WSU livestock management platform named Microsoft ‘AI for Good’ winner

A livestock grazing management tool developed by researchers at Washington State University, the University of Arizona, and the U.S. Forest Service is one of 20 projects to receive recognition and support through Microsoft's $5 million "AI for Good" program.

Launched in early 2025, AI For Good supports academic institutions, individuals, startups, and other Washington state organizations that use AI to promote sustainability, public health, and human rights. As a winner, StockSmart will receive $100,000 in Microsoft Azure cloud computing credits, and its creators will have an opportunity to work with AI for Good Lab scientists.

"I'm excited to use high-level computing technology to explore further applications of StockSmart," said Sonia Hall, WSU agricultural climate resilience specialist and StockSmart project co-leader. "In addition to AI, technologies like virtual fencing open new avenues for informing grazing decisions via more finely tuned information and the ability to respond in real time. The value of rangelands is sometimes unappreciated, so it's a great opportunity to contribute to their sustainable management."

Using detailed, remotely sensed geospatial data with user-provided fences and water locations, StockSmart helps livestock owners and land agency managers in the western U.S. determine how much forage is available to their livestock, and where.

The researchers hope the tool can soon be utilized for other rangeland health benefits like wildfire risk mitigation, said project co-leader and WSU Rangeland and Livestock Extension Specialist Tip Hudson.

Wildfires spread quickly in Washington state's low-elevation grassland and shrub-steppe landscapes, and mitigation via herbicides and mechanical thinning can be costly. A reduction in fine fuels like dried grass, twigs, leaves, and pine needles could change fire behavior, potentially lowering the carbon footprint of wildfires and decreasing the amount of public funds spent on their suppression.

That's where the StockSmart-Microsoft collaboration comes in. StockSmart's creators plan to leverage the tech giant's machine learning and AI expertise to analyze satellite and virtual fence animal sensor data. The results will help them determine specific regions where targeted grazing could help with fine fuels reduction.

"We'd like to use Microsoft's AI capabilities to improve StockSmart's fire behavior and prediction models," Hudson said. "We hope to combine fire behavior models with forage calculations to see where a grazed fire break would impact fire behavior like flame length, height, and temperature. By seeing how many livestock could be sustained for how long, we'll be able to determine how many animals would need to graze in a certain area to achieve the target level of fuel reduction."

The Microsoft partnership also coincides with the recent debut of StockSmart 2.0. Users can now select more comprehensive year ranges for underlying forage production data; they also have additional options for visualizing forage density and volume, among other improvements.

"On large landscapes, sustainable grazing means ensuring that forage harvest doesn't exceed forage supply," Hudson said. "The West has a tremendous diversity of soils, plant communities, and landscapes. One acre of land might have 1,000 pounds of forage, while 20 steps away, that can decrease to 200 pounds."

Usable forage depends on factors like proximity to water and landscape steepness, which impact whether an animal is willing to walk to a food source.

"StockSmart integrates these factors effectively to calculate how much food is really available to livestock in the area that the user defines," Hall said. "Its power lies in its ability to provide easy access to remotely sensed data on forage production and variability."

For Hudson, StockSmart is a perfect example of WSU Extension's mission to help communities in Washington state.

"Extension's job is to use applied science to solve real-world problems for people whose livelihoods depend on it," Hudson said. "StockSmart's grazing management tools already do that, and the platform could soon provide another public benefit - wildfire mitigation. We may literally be able to change fire risk into food by converting flammable vegetation into meat."