Researchers working on AI to help flag up sick animals earlier

A team of researchers is working to develop new ways of identifying sick livestock using artificial intelligence (AI).

A project – led by AI firm Quant Foundry in partnership with Bristol Ve t School and the Edinburgh-based AgriEPI Centre – is assessing the feasibility of using AI methods and behavioural analytics to flag sick animals at an earlier stage.

“The technology used is video analytics from off the shelf surveillance cameras, combined with AI algorithms to identify a number of symptoms in the animals,” said Quant Fo u n d r y managing director Dr Chris Cormack.

“We will be looking in the first phase to identify lameness, then delve deeper into the data to highlight those animals that are ill and track their illness and progress.”

He said a number of subtle indicators of illness had already been identified, and work is under way to train AI systems to pick up on these and provide a reliable early warning system for farmers.

“A s part of the feasibility study we will be working with Agri-EPI and the farming community to build and refine the process to reach a viable cost -effective solution using off the shelf solutions,” added Dr Cormack.

Professor Andrew Dowsey from Bristol Vet School said: “This collaboration is a fantastic opportunity to translate cutting-edge AI approaches to build upon the UK’s high standards in cattle welfare and support farmers in our targets for net-zero emissions.”

Work is going on to get early warning of sickness.