MEDIA TOUR

Newspaper: DALLAS MORNING NEWS

Title: FROM ANT TO ZEBRA, CREATURES MARCH TO SAME DRUMMER

Author: Matt Crenson

Rabbits hop, turtles plod, peacocks strut and cockroaches skitter. In the animal world, there appear to be thousands of ways to get from Point A to Point B. But for all nature's variety, recent research shows that every beast with legs, from ant to elephant, moves the same way.

After years of studying a menagerie of animals -- humans, horses, crabs and cockroaches among them -- biologists have come up with a unified theory of legged locomotion: Animals' running legs work like pogo sticks.

Biologist Robert Full made the case for the pogo-stick theory in Durham, N.C., recently and suggested ways that the idea could be applied to robot design, paleontology and even moviemaking. The University of California, Berkeley, professor illustrated many of his points with movies and stop-action photographs of two-, four-, six-, eight- and many-legged animals walking, trotting, running and galloping. ``All of these animals are bouncing along as they move,'' he said at a meeting sponsored by the Council for the Advancement of Science Writing. It doesn't matter how many legs an animal has, how they're constructed or how the body is connected to them. It makes no difference what an animal's skeleton is made of. Legs basically work like sticks with springs attached, Full said, flexing to absorb energy and then releasing it to propel the animal forward.

He and other researchers developed the pogo-stick theory by doing a natural experiment. Instead of manipulating nature to test a hypothesis, a natural experiment looks for cases that already differ in whatever character a researcher wants to understand better. In the case of legged locomotion, Full and his colleagues looked for animals that move in unusual ways --crabs, kangaroos and centipedes, for example. Despite their apparent differences, they all run the same way. ``This turns out to be an extraordinarily powerful method,'' Full said.

But it's not always pretty. One of the most valuable experimental tools for studying locomotion is the 6-inch-long Madagascar hissing cockroach, Full said. Another species, known as the death-head cockroach, also makes a good subject.

``In most cases they're disgusting animals that we look at. But they do reveal nature's design lessons,'' Full said.

They also misbehave. In one set of experiments, Full used a special substance to determine how cockroaches exert forces on the ground. When a cockroach stepped on the substance, the material conducted light differently, recording the direction and magnitude of the step's force with a dark spot.

Although the setup has yielded a wealth of valuable information, early trials were less than successful.

Full explained that the high-tech substance, known as a photoelastic polymer, goes by a different name in the grocery store -- Jell-O. So when he and his colleagues first set up their experiment, the cockroaches didn't run at all. They chowed down.

The problem was, the experiment was conducted with orange Jell-O. ``Now we use unflavored Jell-O. It works much better,'' Full said.

The Jell-O experiments, along with high-speed films of cockroaches, centipedes, crabs and other animals running on treadmills, clearly showed Full that animals don't progress smoothly as they run along.

``They were hitting the brakes and then stepping on the gas,'' Full said, like a pogo stick lurching foward with each bounce.

Even a centipede, with dozens of legs, lurches forward as it walks. Progress that appears smooth in real time looks like a rhythmic sequence of pushing forward, then pausing, then moving along again.

``Anything that works like that can be described as a spring or a pogo stick,'' said Tom McMahon of Harvard University in Cambridge, Mass.

The first part of each step is the release of energy as the spring uncoils, throwing the pogo stick and its rider up and forward. Then when the pogo stick hits the ground, the spring compresses, storing energy for the next hop.

Full has demonstrated that legs do the same thing, storing energy from one step and then releasing it on the next. And McMahon's research has shown that in a wide variety of animals, legs have the same amount of stiffness no matter what they're doing.

``The leg working as a spring doesn't change its stiffness,'' McMahon said. ``It just works as an ordinary rubber band. And that was a big surprise.''

By looking at locomotion from a biomechanical point of view, the researchers have shown that some animals are really running when it looks like they're walking. McMahon dubbed the process ``Groucho running,'' because the mustachioed Marx brother can be seen doing it in old movies. When Groucho running, an animal uses the leg as a spring, but never lifts more than one foot off the ground at once.

``It looks like the person is walking, but he's not,'' McMahon said. ``That's because in walking the body is highest in the middle of the time that the foot is in contact with the ground.''

But in Groucho running, a person -- or any other animal -- is lowest in the middle of each step, just as in normal running. Running used to be defined as locomotion in which at least one foot is always off the ground. But biologists now think of running as locomotion in which the leg acts like a spring, storing energy from one step and then releasing it on the next.

Walking, on the other hand, is more like rolling an egg end-over-end, McMahon said. With each step, a walking animal pushes itself over its extended leg like an egg being rolled over one of its ends. Then the animal falls downward onto another leg, or legs.

For the last few years, engineers have been applying concepts such as Groucho running and egg rolling to robots. Most robots can't walk very well, Full contends, because they basically work as if they had wheels.

For example, a four-legged robot shifts its center of gravity so it can balance on its two rear legs and one front leg. Then it lifts the front leg that's no longer needed for balance and sets it down farther forward. Then it pulls the rest of the legs forward, one by one, never lifting more than one leg off the ground.

``Dynamically, [robots] are more similar to a turtle than any other animal,'' Full said.

Then he showed movies of one-legged robots that move like pogo sticks. The robots, built by engineers at the Massachusetts Institute of Technology, bounded down hallways and hopped over small obstacles. They looked so lifelike that the crowd cringed whenever one of the one-footed robots toppled over.

The goal of robot design isn't to imitate animals, of course. But Full said his research can help engineers design better robots by revealing the basic principles of legged locomotion.

And it can also help people who do want to imitate animals. For example, Full said, he wasn't too impressed by the dinosaurs in ``Jurassic Park.'' Animators drew those dinosaurs using computer programs that fine-tuned the creatures' motion until it was as real as the artists could get it.

But the dinosaurs didn't obey the basic physical laws he's discovered, Full said, so they didn't look completely real. By building computer animation programs that move animals around essentially like pogo sticks, using real information about their limb strength, weight distribution and other properties, he and his colleagues have been able to generate stunningly realistic animations of creatures running, jumping and falling.

The characters in the movies don't look like much. The stars are imaginary animals that look like headless ostriches and kangaroos. But Hollywood is interested, Full said, because their motion is so real.

Using the same idea, Full said, the physical principles he has discovered could also be used to learn how dinosaurs, mastodons and other extinct creatures moved. In a recent issue of the Journal of Experimental Biology, he and biologist Anna Ahn of UC-Berkeley described a computer simulation of a cockroach leg that attempts to mimic the mechanics of the real thing.

``We are encouraged by the concordance of the model's prediction and the `in vivo' measurements on the animal,'' the biologists wrote.

Because the computer simulation imitates the motion of living animals so well, presumably a similar one could be designed for extinct creatures. Using information from fossils, biologists might estimate how fast a tyrannosaur could run or how high a velociraptor could jump.

Full didn't expect that he'd be dabbling in paleontology, computer animation and robotics when he began his animal-locomotion research.

``It's only in the last three years or so that I've seen how these principles could be applied,'' he said.

But he said the fact that nature has provided him with such a wealth of information illustrates an important lesson: ``Even weird and disgusting animals should be preserved, because you can never predict where this research will lead.''