ATMAS: Airplane Trajectory Missing Alarm System based on Deep Learning

With the rapid development of civil aviation, more emphasis has been placed on airplane safety. However, the traditional method of dealing with the airplane vanishing from the radar is to rely on the controllers to contact the pilot when the disappearances occur, which brings a safety risk owing to the time delay. This paper proposes an automatic alarm system called Airplane Trajectory Missing Alarm System (ATMAS). ATMAS is made up of two components: a Long Short-Term Memory (LSTM) neural network and a Multi-Layer Perceptron (MLP). LSTM extracts the semantic context in trajectory, and MLP determines whether to alarm based on the context. By applying the airplane trajectory in Qingdao controlled airspace as a case study, ATMAS takes the records in the last minute as input and indicates whether the airplane will vanish from the radar in the following minute. The accuracy of the alarm reaches 90.15%.