Data mining techniques are widely used for anomaly detection in the commercial fixed-wing aviation domain to retrospectively improve operational safety. However, analyses of flight data records for anomaly detection in the rotorcraft domain are not as prevalent, potentially due to the lack of installed flight data recorders and vague definitions of phases of flight. In particular, the 2017 National Transportation Safety Board (NTSB) most wanted list recommended expanding the use of recorders in helicopters. This will probably best be tackled by helicopter operators and/or government agencies. In this research, we focus on the definition and identification of phases of flight for rotorcraft operations to pave the way for potential anomaly detection analyses. This paper first reviews the various definitions of phases of flight for rotorcraft and fixed-wing aircraft operations found in the literature. Then, some preliminary definitions for each phase of flight for helicopter operations are proposed and subject matter experts’ opinions are gathered to adjust the aforementioned definitions and determine feasible phases of flight sequences. A set of evaluation criteria is then developed to judge the quality of the results from the phases of flight identification and a methodology to identify phases of flight for various types of operations is presented. Several flight samples with pre-labeled phases of flight were tested using different phase detection techniques. Results show that the proposed regression-based method with sliding window outperforms other candidate methods. With the flight phases information available for rotorcraft operations, different perspectives of flight data records may be revealed and may further facilitate the detection of anomalies in the data.
Hsiang-Jui Chin, Alexia Payan, Charles Johnson, and Dimitri N. Mavris, “Phases of Flight Identification for Rotorcraft Operations", AIAA SciTech 2019 Forum, AIAA SciTech Forum, (AIAA 2019-0139), San Diego, CA, 7-11 Jan 2019, https://doi.org/10.2514/6.2019-0139