Automated Air Traffic Control (ATC)

Air Traffic Control (ATC) has received little attention as part of military simulation environments. Where ATC is played, it is typically only in a limited way, reducing the realism and validity of the results, and using human controllers, thereby increasing the cost of running the exercise. To address these needs, we propose the research and development of an automated air traffic control network appliance, called AutoATC, that can perform ATC tasks in simulation environments by interacting with human and synthetic pilots in natural ways, using standard military communications protocols. The AutoATC capability can be tailored to specific simulation exercises and environments, by allowing the user to switch between standard network protocols, and by providing information particular to the exercise’s scenario before running or by passing dynamic updates while the exercise is in progress. Based on Phase I research in ATC task and language analysis, we propose to use a cognitive systems approach to developing a human behavior model of the ATC tasks, which can be used in different configurations for the needs of particular exercises. By taking a human-like cognitive systems approach, the behavior of the system can be made explainable and, therefore, more trusted and used.