CHINA TOPIX

04/25/2024 04:26:50 pm

Make CT Your Homepage

MIT Develops Incredible Robot Cheetah that Runs up to 30 mph

MIT's robot cheetah can sprint up tp 30 miles per hour using a unique algorithm.

(Photo : Jose-Luis Olivares/MIT) MIT cheetah-bot experiment in Briggs Field. (From left) Sehyuk Yim, Joao Luiz Almeida Souza Ramos, Wyatt L Ubellacker, Sangbae Kim, Xu Sun, and Hae Won Park.

Researchers from MIT have developed a unique algorithm for bounding that enables a robotic cheetah to run free and jump untethered to any support frame.

With this algorithm, researchers can control the cheetah's bounding and adjust it to many speeds and triple the force when the robot performs jumps and obstacles.

Like Us on Facebook

"Robo- cheetah" is able to sprint at 10 miles per hour and jump over hurdles. It can accelerate to speeds of up to 30 miles per hour.

The secret of the algorithm is that it has a program for each of the robot's legs. Each leg exerts a certain amount of force the moment it hits the ground, allowing the robot to maintain a steady sprint at any speed.

Sprinters such as Usain Bolt, who is the fastest person on Earth, don't cycle their legs at a really fast rate to run faster. The secret is they increase the length of their stride, said Sangbae Kim, an associate professor for mechanical engineering at MIT.

This enables a human runner or a robot to maintain a steady frequency by exerting harder force downward, and at the same time, increasing their ground force.

By adapting this technique involving ground force, researchers allowed the Robo-cheetah to run on terrain like a grassy field.

Kim said the majority of robots are slow and heavy because they don't control force at any given high speed situation.

The MIT cheetah is unique because it enables control of this force, even for a short period of time. This control results in a powerful impact on the ground and produces an agile and dynamic robot.   

The robot's machinery includes high-torque-density electric motors controlled by amplifiers that enhance the biomorphic inspired legs. It also adapts well to rough terrain.

This algorithm makes it easier to determine how much force each leg needs to exert within a short period of time it makes contact with the ground.

Real Time Analytics