WindBorne wins up to $49M Air Force contract for balloon data

The U.S. Air Force will buy data collected by WindBorne Systems’ fleet of autonomous weather balloons, under a contract that could be worth nearly $50 million to the Redwood City, California-based startup. The deal lets the Air Force pay only as it needs data, placing orders for specific regions at specific times; the contract is worth up to $49 million, and so far the Air Force has ordered $1.7 million worth, with WindBorne executives expecting more orders to follow. It’s the single-largest contract WindBorne has participated in, and it was awarded without competitive bidding.

WindBorne co-founder and CTO Andrey Sushko said the award is largely a vehicle to support the fleet’s continued growth and to fund further work on the hardware and performance of the balloons themselves. He said the lack of competitive bidding reflects “a determination by the government that for this capability of collecting global data, we were the only company able to provide it in the way that met their needs for these operations.”

Founded in 2015, WindBorne now flies about 700 weather balloons around the globe, collecting atmospheric sensor readings it sells to customers and also feeds into its own AI weather forecasting model. The company has worked closely with the U.S. government, including the National Weather Service, the Air Force and the Navy. CEO John Dean told TechCrunch in August that WindBorne had raised a $37 million Series B round to keep scaling the business, and earlier this summer it released WeatherMesh 6, the latest version of its forecasting model, which the company says outperforms the physics-based model run by the European Centre for Medium-Range Weather Forecasts (ECMWF).

AI weather forecasting has become one of the more productive uses of deep-learning technology since the arrival of transformer model architectures and the ECMWF’s release of a massive weather dataset a few years ago. Models from frontier labs and government agencies are proving more accurate than traditional forecasts, though it remains a jagged frontier where both approaches are still needed — Google DeepMind’s latest weather model, for example, now has cyclone-forecasting tools that are becoming increasingly central to how the U.S. government responds to hurricane season.

In August, WindBorne became the first organization to send an unpowered flight system — one with no engine — to collect data from inside the eye of a typhoon, steering its balloons by adjusting altitude, bleeding air, or dropping ballast to ride winds toward their targets. Sushko called it a novel step made possible by accurate wind and weather models paired with a large enough balloon constellation to have vehicles in the right places, adding that it was one of the first times a deep-learning model was used to inform an actual operational decision.

Why it matters

A sole-source Air Force contract that could be worth nearly $50 million suggests the U.S. military now views WindBorne’s balloon-collected data as difficult to replace, bolstering the startup’s position as AI-driven forecasting increasingly competes with — and complements — traditional physics-based models in government use.