Challenges and Opportunities in Flight Data Mining: A Review of the State of the Art
Alek Gavrilovski, Hernando Jiménez, Dimitri N. Mavris, Arjun H. Rao, Sanghyun Shin, Inseok Hwang, Karen B. Marais · 2016
Incident and accident rates of rotorcraft and fixed-wing general aviation operations are considerably higher than those of commercial aviation. Efforts to improve the safety record of the former have drawn attention on flight data monitoring (FDM), also referred to as flight operations quality assurance (FOQA), as a voluntary safety mechanism for data collection with on-board recorders, analysis, and monitoring for the effectiveness of corrective measures. Flight data monitoring is regularly implemented in commercial operations where the application of data mining techniques has been proposed and demonstrated for anomaly detection. This paper surveys data mining techniques aimed at the reader with little or no prior exposure to this topic, as well as a review of published work on applications to flight data. It then provides an examination of challenges and opportunities for its application in rotorcraft and general aviation operations, where it has yet to be demonstrated on a large scale.