NATS Launches Project Bluebird to Test AI in Air Traffic Control as Traffic Growth Risks Crisis
Updated
Updated · Financial Times · Jun 30
NATS Launches Project Bluebird to Test AI in Air Traffic Control as Traffic Growth Risks Crisis
2 articles · Updated · Financial Times · Jun 30
Summary
Project Bluebird will use digital twins and machine learning on NATS historic data to model traffic scenarios, project aircraft trajectories and flag potential conflicts for controllers.
NATS says the system is meant to advise rather than replace staff, with its research head saying humans will still be needed to guide aircraft and that full machine control is not credible in their lifetimes.
Years of traffic growth after Covid-19 and the long, costly training pipeline for controllers are driving the push, as experts warn the industry faces many more aircraft with ground systems still resembling a 1962-era model.
Safety remains the main obstacle: regulators, controllers and passengers would need rigorous testing before adoption, while GPS jamming, closed airspace over Iran and Iraq, and backup-navigation failures underline the need for redundancy.
If AI manages the skies, can humans still prevent the next 'Miracle on the Hudson'?
With AI managing our skies, how are we protected from a catastrophic glitch or a crippling cyber-attack?
Project Bluebird: Rigorous Trials of Explainable AI for Human-in-the-Loop Air Traffic Management in the UK
Overview
Project Bluebird is entering a crucial stage in June 2026 with the start of live trials, aiming to bring Artificial Intelligence into UK airspace management. These trials build on extensive development and simulation work, using a sophisticated digital twin environment to safely test AI agents in real-world scenarios. The main goal is to collect important data and insights on how AI can improve the safety and efficiency of air traffic control. This effort is paving the way for more advanced, semi-automated applications in the industry, marking a significant step forward for aviation technology.