Any experienced farmer will tell you that there is no need to lead a cow for a costly examination by a veterinarian - a sick animal is already visible. Eats poorly, walks slowly, looks plaintively, mumbles sadly, etc. The authors of the Dutch startup Connecterra decided to replace subjective perception with accurate measurements and developed a system for early diagnosis of diseases in cows based on their behavior.
The source of information about the state of the cow is a modified fitness tracker, the software of which is built on the basis of the open Google TensorFlow platform. It is attached to the neck of the animal and constantly measures the body temperature of the cow, the pace of her movement, the frequency and duration of stops, the duration of meals, the movement of the jaws, chewing gum, etc. In a word, all the physical activity of the individual, comparing with the data for the past days and analyzing the condition of the cow.
Yasser Khokhar, ex-Microsoft employee and CEO of Connecterra, says they can detect an illness 24-48 hours before visible symptoms appear. And with a competent organization of cow care, this will give a 20-30% increase in the efficiency of farm labor. Isn't it worth it to fork out $ 80 for each tracker and another $ 3 per month for network maintenance? Of course not, the herders answer.
“I can recognize the disease just by looking at a cow in the eyes, ” says the authoritative British farmer Mark Rogers, “Because people have been raising cows for 7, 5 thousand years and without any electronics.”
Automating herd data collection will help businesses, it is a fact, but if young pastoralists rely on digital technology, they will lose the experience of generations of their profession. And they will become hostages of electronic toys and credit organizations, while cows will remain cows - they will eat grass and give milk, as they did millennia ago.