Researchers Build Fast Running Robot Inspired by Velociraptor
Image: KAIST Mechatronics, Systems, and Control Lab

UPDATE May 30 1:09 pm EST: The KAIST folks tell me Raptor momentarily achieved 48 km/h, but they decided to use 46 km/h as its official top speed because that's the speed the robot can run in a completely stable manner. And there appears to be some conflicting reports on whether Cheetah's top speed is 45.5 km/h or 47 km/h (I contacted Boston Dynamics trying to clarify but haven't heard back). So given the uncertainties involved, we're declaring this a tie. At least until we can have our robot race!

Researchers have long been interested in fast-running robots with powerful, agile legs. In particular, several groups have focused on bio-inspired designs based on cheetahs.

But when a team at the Korea Advanced Institute of Science and Technology (KAIST) set out to create a new sprinting robot, they didn't look at big cats; instead, they found inspiration in a completely different kind of creature: a velociraptor.

The robot, called Raptor, has two nimble legs and a mechanism that mimics a tail. In a recent experiment, it achieved an impressive speed of 46 kilometers per hour on a treadmill.

That's faster than the fastest human, the Olympic sprinter Usain Bolt, whose top speed has been estimated at 43.92 km/h.

And it's almost as fast as the world's fastest legged robot, Cheetah, built by Boston Dynamics, which reached 47 km/h in a trial two years ago, also on a treadmill.

But before you suggest a robot race between the two machines, note that both Raptor and Cheetah are attached to beams that keep them steady and prevent them from running off the treadmill—and crashing through a wall.

Image: KAIST Mechatronics, Systems, and Control Lab

And despite their similar speeds, the two robots are very different. Cheetah is a hefty quadruped, powered by hydraulic actuators. Raptor is a compact, 3-kilogram machine, with two legs made of lightweight composite material.

Another difference is that Raptor has a tail. Although it looks nothing like a Velociraptor's tail, it works like one, moving rapidly to help the robot maintain its body stable while stepping over obstacles, says Jongwon Park, a PhD student at KAIST's Mechatronics, Systems, and Control Laboratory.

Park developed Raptor along with colleagues Jinyi Lee, Jinwoo Lee, Kyung-Soo Kim, and Professor Soohyun Kim.

A recent video of Raptor shows the robot progressively increasing its speed until it reaches the 46 km/h mark. The video also shows a researcher placing obstacles on the treadmill. Thanks to its tail, Raptor doesn't trip on them.

Using a tail for stability is not a new idea in robotics. MIT's cheetah-inspired quadruped has a mechanical tail, and a UC Berkeley wheeled robot features a tail that is also based on a velociraptor's.

Raptor is not the first robot to sport legs with prosthetic blades, either. Another biped running robot, Athlete, built several years ago, also used carbon-fiber running blades.

Still, Raptor stands out for its simplicity, showing that it's possible to achieve significant speeds with an uncomplicated design.

Unlike other legged robots that rely on multiple actuators, Raptor uses just one motor per leg, which consists of a nine-bar linkage. To recover part of the energy used to make the robot move, the researchers added Achilles tendons that work as springs, absorbing and restoring energy with every step.

The robot's control system is also simple. Raptor runs a computer program known as a running pattern generator, which controls gait and speed.

The KAIST team is currently working to optimize their robot. They're interested not only in achieving ever-greater speeds, but also improving control and stability.

We've seen other legged robots take their first steps on treadmills and then learn to walk outside the lab. One example is HyQ, a quadruped from Italy. And then, of course, there's Boston Dynamics' WildCat (the successor to Cheetah), which was seen in a video bounding and galloping in a parking lot.

We hope the Korean researchers can build a version of Raptor that doesn't need a treadmill and that they can set free outdoors.

And then we can have a robot race.

[ KAIST Mechatronics, Systems, and Control Lab ]

The Conversation (0)

How the U.S. Army Is Turning Robots Into Team Players

Engineers battle the limits of deep learning for battlefield bots

11 min read
Robot with threads near a fallen branch

RoMan, the Army Research Laboratory's robotic manipulator, considers the best way to grasp and move a tree branch at the Adelphi Laboratory Center, in Maryland.

Evan Ackerman

“I should probably not be standing this close," I think to myself, as the robot slowly approaches a large tree branch on the floor in front of me. It's not the size of the branch that makes me nervous—it's that the robot is operating autonomously, and that while I know what it's supposed to do, I'm not entirely sure what it will do. If everything works the way the roboticists at the U.S. Army Research Laboratory (ARL) in Adelphi, Md., expect, the robot will identify the branch, grasp it, and drag it out of the way. These folks know what they're doing, but I've spent enough time around robots that I take a small step backwards anyway.

This article is part of our special report on AI, “The Great AI Reckoning.”

The robot, named RoMan, for Robotic Manipulator, is about the size of a large lawn mower, with a tracked base that helps it handle most kinds of terrain. At the front, it has a squat torso equipped with cameras and depth sensors, as well as a pair of arms that were harvested from a prototype disaster-response robot originally developed at NASA's Jet Propulsion Laboratory for a DARPA robotics competition. RoMan's job today is roadway clearing, a multistep task that ARL wants the robot to complete as autonomously as possible. Instead of instructing the robot to grasp specific objects in specific ways and move them to specific places, the operators tell RoMan to "go clear a path." It's then up to the robot to make all the decisions necessary to achieve that objective.

Keep Reading ↓ Show less