Situational Anomaly Detection Using Multi-agent Trajectory Prediction for Terminal Airspace Operations

Most existing methods for trajectory anomaly detection in the air traffic domain have mainly assumed each trajectory to be independent. However, an aircraft’s trajectory is dependent on the trajectories of nearby aircraft, especially in a terminal airspace, as air traffic controllers actively alter the aircraft’s trajectory based on traffic density, separation, and scheduling to maintain the safety and efficiency of aircraft operations. To capture the interaction between flights and find anomalies in relation to the situation, we propose a situational anomaly detection framework based on multi-agent trajectory distribution prediction with an agent-aware attention mechanism. The situational anomalies are defined based on the predicted distribution of the trajectory. The proposed framework is demonstrated with real air traffic surveillance data recorded at Incheon International Airport, South Korea to show its effectiveness in identifying operationally significant anomalies.