An efficient cloud prognostic approach for aircraft engines fleet trending

Abstract

ABSTRACTThe implementation of prognostics and health management solutions is becoming increasingly important. Many industrials are interested in this maintenance process, especially Predix by General Electric and EngineWise by Pratt & Whitney. Predix allows to create innovative industrial internet applications that transform operational data in real-time into actionable information in several domains (aviation, healthcare, …). EngineWise is a world leader in the design, manufacture and service of aircraft engines and auxiliary power units. The prognostic process is a main step to predict failures before they occur by determining remaining useful life (RUL) of equipment. However, it also poses challenges such as reliability, availability, infrastructure and physics servers. To address these challenges, this paper investigates a cloud-based prognostic system that defines an approach ‘Prognostic as a Service.’ This approach will provide an efficient prognostic solution in the cloud computing. In this paper, three data-driven algorithms, Artificial Neural Network, Neuro-Fuzzy System and Bayesian Network, are discussed and implemented to estimate the RUL. They are tested for aircraft engines fleet. Furthermore, and in order to test the efficiency of our solution, we have studied the performance of the prognostic system according to the accuracy, precision and mean squared error.