When roboticists want to give their creation a specific skill, they often turn to the natural world for inspiration. Boston Dynamic’s CHEETAH robot can run faster than 29 MPH, thanks to a combination of foot, leg and body movements that are inspired by living cheetahs. But before they can copy nature, roboticists need to know how an animal actually moves. Some animals are so fast or so rare that no one knows their exact motions.
A mechanical engineering professor at Stanford University faced that problem when he started researching birds to build a better flying robot. Thousands have never been filmed with a high quality camera, leaving their flight mechanics a mystery. He started shooting videos with an ultra-high-speed camera and asked his students to take videos too. The resulting footage has revealed brand new bird behavior and some of the finer points of bird flight.
“Our camera shoots 100 times faster than humans’ vision refresh rate,” assistant professor David Lentink said in a release. “We can spread a single wing beat across 40 frames, and see incredible things.”
His Phantom cameras shoot at more than 3,300 frames per second. If the resolution is lowered, they can hit 650,000 frames per second. That’s enough to easily slow the flight of hummingbirds, which flap their wings 50 times a second, to a speed easily visible to the human eye.
One of his students discovered that hummingbirds exhibit a previously unknown behavior: Like a wet dog drying itself, the birds rapidly twist their body after take off. They twist 55 times per second, which is why no human observer had picked up on the movement in the past.
Lentink and his students are currently applying their bird research to their work with robots. Search-and-rescue is an ideal application for which to build better flying robots, as they need to be able to take on unpredictable conditions.
“Hummingbirds are amazing at hovering, but it’s not a very efficient form of flight,” graduate student Waylon Chen said in the release. “A swift flies a lot, so it has a very efficient wing platform, but its legs are too short to land. As we lay out the goal of our robotic design, we can pick and choose which natural mechanisms will be useful, and incorporate only those.”