May 3, 2017 @ 20:21 |
A bipedal robot, called Planar Elliptical Runner doesn’t need orientation sensors and computer brain to balance itself, its mechanical design is sufficient to provide dynamic stability as it runs.
Inspired by Ostrich, Institute for Human and Machine Cognition (IHMC) researchers have developed a bipedal robot, called Planar Elliptical Runner. In this project, the team of researchers tried to explore the different mechanical designs for bipedal locomotion.
Now in a demonstration video they came up with a novel mechanical design that doesn’t use sensors and computer brain to balance the robot. Instead, its mechanical design provides dynamic stability as it runs. “All the intelligence is in the physical design of the robot itself,” says Jerry Pratt, a senior research scientist at IHMC who leads the team that developed the robot to MIT Technology Review.
The Planar Elliptical Runner is a running robot that can run about 12 miles per hour on a treadmill. It is sandwhiched between two plates of glass, which keep it in the sagittal plane. It cannot turn or tip sideways, but it can tip forward/backward and fall down.
The robot is “open loop stable”.
There is only 1 motor driving the legs. There are no sensors or computers on board. An RC car radio controller determines how much power to apply to the motor. Give the robot more power and it runs faster. Less power and it runs slower.
In simulation researchers are exploring how to extend this concept to 3D running. Initial simulations are showing promising results.
- ATRIAS: “Spring-Mass” Technology Heralds the Future of Walking Robots
- PROXI: 20 percent High Efficient Humanoid Robot
The design might feed into future systems. “We believe that the lessons learned from this robot can be applied to more practical running robots to make them more efficient and natural looking,” Pratt adds. “Running will be eventually useful for any application that you want to do quickly and where wheels can’t work well.”
Agility Robotics, a spin-off of Oregon State University has also developed a same kind of dynamically stable bipedal robot named Cassie.
University of Michigan is also developing advanced algorithms that allow for more efficient and graceful dynamic locomotion.
Machines that balance themselves dynamically can traverse difficult terrain, but they are complex, expensive, and power-hungry.
Keywords: Bipedal Robots, Mechanical design, Locomotion
- Image-1: IHMC