Cows residing at Hart Agriculture’s Seven Oaks Dairy farm in Waynesboro, Georgia are ready for the spotlight as they’re currently sporting the latest artificial intelligence technology.
Through a partnership between Amsterdam-based startup Connecterra and Google’s open-source technology Tensorflow, the Seven Oaks cows are the first in the U.S. to wear Ida — the Intelligence Dairy Farmer’s Assistant — a machine learning-based tracker and app.
“So, whereas a Fitbit allows you to look into your own being’s activity, for a farmer that would not be too handy. We take it to the next step,” says Yasir Khokhar, founder of Connecterra. “We crunch the data and we figure out which cows are not performing the way they should be or are not healthy. We’re not just giving insight into how much they move, we give insight into what kind of action you need to take.”
The tracker, worn around the cow’s neck, helps farmers track, monitor activity, and optimize food production. Ida is already being used in Belgium, the Netherlands, Spain, and Asia, with 20+ farms currently on board.
“There’s a two-fold focus on what we’re trying to achieve from this partnership. We’re a grass-fed grazing, free-range dairy, which is quite a novel approach to dairy farming in the United States. I’m originally from New Zealand and that’s how we do things over there,” says Dr. Richard Watson, owner of family-run Hart Agriculture’s Seven Oaks Dairy.
“We’re marketing our milk, not just on the components or the quality of the milk being grass-fed, but we’re also focusing a lot on traceability and animal welfare. We engage with our retail customers and their customers, giving them a story. Where does your milk come from? How are the cows treated?”
Ida allows Dr. Watson and other farmers to see each cow as an individual on the dashboard, as it records its activity level, behavioral patterns, heat stress, digestive disorders from feeding, plus the best time to inseminate as the tracker has a 93 percent accuracy rating.
As production per cow continues to go up and naming each cow in a large herd is no longer possible, monitoring and diagnosing health issues can take longer — there may be 80-100 cows per farmer, according to Dr. Watson. Seven Oaks Dairy farm has 2,500 cows on site.
Over time, the machine learning platform gets to know the cows and optimizes suggestions for the farmer.
“These remote sensing technologies can drill down to the individual cow level and tell us in real-time what’s going on from a herd health welfare standpoint,” says Dr. Watson. “Without that, we risk losing 10 to 20 percent of our herd, just because we couldn’t catch it in time.”
The TensorFlow technology, an open-source machine learning platform offered by Google, has inspired other use cases aside from farming as well.
“We’re super excited about seeing use cases like this using TensorFlow,” says Google Spokesperson Justin Burr. “When we put it out into the world, we just let anybody do whatever they would like with it. We see these interesting use cases from discovering planets from NASA’s data — we saw that happen in December — to dairy farmers using it to optimize their herds’ efficiency.”
“It’s really cool for us, and it shows great advancement within machine learning and the AI industry,” says Burr.
While this Georgia farm is the first farm in the U.S. to pilot Ida, the wearable will scale to farms across the country very soon.
“I salute Google and Connecterra for choosing to deploy this game-changing technology in Georgia,” said Georgia Agriculture Commissioner Gary Black in a statement. “The information gained through this application will help equip family dairy farmers to hone their business and stewardship practices as they continue to produce nutritious and wholesome products for all of us.”