AI saving UK's endangered birds

Conservationists are using cutting edge artificial intelligence in a bid to save one of Britain's fastest declining birds.

Researchers from the Game & Wildlife Conservation Trust have teamed up with Liverpool John Moores University to deploy an AI system that can spot curlews and their chicks in real time, helping to protect vulnerable nests during the crucial breeding season. 

Lee Oliver, director of the trust in Wales, says the birds are in trouble and could be extinct as a breeding population within 10 years.

“We're doing all we can to reverse that and to increase the population. Yes there are plenty of curlew passing through from other countries but the actual established breeding population is extremely low.”

Animal 'facial recognition'

Computer scientists at Liverpool John Moores developed the system which delivers almost 100% recognition of species via a machine learning tool. Under the banner Conservation AI Group, it runs through Trap Tracker and counts dozens of clients and partners worldwide, including Durrell, Natural England, Chester Zoo and Mount Kenya Wildlife Conservancy.

In the case of birds, its algorithms can recognize when a specific species is on the nest and sends an alert when any chick hatches. Alternatively, it can communicate to rangers when birds are in trouble or under predation.

“The technology gives us a tool to monitor a variety wildlife on farms, not just the curlew,” adds Lee. “It has enabled us to start a ‘Farm Gate’ project where cameras monitor all the biodiversity and wildlife passing through gates, over and through gates.”

Kate Appleby, who works on the Curlew Connections Wells project, says not enough chicks are fledging to adulthood, says serious monitoring, especially of predators is urgently needed, and the technology has been game-changing, as it allows monitoring to be much more flexible – anytime, anywhere!

'Game-changing'

And, she says, conservationists are now more aware of the problems the birds are facing.

“The technology has been brilliant; it doesn't replace field workers and it's it works really well alongside and it's another tool in the tool box for support. The use of the drone and use of cameras is absolutely game-changing.”

Paul Fergus, professor of applied artificial intelligence at LJMU's School of Computer Science and Mathematics, said the cameras connect via 3 or 4G, so you put a SIM card in you can put SD cards, collect images or video.

“When it sees a heat signature, it basically triggers that camera or take a photograph of whatever is in front of it. Then that image then gets sent back to the conservation team.”