Architecture of an AI-based aircraft maintenance defect management system

Kurzfassung

Aircraft maintenance is a mission-critical domain where precision, safety, and compliance are absolute priorities. The primary goal of aircraft MRO (Maintenance, Repair and Overhaul) is to minimize downtimes while ensuring while ensuring airworthiness through rigorous compliance with industry standards. In order to achieve this, it is essential to detect, identify and rectify defects during routine inspections and to predict further defects based on previously detected stages. For this, methods are being developed to predict further damage based on historical data on damage patterns, existing maintenance records, and anticipated aircraft usage. The integration of advanced AI technologies—specifically ontologies for semantic reasoning, structured knowledge graphs, Retrieval-Augmented Generation (RAG), and Large Language Models (LLMs)—presents unprecedented opportunities to transform this field. This paper will present an architecture that leverages these innovations to create a next-generation AI-driven aircraft maintenance defect management system, enhancing decision-making precision and operational resilience in aircraft maintenance workflows.