AI-powered 'My Health Records' for cows to manage clinical mastitis

Scientists, technologists, retailers and farmers are teaming up, using AI and creating the equivalent of My Health Records for dairy cows, to better manage clinical mastitis.

The collaboration aims to reduce the use of antibiotics in the milk supply chain and improve the health and management of dairy cows with mastitis.

The Clinical Mastitis Decision Support Tool Project is a $3.5 million, three-year collaboration between Dairy Australia, Coles, DataGene, Food Agility CRC, University of Sydney, University Technology Sydney and Charles Sturt University.

Together they will create a digital tool for dairy farmers and vets that details the mastitis history of each dairy cow, just as My Health Records are digital records of people’s medical history to support individualised healthcare.

Then, using model-driven AI, the tool will provide management advice to farmers and vets based on the history of individual animals and the source of infection. Advice will include which antibiotics to administer, if any.

“There are a lot of factors that influence how we can best manage a cow with clinical mastitis, including cause of infection, the history of the animal and the different types of antibiotics,” Dairy Australia animal health and fertility lead Stephanie Bullen said.

“This digital tool will draw together a range of data, including individual cow history and information on the cause of the mastitis to give a more accurate management recommendation,” Dr Bullen said.

“Ultimately, it will mean healthier, happier cows, more targeted use of antibiotics and less milk down the drain.”

Coles’ Charlotte Rhodes said the project was one way the supermarket was working with the industry through the Coles Sustainable Dairy Development Group to improve animal welfare outcomes, optimise farm productivity and help ensure a sustainable future for Australian dairy.

Food Agility CRC chief scientist David Lamb said the project was "true industry-led, multi-disciplinary research at its best", with farmers and retailers working alongside animal health specialists, data analysists and technology developers.

“What’s really exciting is that the tool will use artificial intelligence to constantly learn and update based on new information about the animal, infection source and available management options,” Professor Lamb said.

Mastitis is the number one health issue in dairy cows, costing the industry $150 million annually, and is responsible for two-thirds of antibiotics supplied to dairy farmers by vets.

Currently, the most common method of diagnosis is through a visual assessment of the milk. However, research shows that not all cows displaying signs of clinical mastitis need antibiotics.

The project team will conduct research and trials on dairy farms across NSW, south-west Victoria and Tasmania, with a prototype tool to be tested on-farm within 18 months.

The final tool will be released more widely to Australian dairy farmers, via DataVat, the Australian dairy industry’s central data repository.