Search

Remaining useful life prediction models are a central aspect of developing modern and capable prognostics and health management systems. Recently, such models are increasingly data-driven and based on various machine learning techniques, in particular deep neural networks. Such models are notoriously “data hungry”, i.e., to get adequate performance of such models, a substantial amount of diverse training data is needed. However, in several domains in which one would like to deploy data-driven remaining useful life models, there is a lack of data or data are distributed among several actors. Often these actors, for various reasons, cannot share data among themselves. In this paper a method for collaborative training of remaining useful life models based on federated learning is presented. In this setting, actors do not need to share locally held secret data, only model updates. Model updates are aggregated by a central server, and subsequently sent back to each of the clients, until convergence. There are numerous strategies for aggregating clients’ model updates and in this paper two strategies will be explored: 1) federated averaging and 2) federated learning with personalization layers. Federated averaging is the common baseline federated learning strategy where the clients’ models are averaged by the central server to update the global model. Federated averaging has been shown to have a limited ability to deal with non-identically and independently distributed data. To mitigate this problem, federated learning with personalization layers, a strategy similar to federated averaging but where each client is allowed to append custom layers to their local model, is explored. The two federated learning strategies will be evaluated on two datasets: 1) run-to-failure trajectories from power cycling of silicon-carbide metal-oxide semiconductor field-effect transistors, and 2) C-MAPSS, a well-known simulated dataset of turbofan jet engines. Two neural network model architectures commonly used in remaining useful life prediction, long short-term memory with multi-layer perceptron feature extractors, and convolutional gated recurrent unit, will be used for the evaluation. It is shown that similar or better performance is achieved when using federated learning compared to when the model is only trained on local data.

The 21st century aviation and aerospace technologies have evolved and become more complex and technical. Turbofan jet engines as well as their cousins, the rocket engines (liquid/solid) have gone through several design upgrades and enhancements during the course of their design and exploitation. These technological upgrades have made engines very complex and expensive machines which need constant monitoring during their working phase. As the demand and use of such engines is growing steadily, both in the civilian and military sectors, it becomes necessary to monitor and predict the behavior of parametric data generated by these complex engines during their working phases. In this paper flight parameters such as Exhaust Gas Temperature (EGT), Engine Fan Speeds (N1 and N2), Fuel Flow (FF), Oil Temperature (OT), Oil Pressure (OP), Vibration and others where used to determine engine fault. All turbo fan engines go through several distinctly different working phases: Take-off phase, Cruise phase and Landing phase. Recording generated parametric data during these different phases leads to a massive amount of in-flight data and maintenance reports, which makes the task of designing and developing a fault diagnostic system highly challenging. It becomes imperative to use modern techniques in data analysis that can handle large volumes of generated data and provide clear visual results for determining the technical status of the engine under investigation/monitoring. These modern techniques should be able to give clear and objective assessment of the object under investigation. Cluster analysis methods based on Neural Networks such as c-means, k-means, self-organizing maps and DBSCAN algorithm have been used to build clusters. Differences in cluster groupings/patterns between healthy engine and engine with degraded performance are compared and used as the bases for defining faults. Fault diagnosis plays a crucial role in aircraft engine management. Timely and accurate detection of faults is the foundation on which maintenance turnaround times, operational costs and flight safety are based. The data used in this paper for analysis was obtained from flight data recorder during one flight cycle. The final decision on a fault is taken by an engineer.