Robot lawn mover
Scientists seeking to understand the mechanics of bird flight have constructed PigeonBot, a robot made from 40 pigeon feathers (and a few other components).
While airplanes maneuver by altering their wing elements, birds can morph the shape of their entire wings to dive, bank, and coast through the air, increasing both their efficiency and agility. This new study on pigeon wings has not only provided a simpler model for how bird wings work but allowed engineers to integrate that knowledge into a nimble flying machine. The researchers hope that PigeonBot will provide an inspiration for those constructing flying machines, as well as those studying birds.
“You can simply use the cadaver of a bird, and there are many in museums, to develop a robot without harming any animal to study their flight,” David Lentink, the studies’ corresponding author and assistant professor of mechanical engineering, told Gizmodo.
“I first started with just one question: How do individual feathers work together?” said Laura Matloff, Stanford University graduate student. Matloff had long been interested in animals, having volunteered at wildlife hospital as a child, and was interested in incorporating knowledge from biology into engineering. She led one of the studies that measured cadavers of food-grade pigeons using motion-capture systems, taking measurements on how the feathers moved as they manipulated the bones.
Aerospace engineers once envisioned an aircraft based on pigeon wings, in which pilots could control each individual feather. But in reality, the real pigeon wing operated much more simply. The team’s measurements allowed them to create a model of pigeon flight that manipulated just two variables: the angle of the overall wing and the angle of the finger joint halfway through the wing. A flexible tendon, like a rubber band, alters the angle of all the feathers in tandem.
But how do the feathers stay locked together with air flowing past them? Micro-CT scanners and an electron microscope shed more light on the function of the rich microscopic system of barbs and hooks that activates when the wings are spread, like avian Velcro. They measured how this “directional Velcro,” as they called it, withstood strong forces in a wind tunnel. They published their results in the journal Science.
Stanford graduate student Eric Chang’s childhood interests drew him to flying creatures like birds, bats, and insects, and he joined a competitive small aircraft design team as an undergraduate. He used his knowledge to build on Matloff’s research and on two decades of bird-inspired robot knowledge more generally. The team put 40 real pigeon feathers onto an artificial skeleton that could move in two places, at the base and at the finger joint, with rubber bands controlling the angle of the feathers, recreating what the researchers saw in the cadaver studies. They combined it with a propeller, artificial tail and rudder, controllers, and sensors, and tested it in a wind tunnel as well as outdoors using a remote-control. They published their research in Science Robotics.
The team felt relief when they finally got the robot flying. “I remember the first day that it flew, after we landed it successfully and it was all in one piece, I collapsed on that ground,” Chang said. “It was this feeling of ‘oh my gosh, it actually worked, and I can breath easier right now.’”
PigeonBot sidesteps the aerospace engineering instinct to articulate every piece of the flying machine, in favor of a simpler model that flew with ease.
Lentink saw various applications for the research. Perhaps a company could use their measurements to develop a new kind of Velcro, and the model is another way for aerospace engineers to consider a simpler model of flight. But he’s not as interested in applications as he is in research and teaching. He imagined a world in which museums could better study bird flight by creating robots based on specimens already in their collections. The team included measurements comparing the forces that the feather Velcro could withstand in species including the Cassin’s kingbird, bald eagle, and endangered California condor.
“You can recreate a condor robotically to understand its flight behavior, and use this insight to help the species,” he said.
(Also, for those of you worried that birds are government drones, this study demonstrates that engineers still aren’t completely sure how birds fly. But maybe you can worry now.)
The researchers plan to continue their studies on the robot to further explore the directional Velcro function and to take more measurements to further improve PigeonBot.