Abstract
Predictive maintenance has received considerable attention in the aviation industry where costs, system availability and reliability are major concerns. In spite of recent advances, effective health monitoring and prognostics for the scheduling of condition-based maintenance operations is still very challenging. The increasing availability of maintenance and operational data along with recent progress made in machine learning has boosted the development of data-driven prognostics and health management (PHM) models. In this paper, we describe the data workflow in place at an airline for the maintenance of an aircraft system and highlight the difficulties related to a proper labelling of the health status of such systems, resulting in a poor suitability of supervised learning techniques. We focus on investigating the feasibility and the potential of semi-supervised anomaly detection methods for the health monitoring of a real aircraft system. Proposed methods are evaluated on large volumes of real sensor data from a cooling unit system on a modern wide body aircraft from a major European airline. For the sake of confidentiality, data has been anonymized and only few technical and operational details about the system had been made available. We trained several deep neural network autoencoder architectures on nominal data and used the anomaly scores to calculate a health indicator. Results suggest that high anomaly scores are correlated with identified failures in the maintenance logs. Also, some situations see an increase in the anomaly score for several flights prior to the system’s failure, which paves a natural way for early fault identification.
Highlights
Prognostics and health management (PHM) has drawn growing interest from industrial and academic research in the last few years, especially in sectors like aviation where profit margins are small and operational costs are critical
In this sub-section, we briefly review progresses made in anomaly detection applied to aviation in general, before focusing in the subsections on data-driven methods adapted to the prognostics and health management (PHM) field
We can mention hybrid neural network approaches such in [24], where a Convolutional Neural Networks (CNN) is combined with a gated recurrent unit (GRU) Recurrent Neural Networks (RNN) in order to extract spatio-temporal features from sensor data and detect anomalies in rotating machinery
Summary
Prognostics and health management (PHM) has drawn growing interest from industrial and academic research in the last few years, especially in sectors like aviation where profit margins are small and operational costs are critical. Oehling et al [12] approach based on Local Outlier Probability (LoOP) [13] was applied to an airline dataset of 1.2 million flights in order to detect anomalies related to safety events Classical approaches such as clustering or distance-based methods like k-Nearest Neighbours (kNN) suffer from the curse of dimensionality issue when applied to highdimensional data. With large-scale high-complex data as the one generally available in aviation, deeplearning approaches are supposed to perform better than traditional machine learning methods [5]. They present significant limitations when applied to long sequences
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
