Artificial intelligence improves welfare of cows

Mortellaro's disease, also known as Italian foot rot, is an inflammation of the skin of the hoof. Cows suffer greatly and often go lame due to a tendency to take their weight off the affected hoof. The infectious disease is caused by bacteria passed from one cow to another, and it can survive in the environment for a long time.

“It is not feasible for farmers to assess whether cows are infected with it on a daily basis: they would have to lift and inspect all animals legs every day,” says lecturer Piter Bijma of Wageningen University & Research (WUR). “Practically, problem cases are only discovered once cows are evidently lame. We are conducting research into a system that automatically detects infected cows early on. It will help farmers prevent the spread of the disease and help them to take care of their cattle.”

Researchers at the WUR innovation centre Dairy Campus are equipping certain areas with cameras and antennae. The cameras will be placed in the milking parlour and take pictures of the clefts in the cows’ hooves. These images will then be uploaded to a computer. The researchers will then train the system’s artificial intelligence algorithm. They will for example highlight the hooves that are infected with Mortellaro’s disease by putting a coloured box around them. This allows the system to automatically detect cows with infected hooves.

Bijma: “The cows will also be fitted with a collar with a battery and a sensor. The antennae pick up signals from the sensors. This way, we know which cows have been where in the barn and we can follow the transmission of infection. With this information, we hope to be able to find the cows that make a large contribution to spreading infection; the so-called ‘super-spreaders’. Perhaps these cows are more easily infected, secrete more bacteria or have more contacts. These cows could possibly be disposed of or excluded from breeding.”

Farmers will soon be able to detect Mortellaro’s disease earlier and to intervene faster. They can trim the hooves themselves and disinfect the materials, or hire a hoof trimmer. Researchers can use the system to investigate the effects of preventative measures, such as alternative flooring. Bijma: “In addition, based on our research results, breeding organisations can breed animals that are less susceptible to Mortellaro’s disease in the future. We know from previous research that cows that are less likely to get sick are also less likely to infect other animals. They can continue to breed with these animals.”

In addition to the research being performed by Bijma’s team, a study into cows’ ability to walk is currently being conducted at the Dairy Campus. Researchers ultimately want to link the diagnosis of Mortellaro’s disease to locomotion to discover what the link between the two is.

Both studies are part of the data and modelling work carried out by the Next Level Animal Sciences innovation programme. The aim of this line of research is to transform large and complex data sets into usable and meaningful information about animal behaviour, health and welfare.