This is not a NASA Website. You might learn something. It's YOUR space agency. Get involved. Take it back. Make it work - for YOU.
Exploration

NASA's Multimillion Dollar Dancing Droid

By Keith Cowing
NASA Watch
December 17, 2015
Filed under , , ,

Keith’s note: NASA’s R5 robot can’t complete any of the tasks it was designed to do and placed last in DARPA’s challenge in 2013. So NASA sent the robot to two universities in 2015 to see if some students could fix it. NASA refuses to tell you how much this robot cost, why it was developed, or how it works. But before they sent their broken droids off to college for repairs they decided that R5 was good for one thing: dancing in music videos – just in time for Star Wars.
Is this what NASA calls “dancing”? This looks more like a slow motion mime with stiff joints. The video’s caption says “NASA’s latest robotic addition had been created to “perform in extreme environments.” The space agency is investing in robotics for deep space exploration and Valkyrie will compete in their Space Robotics Challenge in 2016.”
Oh boy – a robot dance off. I can’t wait.

NASA JSC’s Failed R-5 Robot Project Refuses To Explain Itself, earlier post
Is JSC’s R5 Droid Worth Fixing?, earlier post
Never Ask NASA a Simple Question, earlier post
NASA Awards Two Robots to University Groups for R&D Upgrades, earlier post
Does NASA Have a Robotics Strategy? Did It Ever Have One?, earlier post
NASA JSC Has Developed A Girl Robot in Secret (Revised With NASA Responses), earlier post

NASA Watch founder, Explorers Club Fellow, ex-NASA, Away Teams, Journalist, Space & Astrobiology, Lapsed climber.

9 responses to “NASA's Multimillion Dollar Dancing Droid”

  1. Skinny_Lu says:
    0
    0

    I like her. It may not be like the (computer generated) Transformer robots but she seems to be anatomically correct… not what you may think. I am simply referring to the hips, spinal column joints, legs and arms. Her skeleton mimics the human body. Yes, I understand her limitations and remember that she was beaten in competition by the rest of the robots in that field. Maybe is the yogi inside me, but I really like her 1-leg balancing poses. Maybe she can come to yoga class and get a bit more flexible! =)

  2. numbers_guy101 says:
    0
    0

    Ahh robotics. Another technology filled with so much unfulfilled optimism from the 90’s. I knew things were going south when the demand for ever faster, cheaper computing started to slow, about 2000-ish, as I started to hear the phrase “good enough for most users”.

    I was waiting for my actively cooled brain atop my desk, to which attaching a robotic body would be a natural next step. I’d talk to that brain / computer thing by now and those massive number sets would crunch and graph and distill, guided by a mix of AI capability and human questioning with leaps of intuition. Still waiting.

    But hey, we have dancing robots. And twitter. And kitten videos on demand. And tablets that let us read the newspaper without paper. And see the kitten videos.

    Ok…I’m ranting. It’s just the future is not what it used to be.

  3. Daniel Woodard says:
    0
    0

    Robotics has a lot of potential, but the human anatomical form is something of an accident of evolution, and is based on both our evolutionary past and the need to survive in a pre-technological world. It is counterproductive to artificially constrain robots to look like humans, and the rationale that robots would then be able to use human tools is not, in my opinion, supported by experience. Robot-tool interfaces can and should be different and simpler than human-tool interfaces. Form follows function, and both form and function in robotics are changing a lot faster than human anatomy.

    However I think it is actually productive to allow individual researchers within NASA the resources to explore their ideas through modest but consistent discretionary organizational funding, as was apparenbtly used in this case, rather than the daunting overhead, limited chance of success, and questionable precision of peer review. What is frustrating is that only a small minority of NASA’s capable and dedicated researchers actually get the opportunity to use their imaginations and implement their ideas.

    • Hug Doug ✓ᵛᵉʳᶦᶠᶦᵉᵈ says:
      0
      0

      Robots need to interact with a human-shaped world, and may have need to use human tools, but they don’t need to be exactly shaped like a human to do either of those things 🙂

      That is something that I think the DARPA robotics challenge showed quite powerfully. The humanoid robots looked so clunky, and it was the ones that had not-human articulation that performed the best.

      • Daniel Woodard says:
        0
        0

        Yes, it would be rather bizarre to design a robot to turn a manual screwdriver with a human-like hand using dozens of motors and actuators, which has to let go of the screwdriver and grip it again multiple times to turn the screw, when one motor and a socket for the driver tip is all that is needed.

        Of course we haven’t really explored optimal robotics for weightlessness, where wheels and treads aren’t needed. They might look more like undersea robots. Maneuvering fans (inside) or jets (outside), grippers to stabilize the robot while working, and several articulated arms with manipulators or tool sockets?

        • Hug Doug ✓ᵛᵉʳᶦᶠᶦᵉᵈ says:
          0
          0

          They do have Robonaut on the ISS.

          http://robonaut.jsc.nasa.go

          Right now it is mounted on a post,

          http://www.nasa.gov/sites/d

          but they have developed legs for it to grasp onto the handles that are everywhere on the station.

          https://apollomapping.com/w

          It isn’t very far along, but it is a start. Slowly, we are inching our way into a future we can barely imagine right now.

          • Paul451 says:
            0
            0

            It’s funny that the “legs” are better designed limbs for a free-fall robot than the entire idiotic faux “EVA suit” upper torso/arms/head.

            [Four of those “legs” symmetrically around a central body with the main cameras and processing systems. Smaller satellite cams on the “wrists”. Replaceable/adjustable “hands”. Bam! Done. General purpose ISS climber bot.
            Of course, I’ve just described a slight variant on JPL’s RoboSimian; just moving the “head” to the centre of the body.]

            [Edit: Looks like the legs have a camera on each gripper, not back on the wrist. If it had replaceable tools/hands, instead of fixed grippers, it might be better to put the camera on the limb itself. OTOH, cameras aren’t expensive, so it might be worth having them on the tools. That way the position is optimised for each task, and if a camera is damaged, it only affects that tool.]

          • Daniel Woodard says:
            0
            0

            Clearly the legs are not articulated like human legs. Is the robot able to translate from one location to another using them?

          • Paul451 says:
            0
            0

            Is the robot able to translate from one location to another using them?

            Errr, that’s their only job.

            Video of the legs in action:

            https://www.youtube.com/wat

            You can see it (slowly) using ISS-style grab-rails from 1:42.