Graphic representation of measurements that can be used in facial inference algorithms.
Graphic representation of measurements that can be used in facial inference algorithms.

Ph.D. Student Develops AI Platform for Dairy Farmers

DairyFit allows farmers to search data for behavioral patterns to better manage herds

Catie McVey wants to bring big data to the small farmer.

The Ph.D. candidate in animal biology at UC Davis has developed a customizable artificial intelligence platform called DairyFit to help dairy farmers get a better sense of what is happening in their herds.

McVey combined biostatistics and machine learning into an algorithmic platform where farmers can visualize data that is already being captured and analyze it to look for insights into behavior.

“DairyFit helps to pull out patterns hiding in big messy data streams,” she said. “Then it lets the farmer decide the right management intervention. I want to empower farmers with data, not replace them.”

McVey’s work, which is the subject of her dissertation, is getting attention. Earlier this year she won the Animal Health + Industry Award as part of the Big Bang! Business Competition.

Using what is already available

She describes DairyFit as “big brother for cows” and said the technology can analyze data from sensors, ear tags and other precision livestock farming technology that documents eating, chewing cud, walking, time at rest and other behaviors. Mining that data can highlight patterns that can be difficult or time consuming to observe in person.

The platform can be used to audit operations to determine a number of things, including if cows are laying down enough, spending the right amount of time in the milking parlor or if they have spent too much time on their feet. It can also detect shifts in behavior.

“You can kind of let the cows speak for themselves,” said McVey, who worked with her advisor, animal behavior professor Kristina Horback, and biostatistics professor Fushing Hsieh on this project. 

The platform can be deployed for hundreds of cows, small herds and individual animals, opening up the world of big data analysis to small farmers. “We should be giving them tools to engage with their data,” McVey said. “I’m never going to make an algorithm that knows better than the farmer.”

Identifying patterns

One pattern that emerged during testing: Researchers expected sick cows to walk into pens at the rear, with the big, powerful ones in front. What they found was that the healthy cows led from the front and chased from behind.

“We started referring to them as caboose cows,” said McVey, who is also a member of the Animal Behavior and Cognition lab in the Department of Animal Science. “They seem to make sure the herd stays together,” she said.

Having that kind of intelligence can help manage herds. To that end, McVey has made the code public, is testing the platform with a small group of farmers and hopes to offer it as a smartphone app.

Supporting entrepreneurship

The Big Bang! Business Competition is organized by the UC Davis Mike and Renee Child Institute for Innovation and Entrepreneurship, which helps researchers develop and advance their ideas into business ventures via competition, workshops, mentoring and networking opportunities. Prize funding comes from corporate, nonprofit and other sponsors.

The award came with a $10,000 prize and $5,000 in in-kind services from AgStart, an agriculture, food and health accelerator that helps entrepreneurs bring their ideas to market. 

Media Resources

Primary Category

Secondary Categories

Food & Agriculture