Trends and Challenges of Machine Learning-Based Predictive Maintenance in Aviation Industry

Based on the airline maintenance cost executive commentary FY2020 data published by International Air Transport Association (IATA), aircraft maintenance accounts for 10.3% of airline operating costs, with approximately 3.3 million US$ spent per plane in 2019. Predictive maintenance has grown popularity in aerospace in the last decade due to the increasing availability of condition monitoring data for aircraft components and engines. This paper discusses the emerging trends in Machine Learning (ML)-based predictive maintenance in the aviation industry and explores the challenges in its application. The primary objective is to give a comprehensive view of the current state and potential directions in predictive maintenance by considering both machine learning advancements and the unique complexities of the aviation industry.