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

The U.S. Air Force will buy data collected by WindBorne Systems' fleet of autonomous weather balloons under a new contract that could be worth nearly $50 million to the startup. The contract is worth up to $49 million, but the Air Force pays only as it needs data, and it can place orders for specific regions at specific times. So far it has ordered $1.7 million worth, and WindBorne executives expect more orders to follow.

WindBorne, founded in 2015 and based in Redwood City, California, currently has about 700 weather balloons flying around the globe. They collect atmospheric sensor readings that the company sells to customers and also feeds into its own AI weather forecasting model. According to TechCrunch, WindBorne is one of the first organizations to directly feed empirical data into a deep-learning weather model. The company has worked closely with the U.S. government, particularly the National Weather Service and the U.S. Air Force and Navy, but this is the single-largest contract it has participated in.

Co-founder and CTO Andrey Sushko told TechCrunch the contract is largely a vehicle to support continued growth of the balloon constellation, along with some research and development to improve the hardware and performance of the balloons. He added that the contract was awarded without competitive bidding, which he described as a determination by the government that for collecting global data, WindBorne was the only company able to provide it in the way that met the needs of these operations.

The article adds background on the company. CEO John Dean told TechCrunch in August that WindBorne had raised a $37 million Series B round to keep scaling. Earlier this summer the company released the latest version of its forecasting model, WeatherMesh 6, which it says gives more accurate forecasts than the traditional model run by the European Centre for Medium-Range Weather Forecasts (ECMWF), which simulates the atmosphere using the laws of physics.

TechCrunch frames this against a wider shift. Since transformer architectures arrived and the ECMWF released a large weather dataset a few years ago, AI weather forecasting has become one of the more productive uses of deep learning. Meteorologists previously relied on expensive supercomputers running detailed equations of how the atmosphere behaves; now end-to-end learning can ingest decades of weather data and issue predictions more quickly. Models from frontier labs and government agencies are proving more accurate than traditional forecasts, though it remains a jagged frontier, with both types of forecasts required for a useful understanding of the weather. Google DeepMind's latest weather model offers important new capabilities, and its cyclone forecasting tools 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 in the eye of a typhoon. The company steers its balloons by adjusting altitude, bleeding air, or dropping sand carried as ballast, sending them into winds that carry them where they need to go. Sushko called that a pretty novel step and named two ingredients: good wind and weather models plus the skill to navigate them, and a constellation large enough to have balloons in the right places. In his words, it is one of the first times a deep-learning model was used for actual operational decisions.

Key facts

• The U.S. Air Force will buy data from WindBorne's weather balloons under a contract worth up to $49 million; the Air Force pays only as it needs data and has ordered $1.7 million worth so far.
• It is the single-largest contract WindBorne has participated in, and per CTO Andrey Sushko it was awarded without competitive bidding.
• WindBorne has about 700 balloons flying globally, and feeds their readings into its own AI forecasting model, WeatherMesh 6.
• Sushko says the contract mainly funds growth of the balloon constellation plus research and development on balloon hardware and performance.
• In August WindBorne became the first organization to send an unpowered flight system into the eye of a typhoon, steering balloons by altitude changes, venting air and dropping sand ballast.

Why it matters

For WindBorne this is the biggest contract of its history, and it comes from a customer it already worked with. Sushko reads the absence of competitive bidding as the government's judgment that no other company could collect global data the way these operations required. The story also shows empirical balloon data feeding a deep-learning weather model, which Sushko describes as one of the first times such a model has been used for actual operational decisions.

Who it affects

WindBorne is the direct beneficiary, and the U.S. Air Force is the buyer. WindBorne has also worked closely with the National Weather Service and the Navy. More broadly, the story touches the weather forecasting field, where AI models from frontier labs and government agencies are being weighed against traditional physics-based forecasts.

How to use it

This is a government data purchase, not a product launch. The Air Force pays only as it needs data and can order specific regions at specific times, so usage is on-demand rather than a fixed flow. Sushko says the contract is a vehicle for growing the balloon constellation and for research and development on the balloons.

How solid is it

The figures and statements about the contract come from WindBorne, reported by TechCrunch: the up-to-$49 million ceiling, the $1.7 million ordered so far, and the account of the no-bid award. No Air Force or government spokesperson is quoted. The claim that WeatherMesh 6 is more accurate than the ECMWF model is WindBorne's own ("which it says").

Risks and caveats

The $49 million is a ceiling, not a committed amount, and the source does not say how much will actually be spent. The expectation of further orders comes from WindBorne executives. No contract duration, start date or award date is given, nor the period in which the $1.7 million was ordered. The source gives no accuracy metrics for WeatherMesh 6 against the ECMWF model. TechCrunch also notes that traditional and AI forecasts are both still required for a useful picture of the weather.

“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.”

— Andrey Sushko, WindBorne co-founder and CTO, to TechCrunch, on the no-bid award