Facial recognition technology might seem creepy when it’s your mug in question, but move it over to the farm and its possibilities become far less dystopian.
Irish startup Cainthus is using it to increase cow productivity and lower costs on dairy farms. And with food industry heavyweight Cargill grabbing a minor steak (please kill me for that) in the company, Cainthus’ reach should expand to more markets in the coming months.
Cainthus’ software uses machine learning and digital imaging software to recognize cows’ faces and hide marks, and in doing so provide data about the animals’ temperature, appetite, water intake, and other behavioral aspects.
Regularly checking up on the herd is one of the most important aspects of dairy farming, since cows’ overall health and wellbeing has a huge impact on the quality of the milk they produce. A change in appetite or activity can signal a problem with the cow, such as illness. If, for example, water intake is low, the cows’ drinking supply may need to be refreshed. If cows’ temperature goes up, it may be too hot in the barn. In both of these cases, overall milk production is at risk.
Cainthus isn’t changing any of these issues with its technology; they’re simply making it possible for farmers to keep track of the cows without having to be physically with the herd, as has been the case historically. As cows are actually happier when people aren’t around, monitoring their activity via this software should be an added bonus for milk production.
Data-driven software could also help to ensure more precise information about cows’ health and environment. Cows are, first and foremost, an investment. The faster dairy farmers can learn about a change in environment—as well as the precise nature of the change—the sooner they can address it without milk production and overall profit being affected.
According to Cainthus’ site, the software is compatible with “all existing technology, including UAV, satellite, CCTV, and smart devices.” So far, their facial recognition technology is in use on farms in New York, California, and Canada. Cost is based largely on the size of the farm in terms of both space and number of animals. The hardware runs between $38 and $50 per animal, and data analytics for each animal are another $50 per year.
Cargill, meanwhile, has had an eye on the startups lately. In addition to their investment in Cainthus, the Minneapolis-based company recently invested an undisclosed sum into lab-grown meat alternative producer Memphis Meats. It’s also a partner in Techstars upcoming Farm to Fork startup accelerator.
Cargill’s investment in Cainthus should make the technology available in more markets, and also accelerate the Dublin-based startup’s plans to expand into areas like poultry, swine, and aquaculture. (Crop farmers can already use the software to get data on crop health, growth rate, and biomass quantification.)
So while facial recognition software will probably stay a hot-button issue where human faces are concerned, the agriculture and livestock sector seems perfectly positioned to embrace it as more and more farms go digital.