Intelligent Flight-Data-Recorders; a Step Toward a New Generation of Learning Aircraft

To understand how aerial accidents occur, the installation of flight data recorders (FDR) and cockpit voice recorders (CVR) have become mandatory by law. However, such devices play a passive role and are used once an accident has occurred. Recent advances in machine learning techniques and their application in solving engineering problems are the keys to creating a more active role for both FDR as well as CVR. Here, we investigate a new approach to bringing intelligence to an FDR (called I-FDR). Through a continuous data-mining process, an I-FDR could bring better situational awareness to the flight crew. An I-FDR, similar to the FDR, records all pertinent flying parameters. In addition, proper data-mining and machine learning techniques bring more insight into the current state of the flight as well as its immediate future. Once the aircraft completes its mission, all recorded flight data is transferred to a data bank for learning purposes on future flights of a similar aircraft type. This study compares three unsupervised machine learning algorithms to examine their effectiveness in guiding flight crew in hazardous situations. Methods include DBSCAN (Density-based spatial) clustering, isolation forest, and LSTM (Long Short-Term Memory). This study uses simulated takeoff data sets and a runway excursion accident as the test case to determine which method is more accurate. The assessment shows that LSTM is superior in detecting anomalous data.