Sunday, July 13, 2008
News

Spy bugs

Georgia Tech scientist designs an artificial insect that can search for Osama bin Laden or signs of life on Mars

By Josh Clark

So you’re making a pretty good living in your native Afghanistan supplying the local farmers’ poppy crops to the Vietnamese, who, in turn, process it into heroin for sale in London, Berlin and New York. Then, everything sours. But the dope’s not the problem; it’s all that funding you’re providing local Taliban insurgents that has come back to bite you.

       This is exactly how you ended up spending your days in mountain caves, talking about the intricacies of head scarves with the same fanatics you gave all that money to. You notice an insect as it buzzes into the cave. While you’re getting a bit tired of all the rhetoric, you trust these guys know how to hide. They’ve been doing this for years. Which is why you’re so surprised when the cave is carpet-bombed by American F-16s. They found you, it turns out, using what you’d thought was a dragonfly.

       You, my friend, have just learned a hard lesson about the emerging field of biomimicry. This fairly recent discipline of design and engineering takes its cues from nature. Remember that myth about how a bumblebee’s lumbering ability to fly defies the laws of physics? Yeah, that’s not true, but biomimicry bees are aerodynamic superstars. Shark skin is excellent at reducing drag through the water. Termite mounds contain built-in air conditioning.

       Instead of reinventing the wheel to achieve their ends, engineers and designers are lifting their newest ideas from 4 billion years of evolutionary progress. So shark skin is the basis for new speedboat hulls, termite mounds are models for no-electricity air conditioning in a Harare office building, and Mercedes produced the Bionic—a diesel concept car based on the streamlined boxfish. And a dragonfly serves as the model for a device designed to search for Osama bin Laden—or anyone else who’s difficult and dangerous to find.

          There’s probably no better person to talk to in Atlanta about artificial insects than Robert Michelson. The Georgia Tech electrical engineer and aerospace design researcher is at the leading edge of the field.

       His designs, he insists, aren’t biomimicry.

     “It’s bioinspiration,” says Michelson. “Biomimicry is when you just copy something you see in nature.”

That mode of design, says Michelson, doesn’t always work; what nature wrought can’t always be applied to what the engineer means to achieve.
      
Take the motion of a winged insect. “We can’t replicate the tiny muscles that do all of the things they do in an insect wing,” he says. What can be done is “seeing how it works and figuring out how it can help something we create.”
     
 Which is precisely what Michelson did to create his “Entomopter.” The design of this unmanned craft isn’t exactly found in nature, nor is it exactly found in any other existent technology on the planet. It’s a new hybrid, a joint venture between man and nature.
      

Flying robots

Michelson doesn’t remember the precise moment he became inspired to create autonomous flying vehicles based on the flight of insects. He mostly remembers the robots.
      
“As the field of unmanned aerial vehicles emerged, I saw that as an interesting application for robots,” Michelson says. The problem was, he was a bit ahead of the curve. At the time, one could create an unmanned aerial vehicle that propelled itself with flapping wings like an insect. Still, all one had was a very small, unusual vehicle—pretty cool, but not necessarily something for which the world was aching.
     
The challenge for Michelson and the handful of other scientists working toward the same ends at the time was to create a flying machine that could think for itself; a flying robot. The challenge still remains, and Michelson and his fellow aerospace researchers have pulled in other fields of research.

“The intelligence part is something that’s being developed in parallel,” he says.

The field of creating thinking machines has had its breakthroughs. You can find sets of sensors on 2009 Mercedes that determine whether you’ve nodded off at the wheel. The sensors feed their separate bits of information to an onboard computer. The key is algorithms, mathematical formulae that say, “If this, then this.” For example: If this guy’s slowly drifting into another lane, and suddenly jerks the wheel, it’s 1 a.m. and he’s driven from Atlanta to Orlando without stopping, then he’s probably asleep. Alarms go off.
      
Affixing these sensors and connecting them to a computer on board a 3,000 pound Mercedes SLK is one thing. Outfitting a 25-gram aerial vehicle that looks like a dragonfly with a similar system is another. Still, Michelson remains confident that the future is bright for the convergence of intelligence and his machine: “It’s really just a function of dollars to shrink it down to the size of an Entomopter.”
        
Which is fairly small. His tiny yellow prototype looks like a miniature biplane. Its wingspan is about six inches across, and at the heart is what Michelson calls its reciprocating chemical muscle. As a cylinder located about where the pilot and passenger would sit in a biplane pumps back and forth, the wings flap up and down. It looks a bit like a pygmy Sybian machine that can fly with a Tech Yellow Jacket sticker planted proudly on its left wing.

The whole operation is powered with highly concentrated hydrogen peroxide—that’s right, the stuff you clean your wounds with. Onboard, the hydrogen peroxide is converted into gas. This change of state produces two important forms of energy: steam and heat. Both are used to power the reciprocating muscle, which pumps away and can create wing flapping at 60 hertz (that’s really fast).
      
With such a small instrument, though, Michelson had to get efficient, figuring how to get the absolute maximum power and function from a minimum amount of gas. Fully fueled, the craft nearly doubles in weight. The added weight decreases fuel efficiency, and thus cuts down on the amount of time the Entomopter can spend rooting out insurgents hiding behind stalactites.

So Michelson figured out ways to make his aerial vehicle more efficient. The machine focuses the waste gas produced by the chemical conversion of the peroxide from liquid to gas into an ultrasonic tone, like a bat’s, to provide the Entomopter with an echo locator with a 12-foot radius, which keeps it from running into objects. Other waste gas is emitted in puffs at joints in the wing to allow the vehicle to turn left or right. The rest is sent to these same joints to lubricate moving parts and reduce drag. All together, it’s one of the more efficient machines flying around campus these days.
      
At a time when the U.S. is entrenched in an unpopular war, such unmanned vehicles are increasingly appealing. Why send human soldiers who have families and friends to do a job a robot can do? Who cares if it’s shot down? It’s just money. So technology like Michelson’s has captured the attention of the military.

DARPA, the military’s Defense Advanced Research Project Agency, has thrown funding at the project. But the Entomopter has also gained the attention of NASA. Michelson’s reciprocating chemical muscle solved a problem that had been plaguing NASA’s Mars missions.

Bugs in space

      The red planet atmosphere consists of 95 percent carbon dioxide, and only about less than half of 1 percent oxygen. This means that you couldn’t light a match on Mars; the mixture of gas won’t support combustion. So any craft that uses a combustion engine wouldn’t fly. This rarefied atmosphere is also thin; gravity on the planet is about 37 percent what we experience on Earth. 

To even stay in the air, any vehicle with fixed, unmoving wings would have to fly really fast constantly—about 300 kilometers per hour (around 186 mph). Any dip in speed, even by 5 km/h, says Michelson, and the craft would “drop like a rock. It’s hard on Mars to have to go 300 kilometers per hour, because you zip right past what you’re looking for.”
      
The Entomopter provides a clear advantage in the Martian atmosphere over fixed-wing aircraft. Its reciprocating muscle is powered by a chemical conversion, not combustion, and it can flap its wings at a rate of 300 km/h while the body of the craft, and any sensors it may have aboard, pass over the planet’s surface at a relatively low rate, allowing the controllers back in Houston or Cape Canaveral to get a good view of what they’re looking at, mark spots and lead rovers on the surface to get a closer look or a sample. Robert Michelson’s Entomopter may be the first low-altitude Martian scout.
       
But it’s going to take a larger version of the machine to stay aloft in the atmosphere of Mars. Michelson estimates a NASA version would have a three-foot wingspan instead of the current six-inch terrestrial version. He’s tinkering with it now. NASA has a technology-readiness level system, a series of numbers assigned to any tech the agency is interested in, from one to nine. A No. 1 means an idea’s been born; No. 9 means the technology has been proven in missions. Michelson estimates his Entomopter is about halfway there.
      
And there are a myriad of other applications for the unmanned vehicle. In addition to flying over Mars’ surface and into Afghan caves, it can also be sent into buildings to check on the condition of hostages (as well as the location of the hostage-taker). An Entomopter outfitted with an olfactory sensor can be a mechanized canary in the proverbial coal mine, sensing and reporting on toxic gas levels in enclosed spaces.
      
Michelson says he isn’t concerned where his machine ends up; he’s not lobbying heavily for space or terrestrial use.

“It’s the technology I’m more interested in, not the applications,” he says. “I’m more interested in seeing it work.” SP