Caltech is training drones to ‘herd’ flocks of birds. Here’s why



“The existing bird-scaring techniques employed to avoid bird strikes on airfields include lasers, distress calls, gas cannons, flares, inflatable scarecrows, dogs, radio-controlled planes, shotguns, and live birds of prey,” Soon-Jo Chung, associate professor of Aerospace and Jet Propulsion Laboratory Research Scientist at Caltech, told Digital Trends. “According to surveys by the International Bird Strike Committee, none of those systems work adequately. Simply put, birds are too intelligent!”
Knowing the limitations of solutions like training falcons or having a ground-based pilot control an RC airplane, Chung’s team set out to create an autonomous technology which could help. Their drone-based system works by figuring out the best paths around a flock of birds to effectively herd the flock in a desired direction. To create the algorithm, the researchers modeled flocking behaviors, and then reverse engineered these to establish the ideal flight path for a drone to move the flock away from protected airspace — crucially, without dispersing it in the process.
The drone-based system was recently tested and was shown to be successful. The only current weakness comes when dealing with multiple, or larger, flocks. To that end, the researchers hope to find ways to scale the project to use multiple drones for situations in which there are multiple flocks of birds.
“Bird and other wildlife strikes on aircraft cause well over $400 million in damage annually to U.S. civil and military aviation, and $1.2 billion worldwide,” Chung said. “I also hear [about] wineries battling crows that are eating fruit before the harvest. Imagine a drone taking off autonomously, and herding these live animals away from your protected zone. That’s the future [vision] we are working on.”

أحدث أقدم