In principle, ATFM and ATC layers could result in conflicting strategies. The proposed AI digital assistant will not only ensure that suggested solutions are compatible with short-term conflict alert and traffic collision avoidance system (TCAS) logic, but will also identify resilient ATFM and ATC policies that aim to maximise long-term benefits. The AI technique selected to achieve these long-term rewards is deep reinforcement learning (RL).

The digital assistant, based on deep reinforcement learning, will ensure that proposed solutions are compatible with short-term conflict alert and TCAS logic, while also minimising on-ground ATFM measures by accounting for environmental uncertainty and dynamics. These tasks will be performed collaboratively through a human–AI teaming approach.

BENEFITS

• Increased airspace capacity and operational productivity

• Improved flight adherence

• Reduced controller workload