Sheep farmers will be able to select more profitable sheep, thanks to AI

The sheep sector has long lagged behind cropping when it comes to new technology removing labour costs. But scientists say AI may be the breakthrough needed.

New Zealand sheep scientist Dr Mark Ferguson

Artificial intelligence is on the cusp of being used by a tech start-up, headed by livestock scientist Dr Mark Ferguson, where cameras will autonomously pair up ewes and their lambs.

Genesmith is expected to commercially launch this technology later this year.

Dr Ferguson, the chief executive of consultancy business Nextgen Agri, started the AI company Genesmith and it is looking at how it can help sheep producers in Australia, and New Zealand.

He spoke about the research direction of his AI work at the Best Wool Best Lamb conference in Ballarat last week.

“Increasingly AI will have a really big role in the sheep industry,” he said.

“Anything that can be automated, will be automated.

“It will help us be much more timely with our decisions and what we do on-farm.

“We haven’t had this is in sheep, we have had nothing that has taken the labour out of things, whereas every other industry has, so this is our opportunity now to do things which we have never thought possible.

Genesmith was applying the power of machine learning, along with infield cameras, and the first areas they were exploring was using facial recognition to match lambs and ewes, in the paddock at a commercial scale.

Dr Ferguson said he believed this would unlock “enormous opportunity”.

This could include identifying ewes which raised the best and heaviest lambs, which could have a big impact on producer’s bottom lines.

Now, the animals have to be physically caught, tagged and matched at birth for the same data to be captured, a process which risks interrupting mothering.

He said it would not replace electronic identification tags but would “supercharge” them, by enabling a greater use of data that could be linked to them.

The beauty of using facial recognition was that the agriculture sector did not have to invest in all the research capital from the beginning; existing technology could be leveraged, he said.

Down the track, cameras could also be used to identify behaviours in the sheep that were indicative of health – such as grazing patterns, worm burdens, as well as assess pasture.