Stink bugs, recognizable to most by the gnarly, unpleasant odor they emit after being unceremoniously crushed, are actually one of the major agricultural pests tormenting farmers in North America and Southern Europe. The pesky invasive species, which may have hitchhiked from Asia to the US via shipping containers in the 1990s, has reportedly contributed to as much as €500 million in crop damage in Europe in 2019 alone. All that destruction makes keeping a close eye on these insects a top priority, but today’s methods of doing so remain labor-intensive and relatively ineffective on a large scale. Now researchers believe that a combination of camera-equipped drones and new AI models could finally give farmers a better way to fight back. A group of Italian researchers writing in the journal Pest Management Science recently released findings from a study in which they used an off-the-shelf, commercial drone to collect high-quality images of stink bugs nesting in a pear orchard. The researchers developed an ‘automated flight protocol’ that allowed them to control the drone and hover over infected crops using a mobile app. The drone was then able to capture hundreds of high-quality images that were used to train an AI algorithm that can successfully identify future stinkbugs. That model, thanks in part to the quality of the images captured by the drone, accurately identified stink bugs 97% of the time. “Overall, this new monitoring system demonstrated the potential of integrating UAV and AI to detect and quantify the presence of insect pests with the size and shape of H. halys,” said Daniele Giannetti, researcher and co-author of the University of Parma in a statement. . Drones equipped with cameras hovered four meters above a pear orchard and collected hundreds of high-definition photos. Credit: University of Modena and Reggio Emilia Drone caused insects to freeze in place Before the drone experiment, observation and surveillance of stink bugs typically involved using pheromone traps to lure the insects in for manual counting. This approach, the researchers write, is limited in its effectiveness, especially when it comes to measuring stink bug numbers in large, dispersed agricultural areas. Giannetti said these more traditional monitoring methods are also limited because of the economic costs associated with hiring workers to conduct that monitoring. For their research, the researchers equipped a DJI Matrice 300 drone with an HD camera and had it perform 16 different flight missions over the orchard. However, once the drone started hovering over the pear, something interesting happened. Rather than scattering or falling from the crops, as they do when human observers approach, the insects froze silently. About 85% of the measured insects showed this freezing behavior when the drone hovered above them at a height of 4 and 8 meters. The unusual freezing effect gave the drone enough time to take a high-quality photo with the bug in the frame. From there, researchers used that footage to train an AI model that can detect stink bugs. Although the drone captured a total of 2,459 images, only 402 of them actually contained stink bugs. Researchers manually labeled these bugs to help train the AI model by delineating the insects with a rectangle. Once properly trained on the images, researchers say the AI was very adept at spotting instances of stink bugs in other images. Although the researchers limited their focus to stink bugs, they say the same underlying principles can be used to monitor a variety of pests. Future researchers should capture their own new drone-based images and then feed them into their own custom AI. Farmers and conservationists in the US and around the world are already readily using drones to assist with crop management, soil analysis and real-time “weed scouting.” “This experience is truly encouraging,” University of Modena professor and co-author Lara Maistrello said in a statement. “We find these results exciting, especially because there are so many future applications.”
