Sign up for the farmnewsNOW newsletter
ID 124267299 © Nikoleta Vukovic Dreamstime.com
Pig Welfare

Using artificial intelligence to track animal welfare

Sep 26, 2023 | 3:10 PM

Researchers are looking to artificial intelligence to track and improve animal welfare.

The Western College of Veterinary Medicine and the University of Saskatchewan’s Department of Computer Engineering are developing new tools to document and analyze animal-based indicators of on farm welfare, such as skin or tail lesions detected on swine carcasses at slaughter.

There are also 14 industry partners involved with support from the Natural Sciences and Engineering Research Council.

Dr. Martyna Lagoda, post-doctoral fellow on Swine Behavior and Welfare, said this work is being done because consumers want to know that the animals raised for meat have had a good quality of life.

“This knowledge can be delivered to consumers by unbiased animal welfare assessments. Currently, on farm welfare assessments take place maybe only once a year,” Lagoda said. “They are not always performed by a third party so they may be biased and on farm assessments are also very time consuming and a threat to biosecurity.”

Lagoda said the computer engineering team is working on a carcass assessment diagnostic tool. She said the team has a fully working computer model that can detect and track the same pig carcass.

“It can also identify the dorsal and the lateral sides as well as specific body regions of the pig. So, for example, the tail, shoulder, or the ear, which are important parts for animal welfare,” she said.

The model is currently being trained for the detection of lesions on carcasses in terms of quantity and quality. Lagoda said the goal of the project is to determine whether there are links between lesions seen on carcasses and those collected on live animals on farm using a much wider range of lesions than previously tested.

“We’d like to confirm as well whether carcass lesions can accurately tell us about the animals’ welfare on farm,” Lagoda added.

The study included on-farm visits by research scientists. Videos were taken at the processing plant to develop AI for carcass scoring.

Lagoda said the on-farm data is currently being analyzed with the results expected to be released by April 2024.

alice.mcfarlane@pattisonmedia.com

On Twitter: @farmnewsNOW

View Comments