Animal welfare in sustainable food systems
Cameras for healthy hooves
Lameness is one of the biggest animal well-being issues in cows. That’s why Wageningen University & Research wants to respond to such discomfort at the earliest convenience. One way to do that is by equipping barns with cameras that use AI to detect lameness automatically, thus allowing researchers – and eventually farmers as well – to take even better care of their animals.
Lame cows are a tricky problem for farmers. The cows may be in a lot of pain, which is undesirable for ethical reasons, not to mention that cows in pain produce less milk. The lameness is usually the result of problems with the hooves, for example caused by an infection. Cows are at risk of Mortellaro’s disease – an inflammation of the hoof skin – if they have to walk on wet, dirty barn floors. The infectious disease is caused by bacteria that are transmitted from one cow to another via their environment. Hoof problems can also be caused by standing for a long time, for example because the places to lie down are uncomfortable or because the cows have to wait for milking. Lameness is not a binary condition.
Severe forms of lameness are clearly visible, but milder forms can also occur with subtle changes in walking patterns.
These milder forms of lameness are much harder to see, but it is important to detect them in good time so that treatment can be started. However, it is too much work for farmers to check every day whether their cows have an infection or problems walking. It would involve lifting the legs of all the cows for a check-up every day, or closely studying how the animals walk. It is therefore important to be able to detect abnormalities at an early stage, says animal scientist Claudia Kamphuis, who works at Wageningen Livestock Research. “The earlier problems are detected, the earlier farmers can intervene, and for example arrange for an extra pedicure from a hoof trimmer or provide more comfortable places to lie down.”
‘Farmers can intervene sooner, and for example arrange for an extra pedicure for the hooves’
To detect infections and problems walking at an early stage, WUR researchers have installed cameras and antennas in the Dairy Campus innovation centre. Kamphuis: “Our colleagues look at Mortellaro’s disease and we study walking patterns and the impact on the cows’ behaviour.”
There are cameras in the milking barn that take pictures of the hooves. The cameras at the exit from the milking barn record the gait of the cows as they walk back from the milking barn to the living areas.
Kamphuis: “We want to develop an algorithm that flags up abnormalities in the gait. To do that, we’re using a system that is based on artificial intelligence. First, we identified various points on the cow that are important in tracking the walking pattern, such as the hooves, hips and knees.
Then we taught the algorithm what a cow’s normal gait looks like. If the gait starts to deviate from this, for example due to a hoof complaint, we can send an alert at an early stage.”
The researchers combine the data from this with the behavioural data measured in the living areas. Kamphuis: “We want to determine the associations between the cow’s gait and its behaviour. One possibility is that cows that develop problems with their locomotion also exhibit different behaviour to cows that don’t develop such problems. So certain behaviour could then be a risk factor pointing to locomotion problems. For instance, cows that are lower down the hierarchy may run a greater risk than cows higher up as they have to stand more and get chased away by cows further up the hierarchy.”
‘Thanks to AI, we can get information that we simply didn’t have access to five years ago’
To combine the video data with the data on behaviour, the cows have been fitted with a collar containing a sensor. The researchers have also installed cameras in one of the living areas. Antennas capture the sensor signals so that the activities of all the cows can be tracked: where does a cow walk, how active is it and what kind of behaviour – for example, standing or lying down – does it exhibit? The cameras are used to check and validate the sensor information, as well as to generate new information. Kamphuis: “This lets us monitor the animals continuously. Thanks to artificial intelligence, we can get information we simply didn’t have access to five years ago. We hope this will let us identify problems early on.”
According to Kamphuis, the trick is to use a smart approach to analysing and storing all the data that is generated. The researchers plan to set up a data warehouse. Kamphuis: “We are considering a data warehouse infrastructure to link the video and tracking data. We have large amounts of data that is multidimensional and complex. Furthermore, researchers need to have easy access to the material that is relevant to their specific study. So we also need to consider how long we should store the different kinds of data, what the smart way of saving the information would be and how we can monitor the data quality.”
‘Cows that are lower down the hierarchy may run a greater risk of lameness than cows that are higher up’
Ultimately, livestock farmers will be able to use systems like this to detect lame cows at an earlier stage, halt the spread of infections and offer cows the right care. In addition, breeding organisations will be able to use the research results to breed animals that are less susceptible to hoof problems. Those animals can be used for further breeding. The system is also valuable to WUR itself. “Dairy Campus and the video tracking system serve as a catalyst for further research,” says Kamphuis.
“We will be able to develop targeted knowledge about preventing locomotion problems, for example by taking measures in the accommodation and feed, because we can follow the cows continuously over a long period with this system. Our colleagues can use these techniques for example to study individual roughage intakes. At Wageningen University & Research, we continue to focus on new, innovative analysis techniques.”
Share this article
WHO Claudia Kamphuis, animal scientist
RESEARCH Analysing cow walking pattern and claw problem data
TEAM Animal Breeding & Genomics
Portrayed researchers: Claudia Kamphuis with Bert Klandermans
MORE INFORMATION This project is part of the Next Level Animal Sciences (NLAS) innovation programme, and is also a collaboration with the SNN Innovation Fund and Noldus Information Technology.
Participating researchers of Wageningen University & Research collaborate with various partners to develop new research methods and technologies within the field of animal sciences. NLAS consists of three research directions, namely sensor technology, complex cell systems and data and models.
