Facial recognition AI helps fast-track research in crop-ruining grape mildew

16 Aug 2021

A grape vine infected with mildew. Image: © Firn/Stock.adobe.com

Researchers have used AI and a robotic camera to cut the time spent examining grape leaves in the lab from six months down to a single day.

Powdery mildew poses a serious problem for grape farmers. When uncontrolled, it can lead to mass infections requiring fungicide and sometimes even results in the mass loss of crops. Luckily, some grapes are more resistant to this mildew than others. The only problem is figuring out which ones.

Biologist Lance Cadle-Davidson, a research plant pathologist with the US Department of Agriculture’s Agricultural Research Service (USDA-ARS), is working to develop grape varieties that are more resistant to powdery mildew. His lab recently had a breakthrough thanks to AI combined with imaging technology.

Until now, the lab has been slowed by the need to manually check thousands of grape leaves for evidence of infection. So, in comes the BlackBird.

Developed through a USDA-ARS programme, the BlackBird robot uses a camera to gather information from the leaves with the detail of an optical microscope. Such detailed information requires an equally sophisticated way of analysing the data, however, which is where deep neural networks come in.

Led by assistant research professor Yu Jiang at Cornell University, a research collaboration applied AI technology previously used for facial recognition to analyse the microscopic pictures of grape leaves.

Researchers even designed a visualisation of the network inferential process so that scientists could help build confidence with the AI models. Now, working together, Cadle-Davidson’s team validates what the robots picture, which enables Jiang’s team to teach the AI how to spot the desired biological traits in grapes more easily.

Cadle-Davidson said the results are “astounding” and that what used to take his lab team six months to complete can now be done by the BlackBird robots in a single day.

“It has revolutionised our science,” said Cadle-Davidson. “And we’re finding that Yu’s AI tools actually do a better job of explaining the genetics of these grapes than we can do sitting at a microscope for months at a time doing back-breaking work.”

The group has already received a two-year grant from the Cornell Institute for Digital Agriculture’s Research Innovation Fund for $150,000 to upgrade the BlackBird robot with infrared technology. The group was also awarded a a $100,000 grant from the USDA-ARS to bring the BlackBird technology to ARS field offices working on other crops.

“We hope to find collaborative labs who can join us in taking advantage of this tool,” said Jiang. “We see potential applications for this research in plant studies, animal fields or medical purposes.”

Sam Cox is a journalist at Silicon Republic covering sci-tech news

editorial@siliconrepublic.com