Cows and robots go together. Throw in facial recognition software, and it’s the perfect trifecta.

This is because cows are happier when they are not around people, since they identify humans as predators. Using facial recognition software to count a herd, or signal when a cow is sick or injured or not eating, is another way to keep humans out of the pastures as much as possible and keep cows happier and more productive.

“No prey animal never wants to see a predator. The less they see the happier they are. A cow doesn’t know what a robot is, so they aren’t scared of it,” said David Hunt, co-founder of Cainthus, a company digitizing agricultural practices, speaking at an Alltech conference in Lexington, Ky.

Digital farms can include robots, but Cainthus has also created facial recognition software that identifies cows so that farmers can identify potential problems in the early stages. With 1.4 billion cows on the planet, this is technology that can be applied to many farms as part of the digital dairy concept.

Going digital on the farm

A digital dairy is a farm that is information enabled with searchable queries available in real time from a smart

device with data flows. Cainthus’ facial recognition software is in the beta stage and will be available on the market in August or September this year, Hunt said. It is targeted for dairies with 2,000 or more cows, not for small dairy famers. The 4K camera can identify any cow, whether solid colored or spotted, since even a solid colored black cow has variations in its coat.

By using facial recognition software, “you know what is going on with every cow in your dairy at every second at the day,” and this helps monitor their milk production, Hunt said.

Future robotic intervention is also essential. “For me, the best cow farm in the world is the one a human never has to go into. And they only way we’re going to change that is if we use robots to intervene,” he said.

Annual losses to farmers

The annual cost to the U.S. combined dairy and feedlot industry is $8.7 billion lost to lameness, because cows produce less milk when lame because they’re eating less and are losing weight. The facial recognition software allows farmers to identify lameness in the first stage. With non-digitized farms, 30% of cows going lame are not identified until stages 3 or 4, when production value is already lost, Hunt said.

“When you’re farming at scale it’s extremely difficult to notice. When you notice lameness early, you just clean the hoof and sterilize it and it’s done,” he said.

“All the analytics we do pretty much justify the cost of the entire system. The single biggest cost saving is lameness,” he said. It costs a farmer an average of $241 per animal when lameness strikes, and the facial recognition software is $10 per cow.

As farmers look for ways to digitize dairies, wearables for cows aren’t a feasible option because of the inherent problems such as cost and a high failure rate. “Cows don’t really like wearing them and they don’t like having things shoved down their ears. It’s also questionable how accurate the data is from wearables,” he said.

Complex cow behavior

“What does a digital farm look like? For me it looks very much like your farms today except with more cows,” Hunt said.

The first step of creating a digital farm is using facial recognition software. It notices when a cow is feeding, and when it is not. It also works at night, when most cows eat. It also includes aggression gesture recognition. “Cows fight far more than we realize and the impact of those fights are significant,” he said.

“My principal thing is that Gary Larson is right. Cows are far more behaviorally complex than we realize,” Hunt said, citing The Far Side creator, who frequently used cows as his comic subjects.

“We are predators to them. They are prey animals,” he said. “Cow management is a 24-hour job. Cows are very active at night, even if we’re not. Many times we’ve seen cows clear the entire feeding trough in the first three hours the farmer has put that down and they don’t eat the rest of the night. To improve our efficiencies we need to start managing cows on their schedule rather than our schedule.”

And then there’s the aggression issue. “We didn’t plan on building a cow fight odometer,” he said. But they realized that a cow fight monitor needed to be included because cows stop eating and watch the fight and leave the trough whenever a fight breaks out. Many of the fights are over the most choice bits of food.

This indicates that farmers should breed for passivity and provide more nutrient-rich food, and those changes will increase milk production.

“We like to think we’re going to find a lot more of this as the technology develops. The rule of thumb with vision technology is if a human can see something then a camera can see it earlier and more consistently. We’re going to be able to see more and more with this and it will be cheaper and cheaper to apply it.”

See also:

Steve Wozniak on artificial intelligence, virtual reality and his favorite new tech

The key to human-robot relationships is vocalization, says Carnegie Mellon

We can’t prevent AI changing the world but we can stop robots cooking cats