Watch: Robotic Legs Walk Like a Human

  • Share
  • Read Later


Researchers at University of Arizona have created a set of robotic legs that mimic a real human gait.

The research team believes the legs are the most biologically accurate model of human walking created. So, how’d they do it? The team gave the robot a computerized version of the neural network that humans use to move and groove, called the central pattern generator, or CPG. The network, which lies in the lower part of the spinal cord, gathers sensory information from other parts of the body — like how much strain a muscle experiences or how much weight a foot is bearing — and then responds by producing rhythmic muscle signals. This is what allows people to walk naturally, without needing to think about it.

(VIDEO: A Faster Stride Predicts a Longer Life)

The robot’s version of the CPG also gathers information from sensors in the legs and feet to mimic human walking. The researchers say the robot will help them better understand human movement — how babies first learn to walk, for example, or how spinal-cord-injury patients may regain their walking capabilities through spinal cord stimulation.

The next step in the research will be to incorporate vision and other tactile sensors to the robot, “so that if you stumble, the system will correct itself and not fall over,” researcher Anthony Lewis of University of Arizona’s department of electrical and computer engineering told the AFP.

The research was published in the Journal of Neural Engineering.

MORE: How Exercise Can Change Your DNA