Fujitsu wins first prize for predictive maintenance in Airbus AI challenge
Winning solution identifies when sensors are functioning unusually and shows early warnings for vehicle faults
12 December 2019 | 0
Airbus SE has awarded Fujitsu first prize in its competition, Airbus AI Gym. The global aerospace manufacturer’s competition was devised to pinpoint the most accurate unsupervised predictive artificial intelligence (AI) capability for helicopters.
Flight engineers attach large numbers of sensors onto test helicopters to capture every hint of behaviour. With vast amounts of data, early-warning signals can be difficult to detect. Airbus’ challenge was designed to support research into accurately locating issues – especially data outliers.
Fujitsu came out on top of 140 participants for developing a new way to use unsupervised AI to detect anomalies in accelerometer data from Airbus pre-certification helicopters. By leveraging the ‘DeepTAN’ Unsupervised AI Model created by its sub-division, Fujitsu Systems Europe (FSE), the solution achieved 93% precision.
The lauded design took data sequences from multiple sensors and analysed them across a fixed time period, detecting abnormal sensor behaviour using a deep learning algorithm based on Multivariate Anomaly Detection with Generative Adversarial Networks (MAD-GEN). FSE both trained and validated the algorithm using 1,677 one-minute-sequences of accelerometer data from test helicopters flying at various locations, angles and flights.
“Winning first prize in this data challenge not only underlines Fujitsu’s world-leading AI expertise and technologies – it also provides concrete evidence of our ability to apply them to real-world business scenarios, said Ian Godfrey, director solutions business at FSE. “The concepts we applied to this specific problem have shown us how these new deep learning techniques not only help manufacturers but the firms working to sustain aircraft in service as well.”