Monitoring of aircraft operation using statistics and machine learning

Abstract

This paper describes the use of statistics and machine learning techniques to monitor the performance of commercial aircraft operation. The purpose of this research is to develop methods that can be used to generate reliable and timely alerts so that engineers and fleet specialists become aware of abnormal situations in a large fleet of commercial aircraft that they manage. We introduce three approaches that we have used for monitoring engines and generating alerts. We also explain how additional information can be generated from machine learning experiments so that the parameters influencing the particular abnormal situation and their ranges are also identified and reported. Various benefits of fleet monitoring are explained in the paper.