Cheetah Robot Jumps Over Hurdles Without Slowing Down

Researchers have created new algorithms to autonomously jump over obstacles with the MIT Cheetah 2 robot.  In a leap for robot development, the MIT researchers who built a robotic cheetah have now trained it to see and jump over hurdles as it runs — making this the first four-legged robot to run and jump over obstacles autonomously.

http://i.imgur.com/i6N6his.gif
Image credit: Haewon Park, Patrick Wensing, and Sangbae Kim

There are now even fewer obstacles separating you from your new robotic overlords. Researchers have created a robotic cheetah that can—for the first time—autonomously detect objects and jump over them without slowing down. The team of scientists from the Massachusetts Institute of Technology (MIT) have demonstrated how the futuristic four-legged robot is able to clear half-meter (18 inch) hurdles, while moving at a speed of 5 mph, in the video below.
Before jumping over a hurdle, the robotic cheetah plans out its path in a similar way to a human runner. It estimates the height and distance of the object it’s approaching, decides on the best position to jump and exerts the force needed to leap over the object and land safely.

The team first tested the robotic cheetah’s jumping ability on a treadmill, where it was tethered in place. It was able to jump over 70 percent of the hurdles. They then tested the robot on a track, where it was able to jump over 90 percent of hurdles. The robot cheetah was able to even clear an 18-inch-tall obstacle, which is more than half of the robot’s own height.
“A running jump is a truly dynamic behavior. You have to manage balance and energy, and be able to handle impact after landing. Our robot is specifically designed for those highly dynamic behaviors,” said Sangbae Kim, an assistant professor of mechanical engineering at MIT, in a statement.
The latest robotic cheetah is a significant upgrade to the last design as it now possesses a visual system that allows it to "see." The data collected from this vision is fed into a three-part algorithm that allows the robot to plan out its path, gaining full autonomous control.
Researchers will continue to work on the cheetah robot, testing it on more complex environments like grass. It will also be entered into the DARPA Robotics Challenge for a chance to win the $3.5 million prize money.


Full text of the article in the IFL Science and MIT News.