The use of Artificial Intelligence (AI) technology as a complementary method of disease surveillance on livestock farms in Wales is being explored as part of an award-winning programme dedicated to reducing antibiotic resistance in animals and the environment.
The AI work forms part of a Practice Syndromic Surveillance Project (PSSP), which is itself part of the Arwain DGC (Defnydd Gwrthficrobaidd Cyfrifol / Responsible Antimicrobial Use) programme. The Welsh Government-funded programme supports farmers and vets in the fight against antimicrobial resistance (AMR) through data-driven decisions, innovative technologies, and promoting best practices.
The PSSP is a collaboration between the University of Liverpool, the Wales Veterinary Science Centre (WVSC), and the veterinary delivery partner, Iechyd Da, and builds on an earlier Arwain DGC pilot.
The syndromic surveillance system uses the university’s FAVSNET (Farm Animal Veterinary Surveillance Network) programme to collect disease symptom and antibiotic use data from a trial network of veterinary practices across Wales. The team is then developing and evaluating several methods, including those based on AI to analyse the information and identify specific disease syndromes – or patterns -occurring in farm animals covered by the participating practices.
Available in near real-time, the information could be used to monitor the health status of livestock, with vets, farmers, and other stakeholders alerted early on to the possibility of new and emerging (or re-emerging) disease and complemented by information obtained from government surveillance. Informed decisions can then be made to avoid further spread of disease and implement timely treatment options.
Early detection reduces the impact of disease outbreaks, helps prevent their spread to other animals, and reduces the need for antibiotic use. This, in turn, helps protect animal welfare, public health, and the economy and contributes to the ‘One Health’ approach to human, animal, plant and environmental well-being.
For the PSSP pilot study, five syndromes affecting farm animals were chosen: abortion, joint ill, mastitis, pneumonia and lameness. In total, 32,799 consultations were collected between February 1st 2024, and January 31st, 2025, mostly from cattle (19,224, 58.6%) and sheep (12,356, 37.7%).
The pilot results have been very encouraging, demonstrating that large amounts of data can be harvested smoothly and continuously from veterinary practice management systems as vets go about their daily work.
The technology also allows the link between antibiotic use and specific disease syndromes to be explored, enabling the identification of syndromes associated with high antibiotic use, and therefore a priority for control.
Iechyd Da vet, Robert Smith, said:
“The project aimed to explore whether this type of ‘live' data could be collected from practising vets in Wales as they went about their farm visits. Also, to investigate whether the information could be collated and analysed promptly to provide a workable disease surveillance system to reduce infectious disease risks and the need to use antibiotics.”
He said the PSSP's collaborative approach had worked well, with the project's next phase to include more Welsh veterinary practices using different practice management systems.
Robert said:
“By increasing the amount of data that we collect and refining the AI analysis techniques, we hope to improve the standard of disease surveillance we are achieving, with the ultimate aim of creating a system that can be used to identify priority diseases and strengthen disease surveillance across Wales.
“Reduced disease incidence and targeted prevention will then contribute to the reduction in reliance on antibiotics and other antimicrobials and slow the development of antimicrobial resistance.”
Professor Alan Radford, Professor in Veterinary Health Informatics at the University of Liverpool, said:
“The FAVSNET project team is massively grateful to the partnering veterinary practices for sharing their data and making this project possible.”
