Just in time for the Animal Agtech Innovation Summit in San Francisco in less than 2 weeks, Belfast-based CattleEye is coming out of stealth with an AI-backed autonomous livestock monitoring platform that allows users to identify dairy cows and find out things about them such as whether they’re sick, becoming lame, going into heat, and what their overall body condition looks like. The technology can help farmers produce $400 more milk per cow, according to co-founder Terry Canning, as things like missing an estrus cycle or a delay in treating disease can cut into already razor-thin margins.
The service starts with the user purchasing a security camera for roughly $150 dollars and mounting it three meters above the exit of the dairy building. The user provides his or her IP address to Cattle Eye and the company connects the user to the cloud.
The idea behind CattleEye was born out of Canning’s discussions with a friend in the healthcare industry who was working on technology for cancerous tumor detection. When a pathologist analyzes the scans, errors are possible. But train a camera to detect microscopic cancer cells and the odds of detection improve.
“We train algorithms using AI to take in video footage and segment out the cow in the video. It could be 10 animals side-by-side or overlapping, but it can pull them out individually and look for ways to identify them first by the patterns and the shapes of the animal,” he explains. “Once we have identified the cow, we look for insights like how they are walking today versus yesterday, have they suddenly gone lame, are they going into heat? etc.”
The software component of CattleEye’s technology provides app-based insights that can be integrated with any farm system platform, overcoming the data sioling roadblock that has left many producers disinterested in adopting technology. It will eventually launch as a subscription-based service but the company is currently focused on piloting the technology; it’s gone from monitoring 90 head to 7,500 in a period of three months, according to Canning.
Canning grew up on a dairy farm and created Farm Wizard, which he describes as the first-ever livestock management platform. It grew over 12 years eventually counting four million cattle on the platform from around the globe. Farmers and retailers including FreshDirect used the platform to manage their beef supply. In 2015, he sold the business to Wheatsheaf as part of an £8 million ($10.24 million) transaction and made a complete exit in March 2019.
With CattleEye, Canning is hoping to reduce some of the friction around existing livestock technologies. Many of them rely on attaching wearable devices to cattle either around their necks on collars or on the tailhead.
“First, we’d rather not interfere with the animal at all. Also, there is a cost to wearables because you need one for each animal. If you have 10,000 dairy cows that means you need 10,000 devices,” he explains. “There’s also a commissioning problem, which refers to switching devices between animals, especially on large farms.”
His new technology has proven easier to pilot compared to wearables because all a participating farmer has to do is hang a camera. Piloting wearables, on the other hand, required fussing around with the cows during a time when everything may not be working quite right.
In the midst of plant-based protein’s popularity, Canning sees an even bigger need for technologies to make livestock production more efficient, welfare-conscious, and lucrative for producers. CattleEye is working with undisclosed grocery retailers in the UK that want their milk suppliers to adopt technology to improve production. This may result in premiums for the producers who adopt technology or who are able to produce milk according to a retailer’s bar-setting production standards.
As a veteran in the animal agtech space, Canning is approaching the summit with hearty optimism.
“It goes back to how fast the landscape is changing since 2004. The fact that we now have these events with a key focus on the investment community is just fantastic.”
CattleEye isn’t the only startup using computer vision to track animal behaviour; Cainthus, also from Ireland, is another example.