible AI principles. An interactive guide supports systematic ML implementation, ensuring transparency and compliance with Industry 4.0. Validated through two turbine blade case studies, the approach addresses key qualification steps, ensuring reliability in ML-assisted workflows. The study highlights the need for refined ML qualification standards to adapt to evolving AI technologies and regulations.