A new tech tool uses machine learning to identify whales' tails – or flukes – and is boosting researchers' work in the Azores. 

Lisa Steiner has been studying sperm whales for 40 years and for most of those has had to spend hours matching photographs of flukes to understand the whales of the Atlantic. 

"We started with black and white film in the cameras that we would develop ourselves in the darkroom and then print contact sheets and then from the contact sheets you pick out the specific fluke pictures that you want to print and then you printed those pictures and then you match them," she explained.

But now she can upload the digital images she takes to Fluketracker – a program developed by Amazon Web Services (AWS) and Capgemini – which is quickly learning individual markings that make the flukes useful for identifying whales, much like human fingerprints. 

"It finds matches that ... looking by eye I would never find," Steiner said. 

The algorithm is also learning to recognize the whale even if the contour of a tail has changed slightly due to wear and tear.

And it's also helped to find new whales by "looking" at images far faster than the human eye. At least 200 new whales are now part of Steiner's database since she began using the software.