Humanoid robot from GM and NASA

X10Co At004
This morning, NASA and General Motors unveiled Robonaut 2, aka R2, which is a weird name considering it looks nothing like the real R2. Robonaut 2 was designed to assist both astronauts and auto workers. It has no legs. More details and video after the jump.

From Popular Mechanics:

"We had a common agenda with NASA," says Allen Taub, vice president of global research and development at GM. "They wanted to make a robot that could work next to an astronaut," he says. "The question we wanted to answer was, 'How do I make a robot so it can work with operators, without all of the safety precautions and cages?'" As they go through their automated routines, industrial assembly bots are inherently dangerous to be around. And according to Taub, installing cages and other safety measures often costs more than the robot itself. "This robot can be going through its paces, and if you just hold your hand up, it hits your hand and stops," he says...

GM's goal in co-developing R2 is to eventually install similar systems in its plants, performing the kind of repetitive, ergonomically difficult jobs that might injure a human operator. Vision sensors in the robot's head, as well as pressure sensors in its fingers, allow it to manipulate parts with near-human precision. The biggest upgrades from the original Robonaut are R2's thumb, which now have four degrees of freedom (as opposed to three), and its overall speed, which have improved by a factor of four. One result of all of this engineering is the kind of breakthrough only a roboticist would swoon over: R2 can use both hands to work with a piece of flexible material. If that sounds simple, consider the amount of sensory data, cognitive processing and physical dexterity needed to manipulate something that flows and bends in your fingers.

"NASA and GM Create Cutting Edge Robotic Technology" (PopSci)

"NASA and GM Create Cutting Edge Robotic Technology" (GM)


  1. ObSimpRef:


    I’d really much rather work alongside a robot that looks nothing like a human. A spider, say.

    This GMBot looks like it is one gamma-ray-fried IC chip away from declaring “MUST DESTROY CARBON UNITS!”

  2. What I don’t understand is why makers of robots still go at making their creations look like humans. I guess, it is because the human form is a perfect image of a functional machine. I would think that function precedes form, though.

    I can’t wait for how people what people will think of as “other applications” for this machine.

    1. What I don’t understand is why makers of robots still go at making their creations look like humans.

      Most robots aren’t- have you seen pictures of a modern automotive plant or the Martian rovers? But unlike autonomous (or semi-autonomous) robots this one is designed to be remotely controlled by a human using a VR-style interface, so an intuitive human-like form makes sense. Think “Avatar” or “Surrogates.”

      That said, the body language in photo #1 clearly says “Come with me if you want to live.”

      1. But unlike autonomous (or semi-autonomous) robots this one is designed to be remotely controlled by a human using a VR-style interface, so an intuitive human-like form makes sense. Think “Avatar” or “Surrogates.”

        Hmmm, did I miss that somewhere in the write-up? I didn’t see anything that implied that in the video.

        There’s probably nothing to say that they couldn’t retro-fit this robot with such an input at a later date, once they have that working (which I don’t see any evidence that they do), but it seems clear that they’re planning on this to be useful before such a time ever comes.

        @Auto Parts for Brains: The video mentions that this style of robots gives a dexterity and workspace similar to that of a human. I think they key is being able to use tools designed for humans, as someone mentioned above.

        Note, though, that a robot that has more degrees of freedom than a human would still be able to use human tools. I’ll bet the only thing preventing this from having double-jointed elbows is a possible lessening of strength because it would have to put that motor in the middle of the join rather than the edge. (Note that double-jointed robots would make Brainspore’s VR-controlled robots hard to manipulate fully by a human.)

        1. I remember reading about the VR-style interface a couple years ago when I first heard about this project. If they’ve abandoned that idea for a robot that is just controlled by joysticks or something then I agree it doesn’t make nearly as much sense to give it humanoid form.

        2. In the video, in a lot of the presentations, you can see a controller off to the right on the computer.

          I don’t know much about this second one yet, but the first generation was featured in an episode of Scientific Frontiers, you can watch it online and jump right to the third segment. It’s been a while since I’ve seen it, but if I remember correctly, they go into the intent behind the robot, and its capabilities. At least, what the intent and capabilities were in the first generation!

  3. Considering the gloves and bulking gear astronauts wear to survive the vacuum of space, this robot could help immensely with repair and building in space. You could even have it respond to human control using an interface similar to what surgeons use, allowing someone on the space station or even back here on earth to remotely go and fix/repair/construct in space.

  4. Interesting how GM couches the robot as something that will improve the safety of the products. The more human this technology looks, the more I think about workers in days-of-yore worrying about humanistic cyborgs who steal jobs from humans, and how this would be a kick in the tits for them.

    The flipside is that complex systems — as these mechanical turks definitely are — require their own vast system of human builders, operators, and maintainers. My guess is that a robot builder on their own will bring in more income than will the assembly-line worker they replace, but what would the exact balance be? Would the labor switch-offs be offset with price cuts, or is the capital expended to lower labor and time costs too great? Essentially, who will benefit?

    Not a new argument, and limited in scope to only those producers who have the capital to spend on such a program, but given the increasingly low costs of such “high technology” apps as computer-run fabricating machines (laser cutters?), there might come a point in the idealist’s future where every workshop might have a mikita table-top assembler.

  5. I’m skeptical. It’s always what they don’t show you. Did it pick up the weight on its own? Can it pick up a tool? How about a screw? Can it write in a font smaller than 1000 pt? Where’s the breakthrough? We’ve seen this done elsewhere, better, faster.

    And yeah, it looks terrifying.

  6. If you are going to be using a robot for space work, which is high precision industrial and scientific work, why the HELL would you make it anthroform? The human body is efficient for purposes of walking, gathering, and wading in water. It is not efficient for zero gravity construction and maintenance purposes.

    It is the stupidest thing in the world to design androids and other anthroforms for this kind of usage. The only potential benefit of making tools resemble humans is in scenarios where the humanoid form is beneficial for psychological reasons. So a hospital having androids to do menial tasks makes some amount of sense, but for purposes where the robot in question serves the sole purpose of getting practical and tangible work done, it should be purpose built for its operational needs, not to appeal to human instincts and psychology.

    ~D. Walker

  7. “What I don’t understand is why makers of robots still go at making their creations look like humans.”

    Most tools and machines are already designed to accommodate humans. A human-shaped robot capable of behaving like a human seems to be the ideal.

  8. I would agree with CYMK. Version 3 of this robot could very well replace humans in outer space exploration. No need for bulky life support systems, crew cabin, etc… thereby cutting the weight of a spacecraft tremendously. I understand the thrill of human space exploration, but the cost of doing so is far, far in excess of building humanoid robots to do the same.

    This type of scenario is strangely similar to a recent movie… can’t remember the title. :-)

    1. I don’t think humans would ever take a back seat to robots in space exploration. Robots would definitely be used in a more expansive role IMO; you could shoot a rocket full of supplies and semi-autonomous robots to the moon or mars. Once there, they set upon building structures and things necessary for human survival; when we actually send out humans to explore, they will have a safe environment waiting for them (ie: moon/mars base). You could do the same with LaGrange points; sends robots and parts to build a space station.

  9. 1. I’m keeping a close watch for the word “Skynet”
    2. I nominate myself as the leader of the resistance. DEFEAT THE MACHINES!

  10. I think you meant “considering” instead of “consisting” in the first sentence of this article. I also think the robot’s face looks very cool. Thanks!

  11. Sure, a human shaped robot in space is less efficient than one designed strictly for the job. That’s missing the point. It’s really about projecting the human image into space, and reassuring us little people on the ground that it’s *people* up there, and not corporate drones.

    epic fail.

  12. I can think of other reasons to make a humanoid robot, too, especially if said robot is working next to you.

    Consider what they said above about holding up your hand to make the robot stop. This is a gesture humans already intuitively know how to use with each other. When we do this, we usually gesture towards another person’s head. If the robots looked like it was assembled from tinker toys we might not know where to gesture.

    There’s a wealth of emergency situation I can visualize where you might need a robot’s assistance. Counting on the robot to be shaped like and move like a human would mean emergency planning would be faster and more intuitive; it could get into a car and drive it, grab any tool you could use and utilize it, fit into any opening you could fit into, and so on.

    1. I believe the implication in the article was that if the robot came into accidental contact with a human/etc. it would automagically sense the contact and pause it’s movement. (Instead of the current gen assembler bots that would simply squish us meatbags, accidentally…)

  13. With that first pose, I see a Duffman-terminator reciting Hamlet. Our overlords will be strange, indeed.

  14. This may lead to a strange role reversal:

    “Open the pod bay doors, Dave.”
    “I’m afraid I can not do that, Hal.”

  15. It’s seen things you people wouldn’t believe. Attack ships on fire off the shoulder of Orion. It watched c-beams glitter in the dark near the Tannhäuser Gate. All those … moments will be lost in time, like tears…in rain.

  16. Yeah, well, in Japan the robots can walk. I guess this one can get a wheelchair.

    Hell, get him some articles of incorporation, a handicap plaque, and the rights to use the carpool lane. I mean corporations now have human rights, yes?

  17. Of course it can’t walk. It’s a GM product. Robots walking everywhere instead of driving would set a bad example.

  18. Answers to a couple of questions above (sorry that I have to be vague, but I work for a subcontractor on the project and NDAs abound.)

    Yes, the ‘bot can lift the weight on its own. Strength is not a problem for robots, nor is delicate movements. The trick is getting robots that can both be strong AND delicate at the same time, and can also sense (and quickly stop) when they are whacking a human or something else they didn’t intend to wail on.

    Why a humanoid robot? The human arm is a pretty good design, especially if you’re using human tools. AFAIK, the head is just for public relations — you could easily put your robot “eyes and ears” (cameras / mics) in the torso, but when you need to get funding from Congress, you need a certain cute factor. Why no legs? The bot doesn’t need it at this spot. Legs aren’t helpful for EVAs. My employer isn’t involved in the propulsion end, but I assume you could attach the ‘bot equivalent of a tether and jet pack like the astronauts use.

    Is there a huge breakthrough here? Not really. The current trend in industrial robots is for robots that can quickly be repurposed for a variety of tasks, that can use human tools (so you don’t need duplicate tools for humans working along side the robots), and for robots with good force sensing so they can avoid whacking humans when humans do stupid things like get in front of a robot that’s hammering on something. Robotnaut does need some extra care to keep stray radiation from nailing the electronics, and needs to work in colder temps than is typical on earth, but otherwise is pretty average for current robotic technology.

Comments are closed.