AI keeping an eye on eggs

Artificial Intelligence is being touted as a solution to labour shortages in the egg industry.

Egg farmer Philip Szepe says AI could help monitor the wellbeing of his chickens and improve productivity. Picture: Supplied

Philip Szepe has always wanted an extra set of eyes to keep watch on his chickens.

Keeping watch for predators and monitoring the birds’ health is crucial to the egg-laying operation on his farm in Kinglake West, Victoria.

Labour shortages and the impracticalities of constant surveillance have made that impossible, but it could be a camera system, backed by artificial intelligence, that solves his problem.

Mr Szepe hopes an AI trial underway by the University of Southern Queensland will unlock the key to 24-hour surveillance and data gathering.

“Typically our birds have a 15-hour day length and we probably have staff around the hens for five or six hours, which leaves a significant gap in terms of supervision.

“While we monitor simple parameters like temperature, food, water and power, we’re not getting any stockmanship feedback.”

The model is based on Japanese technology that was developed for the dairy industry.

Mr Szepe said the technology could help drive improved production and welfare outcomes.

“We can identify specific types of behaviour that, if expressed, can link to an alarm event,” he said.

“Then we could get our team to respond in a timely manner, instead of in the morning when the sheds open and the staff enter.

“We’ve had cameras in sheds for years, but it’s always required someone to monitor it. With an interactive system with alert capability and the smarts to differentiate between a normal event and an abnormal event, that is exciting.”

USQ researcher Cheryl McCarthy has spent several hours every day for the past 18 months watching video of chickens on a trial farm on the Darling Downs, west of Brisbane, to identify patterns in the birds’ behaviour to help log data into the AI software.

“We’re watching the video to start off with and we’re providing pointers and criteria around the images for what kinds of things it (AI) can extract,” she said.

The proof of concept phase is expected to be finished mid-year.

Dr McCarthy said the camera system monitored chickens as they left and entered the shed and analysed their features to identify injuries or illness.

“It’s to provide extra monitoring capability for the farmers, because they routinely walk through and inspect the flock to see how the beds are doing, but a camera can be sitting there watching all the time and not disturbing the flock at all,” she said.

University of Southern Queensland’s Dr Cheryl McCarthy is researching a new way to detect chicken welfare using machine vision.

“What we’re aiming to pick up on is, can we with the artificial intelligence in this camera monitoring, are we able to assess the flock and say it’s all looking great, or are there any conditions that would need follow-up inspection from a vet.”

The research is backed by levy-funded industry organisation Australian Eggs.

Australian Eggs managing director Rowan McMonnies said the project had the potential to significantly benefit the industry, particularly due to its cost-effectiveness.

“Technology has always been a huge part of the industry but that has really intensified in the past five years,” Mr McMonnies said. “It’s all about animal welfare and sustainability.

“This sort of artificial intelligence and machine vision project is very exciting.”