Automating the resolution of flight conflicts: Deep reinforcement learning in service of air traffic controllers
This work addresses the need for more automated and transparent AI systems in safety-critical air traffic control to improve adoption and operational integrity.
The paper tackled the problem of automating flight conflict resolution in dense air traffic by proposing a graph convolutional reinforcement learning method in a multiagent setting, showing it can provide high-quality solutions that address operational transparency issues.
Dense and complex air traffic scenarios require higher levels of automation than those exhibited by tactical conflict detection and resolution (CD\&R) tools that air traffic controllers (ATCO) use today. However, the air traffic control (ATC) domain, being safety critical, requires AI systems to which operators are comfortable to relinquishing control, guaranteeing operational integrity and automation adoption. Two major factors towards this goal are quality of solutions, and transparency in decision making. This paper proposes using a graph convolutional reinforcement learning method operating in a multiagent setting where each agent (flight) performs a CD\&R task, jointly with other agents. We show that this method can provide high-quality solutions with respect to stakeholders interests (air traffic controllers and airspace users), addressing operational transparency issues.