This theoretical, computer-generated walker – devised by engineers Mario Gomes of Rochester Institute of Technology and Andy Ruina of Cornell University – has the most energy-efficient gait of any ever discovered.
Even state-of-the-art robotic walkers require tens of kilowatts to move. And people with prosthetic feet and legs use on average about 25% more energy when walking than people with natural legs.
Watch the walk.
Nature News explains: The key to the gait's success is that the walker's leg is stationary when it touches down and takes the weight.
"It has a perfectly soft landing," says Ruina, who was surprised to find that any kind of walker, even a theoretical one, can move without losing energy to collisions with the ground. "It lets you know that there's no fundamental limit on what you can do."
But it’s still just a simplified 2-D model at the moment, designed with rigid legs and a rigid torso connected by springs, with no friction in its joints. Not to mention it moves its swinging leg through the ground as it walks, an abstraction made necessary by its lack of actual knees.
Using those idealizations, the researchers found a mathematical solution that described a gait that wouldn’t lose any energy when the walker's feet touched the ground and began to take on weight from the other leg.
"It's like a wheel that would roll forever," Ruina says. "It's a crazy-looking thing."
Building a robot that copies the walker's gait exactly is impossible, but Gomes is inspired to build one loosely based on it – a design that consists of a rimless wheel made up of spokes, with an oscillating weight in the middle that can adjust the speed of each spoke to be near-motionless as it hits the ground and takes the weight. "It just kisses the ground and there's no energy transported," Gomes says.
The duo suggests that the need to minimize energy lost when the feet collide with the ground may explain the pendulum-like gait of women of the Luo people of Kenya, who walk many miles carrying heavy loads on their heads.
This post was originally published on Smartplanet.com