[Jongwon Park] and his team of students at the Korea Advanced Institute of Science and Technology have created a fast biped robot based upon the Velociraptor. Raptor weighs in at just 3Kg, and stands only 470mm tall, yet it is capable of running at 46 km/h. That’s almost as fast as Boston Dynamic’s Cheetah.
Raptor uses carbon fiber composite legs to absorb and release energy while running. The system is similar to that used in high performance prosthetic legs. A rotating tail assembly further helps to balance Raptor on rough terrain. We have to admit, the tail system does look a bit dangerous for any humans who might need to interact with the robot. It does work though, as evidenced by Raptor bounding over Styrofoam blocks.
The Raptor robot is quite impressive when running at full speed. Considering this project’s budget was nowhere near the resources of Boston Dynamics, it’s an amazing accomplishment. The video reminds us of Boston Dynamics founder [Marc Raibert’s] early robots at the MIT Leg Lab. We can’t wait to see what this team produces in the future.
[Via Engadget]
Can’t wait to see it in my nightmare.
This is some beautifully elegant work. I wonder what the power usage is.
Am I the only person who can’t get excited about a tethered robot?
Sigh, yes ;-) As cool as it is, the first thing that screams out is “where do you put the power source”! If I remember, the Big Dog machine required a 15HP petrol engine to provide power – ok, much bigger machine, but you get the idea.
High voltage graphene super capacitor with carbon artificial muscle that takes 5kV to power it. It’s where I see potential.
5kV is definitely a lot of potential!
no you arent
this is as much running as that other kickstarter wheel robot
notice how “legs” are hard wired to only move in predetermined pattern, its another low polygon count WHEEL
seriously?…when was the last time you saw a wheeled vehicle where only one wheel touched the ground at any given point?
Tether or not, the kinematics necessary to make this work is a feat in and of itself, its more than just two sticks swinging back and forth….
but I guess for those of you visiting from the future, this old hat…
Not only that, but the tether appears to be rigid making it impossible for it to fall over sideways (one of the most difficult parts of bi-pedal propulsion) or turn.
Check out the actuators on the lower legs. It looks to me like the basic running motion is baked in using cranks, but the lower legs do per-step adjustment. It’s certainly closer to bipedal running than that silly R/C wheg thing.
there is no actuator on the lower legs those are shocks to deal with impact forces
They tether them in order to safely develop the prototype (safe for both the developers and the robot itself). it’s hard enough to come up with a robot that can walk a mile an hour, that much harder to come up with one that can out run you. would you really want to invest hundreds, if not thousands of hours in development, and several thousand dollars (or more) in material, just to have it go flying off because you had a single tiny thing wrong, and destroy itself (and probably other things) in the process?
Develop the mechanics in a controlled environment, then figure out how to power it on board (lipos aren’t that heavy and are quite powerful, just might not have a long run time). A good number of times something is developed that by itself isn’t terribly useful, because it needs other developments to make it useful. Once we can develop a viable high energy, long term, light, and safe power course, these developments (the cheeta, the velociraptor, flying cars, etc) will become much more viable and can then be considered really useful. There are lots of discoveries and inventions out there that aren’t very useful right now because by themselves, they just aren’t really viable. But someday, they will be when the rest of the puzzle pieces are created.
Batteries are a significant % of the weight so I guess they are or will be using dummies, but I agree with everything you said, basically.
“Dynamic stability control” aka “whirling blades of death that both mutilate you and help keep your robot overlord upright.”
A truly elegant solution that solves two pesky and persistent problems at once, dynamic stability control, and humans. I like how you look at things
I’m not the only one (who thought that), then. At first I thought it can’t be a raptor because where’s the claw? Then they showed the whirling blades and everything became clear. (c:
having raised two human toddlers, let me say that I am less impressed by “running” than “walking” and most impressed by a device that can go from standstill, to walk, to run, then safe transition back.
the difference is like that between VTOL in a tiltrotor and a helicopter. One is easy to do in toy grade RC. The other, not many RPV’s actally translating into a forward flight regime and back.
Just like the 14 month old that can lever himself up using the couch and in a rapid, unsteerable, barely controlled fall head towards his parent, or a wall, or coffee table. Fast but almost uncontrollable and unable to hold position without aid. Good thing kids are rather durable, but when it’s robots that can barely steer and have a hard time stopping, not so good.
I’d like to see a robot that can do what I call ditch running. When I was a kid, we lived on a place with a small irrigation ditch along the top of a large, grass covered hill.
I used to run down that ditch, placing my right foot on the left side and my left foot on the right side, criss-crossing my legs as I ran.
If I tried that now I’d probably trip myself and break my nose. ;)
The back stoop was concrete and 4 steps high. I used to run in a crouch on the sidewalk beside the house then jump up over the side of the stoop onto the top step, still in a crouch.
If parkour had been a thing in the 70’s…
Another crazy thing I did then was run around the property (a bit over 2 acres) in the dark, without tripping on the rough spots. How I did it was I kept my ankles loose and ran flat footed. I’d feel for part of my sole to touch the ground then push the foot down – letting it tilt into full contact, then lock the foot’s roll. That would happen as I rotated forward over the ankle.
Robots could work the same way. Figure out the motion from the feet up. Just swing the leg forward and put the foot down, don’t bother with calculating how far, the foot senses when to go from standard motion to the code that aligns the foot to the ground, locks the ankle roll and powers the ankle tilt forward.
I also ran on ice in a similar way, feet down *flat* onto the surface for maximum traction then *pull back* instead of trying to roll forward and push off with the toes. Had to get center of gravity over the forefoot before lifting the hind foot. Sort of like front wheel drive running. If a foot providing forward motion got less than full contact before power was transferred to the other foot, slip and an attempt to fall.
I say attempt to fall because I did and still do have a very good balance sense. Several times not falling on ice has been at the expense of a sprained back. I’d much rather have a sore bum than a sprained back muscle but I couldn’t interrupt the balance reflex.
Yes, people did remark that I had weird ways of running. I was just trying to figure out better methods.
Awesome, but needs more neck frills.