Bristol University pioneers AI project to spot cow illness

Professor Andrew Dowsey
Professor Andrew Dowsey
This article is brought to you by our exclusive subscriber partnership with our sister title USA Today, and has been written by our American colleagues. It does not necessarily reflect the view of The Herald.

Bristol University is pioneering an innovative artificial intelligence (AI) system for monitoring cattle health and welfare through advanced camera technologies.

It aims to track individual animal behaviours and predict early signs of disease, potentially transforming cattle management practices.

For the past five years, researchers at Bristol Vet School have used cameras to identify cattle by their unique coat patterns.

However, the focus has now shifted to characterising behaviours, particularly social interactions, to understand how changes in behaviour might signal the early onset of illness before clinical symptoms appear.

This research is a collaborative effort between experts in veterinary medicine, animal behaviour, computer vision, and AI, led by Professor Andrew Dowsey, who presented the findings at the BCBC conference in Telford.

The John Oldacre Centre – the University of Bristol’s commercially run dairy unit at Langford – has a comprehensive video monitoring platform that tracks each cow’s behaviour 24/7.

This system integrates health, and production records and provides insights into how changes in social behaviour correlate with sub-clinical disease.

While existing camera technologies that can automate disease detection in dairy cows are available, they typically focus on visible symptoms during the later stages of disease.

Bristol’s research aims to identify subtle behavioural changes that can signal illness before visible symptoms appear.

Prof Dowsey explains: “We’re developing a proxy for earlier disease detection. Wearables like pedometers can track movement and basic activities.

“Still, video provides a more detailed and subtle insight, allowing us to identify signs of illness sooner.”

He added: “When cows experience infection or trauma, they often reduce non-essential behaviours, such as social interactions, so changes in these behaviours may serve as early predictors of disease.”

Currently funded by a BBSRC grant, the research focuses on detecting early signs of mastitis and lameness – two significant diseases in the UK dairy industry.

The technology also has the potential to monitor other endemic diseases and production-related issues, such as heat detection.

“By collecting months of video from 64 cameras covering our main barn at the John Oldacre Centre, we can train a model to identify behaviour changes indicative of early-stage mastitis and lameness and determine the optimal camera positions to translate the system to commercial herds,” Prof Dowsey said.

The system will be tested across a network of farms this year. If successful, it could be commercially available within the next five years.

“This technology facilitates the early identification of potential health issues, allowing farmers to intervene before problems escalate,” explains Prof Dowsey.

He added: “While our current research focuses on dairy cows, we are also expanding into beef through collaborations with Rothamsted Research.”

Get involved
with the news

Send your news & photos