Healthy pigs don’t bite as much
Reading time: 4 minutes
BY Albert Sikkema
WUR researchers want to use artificial intelligence to measure the resilience of pigs and be able to give early warning of a raised risk of illness or tail biting. Healthy, happy animals don’t bite each other. By linking camera footage with other signals from the shed, the researchers hope to be able to forestall approaching diseases and problems.
A reduction in resilience can lead to pigs becoming sick or frustrated. When they are run down, they are more likely to fall prey to diseases or to start biting each other’s tails. This leads to an increased use of antibiotics and affects the pigs’ welfare.
Besides the pressure being applied to cut back further on antibiotic use in the livestock sector, European legislation is expected in a few years that will ban pig farmers from docking pigs’ tails. As a consequence, there are growing fears that tail biting among pigs will increase. But how much damage does that do?
Two and a half years ago, Marion Kluivers, who works at Wageningen Livestock Research, monitored a big group of pigs at WUR’s experimental pig farm in Sterksel. Over the period of the test, 10 percent of the pigs suffered tail damage, in spite of intensive supervision and monitoring.
An automated monitoring systems is being developed so as to pick up early signs of disease and of tail biting. The pigs are monitored 24 hours a day.
“Tail biting is a multi-factor phenomenon,” says Kluivers. “It partly depends on the climate in the pen and adequate enrichment of the environment with things the pigs can play with, as well as the health of the pigs. Healthy, resilient pigs are less inclined to start tail biting.”
PHOTO Brigitte Riemer/iStock
So Kluivers wants to find out how healthy and resilient pigs are by monitoring their behaviour, so as to be able to give early warning of the risk of disease and tail biting. Because if you recognize the signs, you can make adjustments and forestall approaching problems.
‘We are looking for early warning of reduced resilience which reflects an increased vulnerability to disease in pigs’
She wants to do this using an automatic monitoring system developed to keep an eye on the pigs 24/7. She is working on this with WUR experts in the fields of animal behaviour, computer vision and deep learning, who record the behaviour and use a smart system to convert it into useful information for the pig farmer.
Researchers want to understand the group dynamics at play among pigs. PHOTO Dolph Cantrijn/Hollandse Hoogte
With these methods, the researchers want to get an impression of the group dynamics at play among pigs. “We want to know how the pigs move around as a group. That is something we don’t know enough about. With the aid of the camera footage, we want to map out the pigs’ normal patterns and spot changes in those patterns that could point to reduced resilience in one or more of the animals.”
‘By making timely adjustments, you can reduce the risk of problems on pig farms and the need to use antibiotics’
Researchers and cameras are also going to register the individual animal characteristics, and details such as food and water intake. “Before they fall really sick, pigs already start eating less. We are looking for the early signals, when an animal is not yet sick but has become more susceptible to disease.”
Tail is curled
From her study of tail biting, Kluivers knows that a pig’s behaviour changes before it starts biting. Sometimes the pigs start moving around more, shows research, and sometimes they move less. “In both cases, there is behaviour change.” She also wants to monitor the posture of the tail and how dirty the pigs are. “You can see from the images whether the tail is curled or hangs down, and whether the animal is dirty. And in this case too we want to see whether we can pick up these signals automatically in good time, as indicators of resilience or warning signs of sickness or biting behaviour.”
Marion Kluivers PhD
Senior animal welfare researcher at Wageningen Livestock Research
Research on the early signs of resilience among pigs and cows
Marion Kluivers works on this interdisciplinary research with WUR scientists in the fields of sustainable livestock, sustainable agriculture and horticulture, and farm technology.
This public-private research project on the automatic detection of resilience is still at the planning stage and still has to be approved by the Agrifood top sector. Kluivers wants to place cameras at a number of pig farms. “Initially we are going to film and collect signals 24/7, but we also want to find out whether you can collect enough information to estimate the pigs’ resilience in a couple of hours a day.”
Because the idea is that the signals lead to a fully automated early warning system that supports the pig farmer in animal management.
Reduced antibiotic use
Kluivers expects this resilience research to benefit the health and welfare of pigs and the pig farm’s performance, and to reduce antibiotics use. If a disease-prone animal is spotted early, and if tail biting is knocked on the head in good time, the need for antibiotics on the farm goes down.
‘Artificial intelligence is already being used in horticulture. We can learn from that knowledge in our animal research’
The innovative element in this research is its use of artificial intelligence. “We were already doing behaviour research using direct observation, and studying video footage from the shed. Now we want to process images automatically and link them with other information from the shed.
“To enable that, the pigs and their tails have to be automatically recognized and compared with normal images. This artificial intelligence is already used in horticulture, for recognizing plants at harvest time, and scanning them for any plants that have ended up in the wrong place. We can learn from that knowledge in our animal research.”