A cheetah can run faster than any other animal. A gecko's feet can stick to almost any surface without using liquids or surface tension. And some roaches scurry at nearly 50 times their body length in one second, which, scaled up to human levels, can be around 200 miles an hour.
The wonders of the animal kingdom are not just for fans of National Geographic. Robotic designer Sangbae Kim, a professor at the Massachusetts Institute of Technology, is trying to understand how he can take some of the mechanisms animals use and replicate them in robots.
The animal kingdom provides the best ideas for creating mobile robots, says Kim. Locomotion and movement are the core parts of an animal's life. "Animals have to find food, shelter; move towards water or away from a predator," he says.
"Moving is one of their biggest functions, and they do it very well. That's why ideas from nature are very important for a robotic designer like me."
Mechanical design derived from biological models is something Kim has been working on for years, first at Stanford University and now at MIT. The simplification and adaptation of the fundamental design principles seen in animals has led to the creation of his bio-inspired robots.
Among the robots Kim and Stanford professor Mark Cutkosky have designed are the Stickybot, a robot that has foot pads based on a gecko's feet, and iSprawl, a robot whose motion is inspired from cockroaches.
Kim's latest project is a robot inspired by the cheetah. The idea is build a prototype robot from a lightweight carbon-fiber-foam composite that can run at at least half the cheetah's top speed of 70 miles per hour.
It's an ambitious project. Current wheeled robots are efficient, but can be slow in rough terrains. For instance, iRobot's PackBot, which is used by the U.S. military, can only travel at speeds of up to 5.8 miles per hour.
"Most wheeled robots today can do very well on flat surfaces, but they are slow," says Kim. That's why he's looking to the cheetah for ideas. The cheetah has an extremely flexible backbone that gives extra speed or force to its running motion.
Over the next 18 months, Kim and four MIT graduate students will start building and testing prototypes. The first step will be to create a computer model to calculate the optimal limb length, weight, gait and torque of the hip and knee joints.
The biggest challenge in this project won't be the structure, but getting enough power from a motor to get to the desired speed quickly, says Kim.
Before the robotic cheetah came Stickybot, a mechanical lizard-like robot that takes its inspiration from the gecko. Geckos can climb walls at almost the same speed — of about 1 meter per second — at which they run on the ground. This remarkable ability makes it the perfect animal to draw upon to create a climbing robot, says Kim.
The secret to the gecko's agility is that it uses a phenomenon called directional adhesion, or stickiness in just one direction, to adhere to walls.
"The gecko's feet can detach very easily as it moves forward," says Kim. "If you take normal sticky tape and press it to the wall, you will find it is tough to detach it quickly. Directional adhesion solves that problem."
The pads of a gecko's feet are covered with tiny hairs called setae and spatulae that can be up to one-thousandth the width of a human hair. The hairs cling to surfaces using molecular interactions known as the Van der Waals force. The force helps support the gecko's weight as it scrambles up vertical surfaces.
Kim has tried to recreate that idea for the Stickybot. The Stickybot's feet is covered with hairs made of rubber silicone. The rubber is thicker than those on a gecko's paw, however, which limits the robot's abilities. It can only climb extremely smooth surfaces such as glass, acrylic or a whiteboard.
Kim says his team is working on refining the Stickybot so that it can adapt to climbing on walls with uneven textures.
If the Stickybot can be improved, there are plenty of applications for it, such as repairing of underwater oil pipelines or even window washing.
How good are claws when it comes to climbing? Kim and his colleagues tested the idea when they created Spinybot, a hexapod robot that would use small spines or micro claws, as they called it, to produce adhesion on a surface. The approach is inspired by the mechanisms observed in spiders, says Kim.
Unlike the claws of a cat, small spines do not need to penetrate surfaces. Instead, they exploit small bumps or pits in a surface to move forward.
Each of the Spinybot's feet has 10 toe mechanisms with about two spines per toe. Each toe mechanism can stretch independently of its neighbors to distribute the load. The robot also has a tail that reduces the forces required at the front limbs.
The SpinyBot technology has been successful enough for the team to start working on adapting it for a heavier robot.
Cockroaches aren't anyone's favorite creatures but most of us have watched them scurry away at amazing speeds.
Roaches don't control their legs very carefully, says Kim. They have six small legs that are thrown about 15 times a second. "They are relying a lot on their mechanical property to move forward," he says. "At the same time it's also not about being extremely precise in how they place their legs."
Studying the movement of roaches led to the development of hand-sized hexapedal robots or a new family of 'sprawl' robots. The robots are designed to test ideas about locomotion dynamics, leg design and leg arrangement.
iSprawl has a battery and electric motor, and a power transmission system that converts rotary motion to leg thrust. It also has a push-pull cable transmission system.
The iSprawl, which was the first of the bio-robots designed by Kim, can cover 7.5 feet per second.
Photos: Sangbae Kim; Stickybot (Mark Cutkosky/Stanford)
