Scottish Farmers Test Machine Vision to Manage Pig Pugnacity

Algorithms aided by 3D cameras predict when pigs are about to nip each other’s tails

3 min read

Photo: Arctic Images/Getty Images
Tasty Tails: In extreme tail-biting outbreaks, up to 30 percent of pigs raised together may be affected by bites so severe that their carcasses are no longer fit for human consumption.
Photo: Arctic Images/Getty Images

Pig farmers want human diners to bite into the delicious pork they produce, not for swine to bite each other. (Yes, it happens.) Now, using 3D cameras and machine-vision algorithms, scientists are developing a way to automatically detect when a pig might be about to chomp down on another pig.

Pigs have an unfortunate habit of biting one another’s tails. Infections from these bites can render up to 30 percent of a pig farm’s swine unfit for human consumption. Docking, or cutting, pig tails can reduce such biting but does not eliminate it, and the routine use of docking is banned in the European Union. There are a wide range of potential triggers for outbreaks of tail biting—among them genetics, diet, overcrowding, temperature variations, insufficient ventilation and lighting, disease, and even the season—so it’s an unpredictable problem. “Tail biting is a very frustrating challenge,” says John Deen, a veterinarian and epidemiologist at the University of Minnesota. “Controlling it has not always been that effective.”

To predict and potentially prevent tail biting, researchers in Scotland monitored 667 undocked pigs on a farm using both time-of-flight and regular video cameras that recorded continuously for 52 days. The pigs were checked at least twice a day for evidence of biting.

Each time-of-flight camera emitted pulses of infrared light from LEDs 25 times a second, and recorded the amount of time needed to detect reflected pulses. This data allowed scientists to track each pig’s position and posture. Machine-vision algorithms from farm-technology company Innovent Technology, in Aberdeenshire, Scotland, then determined which activities might serve as possible early warning signs of tail biting.

The scientists found that before outbreaks of biting, pigs increasingly held their tails down against their bodies. Moreover, the software could detect when these changes in tail posture occurred with 73.9 percent accuracy. “It looks like good technology, and I’m very interested in how it could be applied on a farm,” says Deen, who did not take part in this project.

If farmers think a biting outbreak is likely to happen in a pigpen, they could deploy distractions such as straw, knotted ropes, or shredded cardboard, which tap into the pigs’ instincts to root and chew.

“Another thing that people try is to apply bad-tasting stuff such as Stockholm Tar to tails,” says Richard D’Eath, an animal behavior scientist at Scotland’s Rural College, in Edinburgh, who worked on this research. An early warning system could help farmers use such remedies only when needed, which would save money.

This research was part of a £160 million push by the United Kingdom to support innovative farming technology through its Agri-Tech Catalyst program. Agriculture and food already help generate more than £108 billion annually and support 3.9 million employees. A recent industry-led review suggested that incorporating digital technologies such as robotics and autonomous systems into food manufacturing could add £58 billion to the U.K. economy over the next 13 years.

A three-year project called TailTech is now furthering the development of this early warning system with up to £676,000 in funding from Innovate UK, a government agency. The aim is to test a prototype system on more than 16,000 pigs at nine farms throughout Europe for roughly 18 months, with each time-of-flight camera capable of monitoring up to 300 pigs, D’Eath says.

/image/MzEzNjc1Nw.jpegPig Vision: The TailTech system monitors pigs with cameras that produce 3D images. Each pixel is created by measuring the reflection of infrared pulses as an indicator of the distance from objects. Software then represents the distances with colors.Photo: Marianne Farish/SRUC

TailTech will compare the system’s efficacy for different types of pig farms, including ones with docked and undocked pigs, with and without straw on their floors, and with pigs of varying genetics, diet, and group sizes. The project will analyze what fraction of pigs hold their tails low, and for how long, before scientists are sure a tail-biting outbreak will occur. And it will assess how predictors vary between farms. The researchers also aim to improve the accuracy of their system. “In practice, though, 73.9 percent is good enough for the system to work well,” D’Eath says.

The ultimate aim is to have an early warning system “that reads out continually on a screen and also sends alerts to the farmer’s smartphone,” he adds. “No technical expertise will be needed once the system is installed.” The software will compute trends to give farmers a better idea of their herds’ current level of risk, D’Eath says.

A farmer might not buy technology designed solely to detect tail biting, but D’Eath notes that this system is being developed as an add-on to a camera-based automatic pig-weighing system called Qscan that Innovent already produces to help farms meet contractual weight targets. Deen, the veterinarian and epidemiologist, thinks this strategy could make all the difference. As he says, “If Qscan is already in place, I think farmers can quite easily justify adding this system at little cost.”

This article appears in the October 2018 print issue as “Machine Vision to Curb Pig Pugnacity.”

This article is for IEEE members only. Join IEEE to access our full archive.

Join the world’s largest professional organization devoted to engineering and applied sciences and get access to all of Spectrum’s articles, podcasts, and special reports. Learn more →

If you're already an IEEE member, please sign in to continue reading.

Membership includes:

  • Get unlimited access to IEEE Spectrum content
  • Follow your favorite topics to create a personalized feed of IEEE Spectrum content
  • Save Spectrum articles to read later
  • Network with other technology professionals
  • Establish a professional profile
  • Create a group to share and collaborate on projects
  • Discover IEEE events and activities
  • Join and participate in discussions