WindBorne Secures Nearly $50 Million U.S. Air Force Contract for AI-Enhanced Weather Balloon Data

Image: TechCrunch · Source

WindBorne Systems, using AI and autonomous weather balloons, landed a major U.S. Air Force deal potentially worth $49 million for precise atmospheric data to improve weather forecasting.

WindBorne Systems, a Silicon Valley startup specializing in weather forecasting through AI and autonomous balloons, has been awarded a significant contract by the U.S. Air Force. The deal, potentially valued at $49 million, entrusts WindBorne to provide atmospheric data collected by its global fleet of approximately 700 autonomous weather balloons. These sensors gather empirical weather readings that feed directly into the company's deep-learning weather prediction models.

Founded in 2015 and headquartered in Redwood City, California, WindBorne has established collaborations with governmental bodies such as the National Weather Service, the Air Force, and the Navy. This new agreement marks the largest contract the company has secured to date. Under the contract framework, the Air Force will order data as needed, specifying particular regions and timeframes. To date, it has utilized $1.7 million of the contract with expectations of further orders.

Co-founder and CTO Andrey Sushko explained that the contract facilitates expansion of the balloon constellation while supporting research and development aimed at enhancing balloon hardware and performance. The contract was awarded without competitive bidding, reflecting governmental recognition that WindBorne uniquely satisfies the Air Force's operational data collection requirements.

WindBorne recently raised a $37 million Series B funding round, bolstering its growth and scaling aspirations. The company also released WeatherMesh 6, an advanced forecasting model that reportedly outperforms traditional physics-based methods such as those employed by the European Centre for Medium-Range Weather Forecasts (ECMWF).

The evolution of AI-driven weather forecasting has gained momentum with transformer architectures and voluminous datasets from ECMWF, enabling faster and sometimes more accurate predictions than conventional supercomputer-based simulations. WindBorne has contributed to this field notably by deploying unpowered balloons into extreme weather phenomena, including a historic mission to collect data from the eye of a typhoon. These balloons navigate by modulating altitude to catch directional winds, informed by real-time deep-learning models.

Such advances reflect a broader shift in meteorology towards integrating empirical sensor data with AI-enhanced modeling to improve operational weather forecasting capabilities, with WindBorne positioned at the forefront of this innovation.