UK and Google Launch AI Trial to Reduce Aviation Contrails

The UK government and Google are launching a 30 month trial called Operation Blue Skies to determine if artificial intelligence can help aircraft avoid creating climate warming contrails. The project will focus on the Shanwick Oceanic airspace in the eastern North Atlantic corridor, an area that accounts for approximately 5 percent of global contrail warming. By combining Met Office weather forecasts, satellite images, and flight data, the AI will predict where persistent contrails are likely to form and suggest altitude or route changes to air traffic controllers. These adjustments, which may involve changing altitude by around 2,000 feet, will be tested during selected evenings and nights in the winters of 2026 to 2027 and 2027 to 2028.

Contrails form when hot engine exhaust meets cold air at high altitudes. While many disappear quickly, some persist and spread into clouds that trap heat. Dr. Paul Hodgson of Google stated that contrail warming may be responsible for about a third of all aviation climate warming. The Department for Transport is contributing 2.65 million pounds to the project, while Google is providing 1.4 million pounds in research and infrastructure. Other partners include the University of Cambridge, Imperial College London, and National Air Traffic Services.

A primary concern for the trial is the trade off between reducing contrails and increasing fuel consumption. Changing a flight path can burn more fuel, which increases carbon dioxide emissions. The trial aims to establish if the climate benefit of avoiding contrails outweighs the cost of additional fuel. Aviation Minister Keir Mather described the project as a world first led by British ingenuity.

Both sources, which share a center lean, frame the event as a practical technological solution to an environmental problem. However, one source adds a critical perspective by noting that the trial occurs alongside government backed airport expansion and highlighting the high energy demands of the data centers that power AI.