Exploring AI support for air traffic controllers in conflict resolution

Artificial Intelligence is increasingly being explored for its potential to assist air traffic controllers (ATCOs) in real-time decision-making. But how effective can AI be in supporting ATCOs during conflict resolution, and how can we refine these systems for practical use?

As part of the SESAR JU JARVIS project, ENAC and Airbus Defence and Space recently conducted a session to validate "Air Traffic Control Digital Assistant ATC-DA) for Conflict Resolution", one of the three solutions that the project is developing. It assists air traffic controllers in different operations: integration of new entrants, growing interconnections, optimised environmentally friendly decisions, and overall efficiency of air transport. In addition, AI-based data-driven algorithms support operators in different roles and phases of the flight reducing workload for busy tasks.

Simulation Environment

The experiment took place in early March in a dedicated test room at ENAC, using the GENETICS Airbus Defence and Space platform. The setup included:

• DALIA radar interface for visualization and clearances
• TESLA Tactical Controller Tool (TCT) for conflict detection and resolution, developed by Airbus DS

Methodology

In the first session, each of the eight participating ATCO held a valid en-route radar license with over 10 years of experience and the pilot and adjacent sector coordination were automated.

During the 3 days of experimentation, participants performed traffic separation under different scenarios:

• Baseline Scenario: Using only the TCT for conflict detection
• Solution A: TCT + AI-powered resolution advisory, trained on real ATCO decision-making behaviour
• Solution B: TCT + resolution advisory based on expert judgment

While it's too early to communicate provisional outcomes, the tests are a critical step in understanding how digital assistants like the ATC-DA can realistically support controllers and be fine-tuned for operational use.

More about the project