WindBorne Wins Hefty US Air Force Contract for Balloon Data

The U.S. Air Force has entered into a significant new contract with WindBorne Systems to purchase data collected by the startup's fleet of autonomous weather balloons. The agreement could be worth nearly $50 million, marking a major milestone for the Redwood City, California-based company founded in 2015. WindBorne currently maintains approximately 700 weather balloons circling the globe, gathering crucial atmospheric sensor readings that are both sold to commercial and governmental customers and fed directly into the company's proprietary artificial intelligence weather forecasting models. This direct integration of empirical data into deep-learning weather systems represents an innovative approach in modern meteorology.

Although WindBorne has previously collaborated with U.S. government entities including the National Weather Service, the Navy, and the Air Force, this agreement is by far the largest contract the organization has engaged in. The financial structure of the contract establishes a ceiling of up to $49 million, but the Air Force operates on a pay-as-you-need model. Rather than paying the full amount upfront, the military branch can place targeted orders for specific geographic regions and timeframes. Initial orders under this agreement have amounted to $1.7 million, with company executives anticipating steady additional purchases as operational requirements evolve over time.

According to Andrey Sushko, WindBorne co-founder and chief technology officer, the contract serves primarily as a funding vehicle to support the continued growth of the company's balloon constellation. It will also finance ongoing research and development initiatives aimed at improving the underlying hardware and flight performance of the balloons. The contract was awarded through a non-competitive bidding process. Sushko noted that this decision reflects a determination by the government that WindBorne is currently the only entity capable of providing global data collection capabilities in a manner that precisely satisfies their operational needs.

This government contract follows a series of notable milestones for the startup. In August, WindBorne announced a $37 million Series B funding round to accelerate business scaling. Earlier in the summer, the company released the latest iteration of its forecasting model, WeatherMesh 6. WindBorne asserts that this model outperforms traditional physics-based simulation models, such as those run by the European Centre for Medium-Range Weather Forecasts. The convergence of transformer model architectures and expansive historical weather datasets has transformed weather prediction into one of the most practical applications of deep-learning technology.

Traditionally, meteorologists depended heavily on expensive supercomputers executing complex physical equations to simulate atmospheric behavior. In contrast, modern end-to-end machine learning models ingest decades of historical observations to generate rapid forecasts. While AI forecasting remains an evolving field where traditional methods still provide essential validation, models from private labs and government agencies are increasingly demonstrating superior accuracy. For instance, Google DeepMind models and specialized cyclone forecasting tools are becoming integral to emergency response efforts during severe weather seasons.

WindBorne has also demonstrated unique operational capabilities in extreme weather environments. In August, the company became the first organization to guide an unpowered flight system directly into the eye of a typhoon, specifically Typhoon Dolphin. By dynamically adjusting altitude, venting air, or releasing ballast sand, the company's autonomous balloons can navigate complex meteorological phenomena. As the U.S. military and civilian agencies increasingly integrate AI tools and novel data streams into their operations, WindBorne's balloon constellation positions the startup at the forefront of modern meteorological technology.