Winged Victory: Fly-Size Wing Flapper Lifts Off
The hope is to build robotic flies that could work in any situation in which it would be better or safer to send them instead of humans
Photo: NMEM/Science & Society
Lucien Bull demonstrates the high-speed camera he invented in 1903, the first to record insect flight.
Despite the fact that we’d likely drown in our own waste if it weren’t for their enormous fondness for eating our garbage and excrement, flies don’t get any respect. Not from most of us anyway, who see only pests and disease vectors as we swat with abandon whenever they’re around.
But there are a few good scientists and engineers who reach for their notebooks and video cameras instead, amazed and astounded by flies’ aeronautic wizardry. With a brain that’s home to several hundred thousand neurons—we each house 100 billion in ours—a fly can duck, dive, hover, rotate, and fly with easy accuracy and endurance, despite having a lousy field of vision and a carb-based Dumpster diet for fuel.
One of the fly-enamored among us is Harvard University’s Robert Wood, author of this issue’s ”Fly, Robot Fly,” in which he recounts efforts to build tiny flying robots based on the fly’s native wing-flapping skills. His own robotic fly, at 60 milligrams about the size of a chubby real fly, is the first of this robot class to become airborne.
The scientific study of insect flight has its roots in the second half of the 19th century. Queen Victoria was on the throne, technological innovation was in high gear, and the Industrial Revolution careered along with it. Natural history and flora and fauna worship were all the rage—although they would soon give way to more empirical disciplines like physics and science-based medicine—and museums and societies sprang up to serve this public interest.
Photography had also permeated the 19th-century zeitgeist. British-born Eadweard J. Muybridge became the first to isolate locomotion in his famous stop-action studies of humans and animals. Meanwhile, in France, physiologist, inventor, and chronophotographer Étienne-Jules Marey, who discovered that insect wings carve figure eights during movement, was inventing cameras and devising experiments to tease out the details of bird and insect flight. Some years later, his assistant and successor Lucien Bull invented the stereoscopic spark-drum camera, which took pictures at up to 2000 frames per second. Bull used his invention to make the first-ever movies of insect flight (go to http://www.expo-marey.com/indexFR.htmto see examples of their work).
Fast-forward to the late 20th century: the world went digital and cameras and experimental methods improved significantly, but questions about how flapping-wing flight works at fly scales remained essentially wide open.
Then in the 1980s and ’90s, an eclectic and far-flung cohort of researchers, among them Charles Ellington from Cambridge University, Michael Dickinson, now at Caltech, and Ronald Fearing from the University of California, Berkeley, set out to describe the kinetics of insect flight. Through ingenious experiments with live insects and dynamically scaled-up insect models, they managed to pin down some of the major mechanisms and physical forces that propel a fly through its airspace.
Now the hope is to use that know-how to build better houseflies: inexpensive, tiny robot flies that could work together, lots of them, on search-and-rescue missions, environmental monitoring, planetary exploration, military surveillance, and in virtually any situation in which it would be better or safer to send a batch of robotic flies instead of humans.
Professor Wood’s accomplishment is remarkable, but much remains to be done before you’ll see robotic flies buzzing over a fire or monitoring tornado damage in your neighborhood. Small doesn’t mean simple. The obstacles facing tiny robots’ flying and working in uncontrolled environments are as daunting as those faced by your pet AIBO. But as is often noted, flies have had a hundred million years or so to work out the kinks in their evolutionary flight plans. Perhaps in another 10 we’ll be able to say the same about their electronic counterparts.