AI to help track declining curlew

Conservationists tracking curlew across Wales have ditched the tedious task of manually sorting through thousands of camera trap images, thanks to new artificial intelligence technology.

Researchers from the GWCT and Liverpool John Moores University have developed AI cameras that correctly spot curlew over 90% of the time, without any false alarms. During trials across 11 Welsh sites in 2024, the technology processed more than 1,000 images in real-time via mobile networks. The system even picks out well-camouflaged chicks that human eyes often miss.

“Where historically with traditional trail cams each image would need to be sifted through manually, we are now able to see in real time what is happening and spot the chicks when they hatch,” explained Katie Appleby from GWCT Wales.

The breakthrough addresses a major bottleneck in conservation work. Previously, teams spent countless hours reviewing footage while they could have been implementing protection measures in the field. Now they can respond immediately when birds are detected, focusing their efforts on practical interventions like nest protection and predator management.

The AI was trained using nearly 39,000 images and can also identify 25 other species, including foxes, deer and rabbits. This technology could transform wildlife monitoring across British moorlands and grasslands.

Did you like this article? Read more news from Shooting UK, or subscribe to Shooting Times & Country magazine.