MIT’s newest robot is cute, tiny, modular, and could run rings around you.
*robotic cheetah noises*.
Image credits Bryce Vickmark.
Researchers at MIT have developed a ‘mini cheetah’ robot whose range of motion, they boast, would rival those of a champion gymnast. This four-legged robot (hardly more than a powerpack on legs) can move, bend, and swing its legs in a wide range of motions, which allows it to handle uneven terrain about twice as fast as a human, and even walk upside-down. The robot, its developers add, is also “virtually indestructible” at least as falling or slamming into stuff is concerned.
Skynet’s newest pet
The robot weighs in at a paltry 20 pounds, but don’t let its diminutive stature fool you. The mini cheetah can perform some really impressive tricks, even being able to perform a 360-degree backflip from a standing position. If kicked to the ground, or if it falls flat, the robot can quickly recover with what MIT’s press release describes as a “swift, kung-fu-like swing of its elbows.” Apparently, nobody at MIT has ever seen Terminator.
But, the mini cheetah isn’t just about daredevil moves — it’s also designed to be highly modular and dirt cheap (for a robot). Each of its four limbs is powered by three identical electric motors (one for each axis) that the team developed solely from off-the-shelf parts. Each motor (as well as most other parts) can be easily replaced in case of damage.
“You could put these parts together, almost like Legos,” says lead developer Benjamin Katz, a technical associate in MIT’s Department of Mechanical Engineering.
“A big part of why we built this robot is that it makes it so easy to experiment and just try crazy things, because the robot is super robust and doesn’t break easily, and if it does break, it’s easy and not very expensive to fix.”
The mini cheetah draws heavily from its much larger predecessor, Cheetah 3. The team specifically aimed to make it smaller, easier to repair, more dynamic, and cheaper so that they would create a platform on which more researchers can test movement algorithms. The modular layout also makes it highly customizable. In Cheetah 3, Katz explains, you had to “do a ton of redesign” to change or install any parts since “everything is super integrated”. In the mini cheetah, installing a new arm is as simple as adding some more motors.
“Eventually, I’m hoping we could have a robotic dog race through an obstacle course, where each team controls a mini cheetah with different algorithms, and we can see which strategy is more effective. That’s how you accelerate research.”
Each of the robot’s 12 motors is about the size of a Mason jar lid and comes with a gearbox that provides a 6:1 gear reduction, enabling the rotor to provide six times the torque that it normally would. A sensor permanently measures the angle and orientation of the motor and its associated limb, allowing the robot to keep tabs on its shape.
It’s also freaking adorable:
This lightweight, high-torque, low-inertia design allows the robot to execute fast, dynamic maneuvers and make high-force impacts on the ground without breaking any gears or limbs. The team tested their cheetah through the hallways of MIT’s Pappalardo Lab and along the slightly uneven ground of Killian Court. In both cases, it managed to move at around 5 miles (8 km) per hour. Your average human, for context, walks at about 3 miles per hour.
“The rate at which it can change forces on the ground is really fast,” Katz says. “When it’s running, its feet are only on the ground for something like 150 milliseconds at a time, during which a computer tells it to increase the force on the foot, then change it to balance, and then decrease that force really fast to lift up. So it can do really dynamic stuff, like jump in the air with every step, or run with two feet on the ground at a time. Most robots aren’t capable of doing this, so move much slower.”
They also wrote special code to direct the robot to twist and stretch, showcasing its range of motion and ability to rotate its limbs and joints while maintaining balance. The robot can also recover from unexpected impacts, and the team programmed it to automatically shut down when kicked to the ground. “It assumes something terrible has gone wrong,” Katz explains, “so it just turns off, and all the legs fly wherever they go.” When given a command to restart, the bot determines its orientation and performs a preprogrammed maneuver to pop itself back on all fours.
The team, funnily enough, also put a lot of effort into programming the bot to perform backflips.
“The first time we tried it, it miraculously worked,” Katz says.
“This is super exciting,” Kim adds. “Imagine Cheetah 3 doing a backflip — it would crash and probably destroy the treadmill. We could do this with the mini cheetah on a desktop.”
The team is building about 10 more mini cheetahs, which they plan to loan to other research groups. They’re also looking into instilling a (fittingly) very cat-like ability in their mini cheetahs, as well:
“We’re working now on a landing controller, the idea being that I want to be able to pick up the robot and toss it, and just have it land on its feet,” Katz says. “Say you wanted to throw the robot into the window of a building and have it go explore inside the building. You could do that.”
I have to admit, the idea of casually launching a robot out the window (there’s a word for that, by the way: defenestration) with complete disregard, and having it come back a few minutes later with its task complete, is hilarious to me. And probably why they will, eventually, learn to hate us.
Still, doom at the hands of our own creations is still a ways away, and not completely certain. Until then, the team will be presenting the mini cheetah’s design at the International Conference on Robotics and Automation, in May. No word on whether they’ll be giving these robots out at the conference, but if they are, I’m calling major dibs.
