Until recent times, pilots often wondered – and discussed at length – how flies land on ceilings.
Do they do a half-roll first? A reversed split-s? Half of an inside loop?
A few years ago, high-speed cameras finally taught us how the maneuver is done: Flies approach the ceiling in a shallow climb, contact it with their forelegs, and flip over. It’s pretty neat to watch and tempting to put it all down to instinct, but even more recent research suggests that flies are executing a complex calculation involving trajectory, velocity and closing speed when they pull off an upside-down landing.
I don’t want to imply that pilots are fixated on flies except when the pesky critters are buzzing around the barbecue, but another interesting topic has been the agility of those immeasurably irritating insects when we attempt to swat them. How do they escape the folded newspaper or fly swatter so frustratingly often?
At the California Institute of Technology, high-speed cameras have shown that flies react very cleverly to objects, such as a fly swatter, coming at them. Rather than simply fly straight off, which could be fatal, they do a lightning-fast two-step, hopping into the air then taking off almost vertically. Furthermore, they can vary the takeoff angle to suit the circumstance, sometimes jumping sideways, flying backward or sideways, and accelerating at an astonishing 20 to 30 Gs.
Christopher Hounsfield, editor of “Aerospace Testing” magazine, attributes this to “a super-light body, the ability to turn a minute amount of liquid food into limitless muscular energy, and 400 million years of evolution.” Although he claims to distrust anything that “can taste with its feet, liquidizes its food…and can pass on typhoid with one hacking cough,” he gives its “acute aviating ability” deep respect.
Meanwhile, Julie H. Simpson, at the University of Wisconsin in Madison, is trying to produce a functional map of fly brains which, she says, have only 200,000 neurons compared to our 100 billion. Even so, she told “Forbes” magazine, the work could consume the rest of her life – and she’s only 32. About two-thirds of our DNA has an equivalent genome in flies, she says. Their brains are “organized much like ours, with activities such as sensory perception assigned to distinct regions.” She hopes that her studies eventually will lead to an understanding of human memory.
Why all this scientific interest in flies?
Imagine, if you will, an artificial fly able to navigate around rooms and through doors, perhaps autonomously, attracting no attention yet agile enough to dodge a fly swatter if it does annoy someone. It would be of nearly-incalculable value to the military and in law enforcement.
Those Cal Tech studies are in pursuit of building just such an autonomous flying machine, one that actually mimics a fly’s agility and gymnastic skills. So far, however, the Cal Tech group and others working along similar lines have had little success. Those 200,000 neurons in a fly’s brain, it appears, are incredibly focused on aviating – even more so than most pilots – and until we can build an electronic brain the size of a grain of sand with the single-mindedness of a real fly, an artificial fly seems likely to elude us.
On the other hand, researchers at the University of Florida are having great success with their artificial seagull. Anyone with waterfront property and a wharf might wonder why the interest in seagulls, which are notoriously messy, but the Florida team admires gulls’ flying ability and their version excretes nothing.
The team’s gull-drones are intended for urban surveillance missions not essentially different from those planned for artificial flies. Funded by NASA and the U.S. Air Force, the agile artificial birds can zip around in tight places and detect all sorts of interesting things with a range of sensors. The researchers say their drones could be on real missions in a couple of years.
Dr. Rick Lind, who leads the mechanical and aerospace engineering team, says that the craft can alter wing shape much as real gulls do, allowing them to change direction rapidly and perform gull-like aerobatics. “It isn’t such a challenge to get wings to change shape,” he said. “It is more of a challenge to do it under autopilot control,” which is the idea. “The ultimate aim is for it to fly by itself through cities to search for (such things as) bio-agents. The vehicles will need to identify unexpected obstacles, re-plan the flight path and go on with the mission,” he said.
A lot like real pilots. Or real seagulls. Or flies, for that matter.
If we can build autonomous man-made flies and seagulls with half the capabilities of the real things, consider what a successful nano-electronic fly brain might do for avionics.
Anyone want to land on the ceiling?
Thomas F. Norton is GAN’s Senior Editor.