Since it was first detected in Berks County, Pennsylvania, in 2014, the spotted lanternfly has become the focus of gardeners’ ire — and the target of many stomping feet — with the coming of every recent spring and summer. The invasive species has spread to at least seven states in the mid-Atlantic and shows no signs of stopping. Now, a research collaboration between the College of Engineering and the Academy of Natural Sciences is trying to help.
Over the winter months, most adult lanternflies will die off, leaving behind egg clusters that will begin the life cycle anew next spring. By collecting user-submitted photos of the clusters and taking advantage of machine learning, the team can teach computers to identify them in hard-to-reach places.
“It’s likely that lanternflies came to the United States as egg clusters attached to some kind of cargo. They can be laid in places that are dangerous for people to look for them, like under cars or trains,” explains Maureen Tang, PhD, associate professor of chemical and biological engineering and one of the project’s lead researchers. “By teaching computers what the eggs look like, we can identify and eliminate them more efficiently, hopefully stopping the spread.”
Drexel researchers have previously used similar techniques to identify cracks and defects in infrastructure and manufacturing.
“Our approach in this project is to leverage our laboratory and computational capabilities in image collection and processing to determine the most efficient ways to quickly and reliably detect egg masses,” says Antonios Kontsos, PhD, associate professor of mechanical engineering and mechanics and another project leader. “I am excited about this opportunity because the project aligns with my group’s expertise, but its broader societal impact is quite different from what our research has been used for so far.”