Overview: On September 1st, according to a headline on the NATS website, the UK’s air traffic management service provider, NATS, announced the official launch of “Operation Blue Skies.” This is a landmark trial being conducted in the oceanic airspace over the Atlantic, designed to test an AI-driven contrail avoidance solution with the aim of reducing carbon emissions from air travel. According to Aerospace America, a publication of the American Institute of Aeronautics and Astronautics, the trial is being jointly conducted by NATS and the partners of the “Bluebird” project, namely the Alan Turing Institute in London and the University of Exeter.
According to Aerospace America, all parties involved in this collaboration have set a specific milestone: in a 2026 experiment, an AI agent will be fed with real-time data, including aircraft location and flight path information, with its information range being comparable to that available to human air traffic controllers. The agent will then send instructions to digitally represented aircraft within a digital twin airspace. The final step of the experiment will involve comparing the performance of this AI agent with the actual performance of human air traffic controllers in the Swanwick operations center. In other words, the AI will not interact with real flights for the time being, but will first “compete” with human controllers in a digital airspace, and the results will be used for comparison.
According to NATS’s official website, the two key terms for this trial are “overwater area” and “contrail avoidance.” The idea behind contrail avoidance is to adjust flight paths and avoid airspace where contrails are likely to form, thereby reducing the impact of aviation activities on the climate. Conducting this type of trial on transatlantic routes also provides a relatively independent scenario for verifying AI-assisted flight path decision-making. NATS calls it a “milestone” overwater airspace trial, and the specific trial results and subsequent arrangements have not yet been announced as of the time of writing.
It is worth noting that NATS is not solely focused on a single approach when it comes to AI. According to Aerospace America, its collaboration with the Turing Institute and the University of Exeter is also part of the "Bluebird" project, which focuses on testing AI applications using real-world air traffic management tasks, rather than simply conducting general concept verification.
Public acceptance is a crucial factor in the integration of AI into air traffic management. According to NATS’s 2026 Aviation Index survey, over three-quarters (76%) of respondents in the UK expressed discomfort with the idea of AI managing air traffic; in contrast, the public is most comfortable with AI taking on information-providing roles. This trend in public opinion also explains the progression of the “Blue Sky Initiative”: it begins with a focus on single-task scenarios, followed by validation in a digital twin environment, and then gradually moves towards actual implementation.
AI is entering air traffic control, and the challenges are half technical and half related to trust. The key aspect of NATS's recent actions lies in the sequence: first, they define the scenario (oceanic airspace, contrail avoidance), then they design a measurable, controlled experiment, turning the question of "whether AI or human controllers perform better" into a problem with a definitive answer, rather than just a slogan. Reducing carbon emissions is one of the few issues with broad consensus in the aviation industry, and by focusing on contrail avoidance, they have found a relatively uncontroversial entry point for AI in air traffic control. The results of the comparative experiment in 2026 will be worth monitoring closely.