What We Can Learn
from Animals


getmedia/34ec85b9-24bc-4163-884d-f2e7d81b8393/What-We-Can-Learn-From-Animals_Robotics-60th.jpg.aspx?width=60&height=60&ext=.jpg

Sangbae Kim, assistant professor of engineering at the Massachusetts Institute of Technology (MIT), has made a career out of his fascination with the movements of animals. It's led to "Stickybot" for one thing, a robot named one of Time's "Best Inventions of 2006" and inspired by the gecko. He also believes it could lead to breakthroughs in many areas of industry and science.

"The freedom and parallel processing animals have and how they control their muscles to run so fast is tremendous," Kim says. "This is about learning from biology, not just copying animal principles but extracting the principles to see the effect."

Stickybot

For Stickybot, it was looking at how lizards managed to stick to vertical surfaces and climb as efficiently as they do. Kim was attempting to mirror it with a pad-like attachment. As explained on his Stanford website from his days pursuing his PhD there: "Stickybot is (a) quadruped robot capable of climbing smooth surfaces, such as glass, acrylic, and whiteboard using directional adhesive … The robot is operated by 12 servo motors controlled by a PIC controller with force sensors. Stickybot is the incarnation of biomimetic design constituents that I invented, including an underactuated hierarchical system, cable-driven actuation, and a passive, compliance-based force control scheme… After a year's furious effort on a foot design, push-pull cable actuation mechanism, Stickybot succeeded to climb a glass surface reliably."

What We Can Learn From Animals - Robotics

Professor Sangbae Kim.

The site goes on to explain that the sticking is done with little preload and lets go by reducing the load.

What became entertaining to some was just one more stop along the road for Kim, noting that the possible applications for animal-inspired robots include search and rescue and prosthetics. "I want to know how that biosystem works and perfectly implement a mechanical system through it," he says. "Imagine if you learn every piece of a movement that's been going on for millions of years and can understand the complete methodology. 'If I want to do the same thing, how would I do it?'"

What We Can Learn From Animals - Robotics

The gecko was the inspiration for the Stickybot.

If it sounds like fun, Kim assures you that it is. Though there can be many hours involved, "edutainment" is how he describes it. "I get to combine two things so many people love: animals and robots."

Many Mysteries

Kim believes one of the major failings in the research sector is a lack of attention to animals and the mysteries they can unlock. "You have to remember," he says, "that we've only found a fraction out about our own bodies. It's the same in the natural world."

And Kim, of course, is right. No matter how technologically advanced airplanes may be, they pale in comparison to the abilities of a bird, one not waiting for clearance from the tower for takeoff. But despite Kim's robots sometimes being compared to specific animals (he's right now studying the cheetah and monkey closely), he cautions that it's not truly about finding one animal's abilities but common ones among many: "When you find that it makes for a greater chance at application," he says. "My ultimate goal is to help people with all my research."

Eric Butterman is an independent writer.

I get to combine two things so many people love: animals and robots.

Prof. Sangbae Kim, Massachusetts Institute of Technology

getmedia/34ec85b9-24bc-4163-884d-f2e7d81b8393/What-We-Can-Learn-From-Animals_Robotics-60th.jpg.aspx?width=60&height=60&ext=.jpg

June 2012

by Eric Butterman, ASME.org