Online Simulation of Methods to Predict the Remaining Useful Lifetime of Aircraft Components

Authors

Daniel Azevedo
Bernardete Ribeiro
Alberto Cardoso

Abstract

This work explores the creation of an online tool where the user can simulate a Prognostics and Health Management (PHM) system, by creating and submitting a specific machine learning experimental scenario in order to predict the Remaining Useful Lifetime (RUL) of aircraft subsystems that were affected by a system fault. In this interface, the user can choose a public dataset from the proposed ones and a specific machine learning method to be applied to the dataset. After submitting the selected configuration the system runs it and the output, i.e., the predicted RUL, is presented in the form of a graph with the possibility of exporting the results in a .txt file. The suggested datasets are made of data retrieved from aircraft sensors and the proposed methods represent different alternatives for RUL prediction. There is also the possibility to choose more than one method and then graphically compare the results. Since the methods are executed remotely, the use of this tool is not computational demanding for the user. The main aim of this work is to create a simple and user-friendly interface, allowing the users to make their own experiences online, simulating a PHM system applied to a given dataset.

Keywords

Aircraft Maintenance, Machine Learning, Prognostics and Health Management, Remaining Useful Lifetime

Subject

Prognostics and Health Management of Aircraft Systems with online experimentation

Related Project

H2020-REMAP – Real-time Condition-based Maintenance for Adaptive Aircraft Maintenance Planning