Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
×
Robotics The Military Hardware

11-lb Robot Can Jump 30 Feet Into the Air 68

Ruvim writes "Boston Dynamics has developed a 'Sand Flea' 11-lb robot that drives like an RC car, but when it needs to, it can jump 30 feet into the air. An onboard stabilization system keeps it oriented during flight to improve the view from the video uplink and to control landings."

This discussion has been archived. No new comments can be posted.

11-lb Robot Can Jump 30 Feet Into the Air

Comments Filter:
  • by Genda ( 560240 ) <mariet@nOspaM.got.net> on Wednesday March 28, 2012 @04:56PM (#39501639) Journal

    Amir... I think we have fleas!!!

  • One day... (Score:5, Funny)

    by SexyHamster ( 174881 ) on Wednesday March 28, 2012 @04:57PM (#39501649)

    One day Boston Dynamics is going to combine all these robots together and kill us all.

  • It's cool (Score:4, Interesting)

    by roman_mir ( 125474 ) on Wednesday March 28, 2012 @05:00PM (#39501699) Homepage Journal

    It's a nice RC toy, but how is it a robot? Just having a receiver, an electric motor to drive and some spring mechanism and a person controlling it over radio makes a robot?

    • From the article: "An onboard stabilization system keeps it oriented during flight to improve the view from the video uplink and to control landings."
      • It's still just a remote controlled vehicle. RC helicopters have onboard gyros to keep them pointing in one direction, doesn't make them robots.
        • by Tsingi ( 870990 )
          What is your definition of a robot then?
          • Autonomous.

          • The simplest definition:
            A machine capable of carrying out a complex series of actions automatically.

            A better definition:
            any machine or mechanical device that operates automatically with humanlike skill.
          • I know i'm being nitpicky, but I guess i'd like the see the word reserved for machines that exhibit some higher level of autonomous operation. In many people's minds, a robot should display goal setting beyond, "maintain attitude". It should recognize threats and impediments to it's task and make it's own decisions to get around them. I feel like a self driving car is more of a robot than the stuff coming out of Boston Dynamics.

            BigDog's ability to maintain balance when kicked is impressive, but I have a h
            • by emj ( 15659 )

              Industrial Robot [wikipedia.org] is what I think of when I think of robot, and this isn't that far from that, and it is autonomous for some parts of it's journey. But my reasoning might be slightly handicapped by the fact that in Swedish robot also means missile. :-/

          • What is your definition of a robot then?

            Something that can't be bargained with, or reasoned with. Something that doesn't feel pity, or remorse, or fear. And something that absolutely will not stop, ever, until you are dead.

      • And from the video, even that doesn't look particularly true. I wouldn't call it "keeping it oriented" if it involves a nearly 90-degree spin before it lands again.

    • back home, and they arent much bigger than a receiver, an electric motor, and some spring mechanism.

    • Right now it is, they are only at platform stage. It's not all that difficult to install gps capable path-finding code on that little thing. Once that is complete, all you really need is a swarm of those with c4 payload to deal with tanks, artilleries, personae... you name it. Just para in like 50 of those with pre-programmed coordinates, and things go "kaboom".

    • no one claimed it was autonomous or intelligent or whatever. just a demonstrator for a mode of locomotion: a small wheeled vehicle that jumps really high and doesn't break in the process. Autonomously navigating that thing would pose one major problem: how is it supposed to know where it is going when jumping onto higher terrain? Like, the roof in the video, how could it see there was in fact a roof and not a bottomless pit? it could possibly team up with an airborne unit relaying its extended field of view
      • > how could it see there was in fact a roof and not a bottomless pit?
        I'd assume the drone flying overhead a few miles away would tell it.
  • by Hognoxious ( 631665 ) on Wednesday March 28, 2012 @05:03PM (#39501747) Homepage Journal

    tries to work out GPE ... too confused to do it with units based on an 11th century King's knackerbag.

    tries to convert units to metric ... too drunk.

    conclusion: I for one welcome our robotic howitzer overlords.

    • by Yvan256 ( 722131 ) on Wednesday March 28, 2012 @05:15PM (#39501923) Homepage Journal

      4.98951607 kilograms robot can jump 9.14400 meters into the air.

      • 4.98951607 kilograms robot can jump 9.14400 meters into the air.

        I have two words for you: significant figures. Everything else is just noise...metric noise.

        • Re:frumpy poise (Score:5, Insightful)

          by Bromskloss ( 750445 ) <auxiliary.addres ... nOspAm.gmail.com> on Wednesday March 28, 2012 @07:52PM (#39503665)

          I have two words for you: significant figures. Everything else is just noise...metric noise.

          Aghnnn! I'm not at all at ease with having the number of digits written out signify the uncertainty. That's just not very elegant. By that method, you can only express certain ranges (for example "1.45 to 1.55", but neither "1.44 to 1.54" nor "-5 to 17") and it's even dependent on what base you write in! Rather, give uncertainty as a separate number.

          Even more sophisticated would be to specify a probability distribution over possible values, but in the above discussion I assume that one wished to express the uncertainty as a range with sharp boundaries.

          In case anyone wonders, I do myself practise the "significant digits" method when the social context calls for it and I want to please people (so that they give me money, for example), but whenever I can, I follow my heart and do what's right.

  • minor issues (Score:4, Insightful)

    by v1 ( 525388 ) on Wednesday March 28, 2012 @05:07PM (#39501795) Homepage Journal

    You can see the camera do take cuts before it jumps most times. Either it's chargin' its laser and they cut for time, or it's got other issues. (it may take awhile to say, charge a compressed spring piston)

    Another possible angle is it may only be able to set up for the jump if it's right-side-up, and we never see it flipping itself over, so if it found itself upside-down, it could be in trouble?

    But those jumps are quite impressive. I got the impression though that it could only do one height of jump, like when you see it jump off the top of the building. I was expecting a short hop to get it off the edge, but it cranks way back and catapults way up and off. So it seems to have a ways to go, development-wise.

    • I got the impression though that it could only do one height of jump, like when you see it jump off the top of the building. I was expecting a short hop to get it off the edge, but it cranks way back and catapults way up and off.

      Huh? The jump off (@ 36 s) isn't even 1/4 the height of the jump up.

    • by tool462 ( 677306 )

      In addition to what others have pointed out, the part of the article that had the words in it mentioned that it is accurate enough to jump through a window two stories up.

      • by atisss ( 1661313 )

        Except, it can't jump back out through the window.. Who needs a robot that can't even do self-defenestration?

        • by Jeremi ( 14640 )

          Except, it can't jump back out through the window.. Who needs a robot that can't even do self-defenestration?

          Add some explosives to that robot and you have Philip K Dick's idea of a hand grenade....

    • I made the assumption that it jumps so high when leaving the rooftop to show how durable it is - that was the maximum falling distance for the playground they'd set up.
    • Re:minor issues (Score:5, Insightful)

      by Tsingi ( 870990 ) <graham.rick@gma[ ]com ['il.' in gap]> on Wednesday March 28, 2012 @07:31PM (#39503449)
      Just to reiterate, I actually watched when it jumped from the top of the building to see if the height would be the same. The jump at the top of the building arced ~5 feet. Not near as high as the jump to get to the top.
    • Another possible angle is it may only be able to set up for the jump if it's right-side-up, and we never see it flipping itself over, so if it found itself upside-down, it could be in trouble?

      As with many jumping devices there is no top or bottom; the device is symmetrical top to bottom. If you look closely at the video you will see that the legs are attached between the wheels and chassis and are themselves symmetrical. They can move in either direction so it does not matter which side is up for the device to be able to jump.

      Sire it would require camera lenses on both sides of the device but that don't weigh much..

    • You can see the camera do take cuts before it jumps most times. Either it's chargin' its laser and they cut for time, or it's got other issues. (it may take awhile to say, charge a compressed spring piston)

      Or they are changing from a closeup that shows the elevation and charging to a wide shot to show the jump.

  • ...and where can i get one?
  • Comment removed based on user account deletion
  • The Mach-5 was the first thing I thought about when I read the title, but this is definitely something else.

    The second thing I thought: "Do not give this to my children. I'll be climbing on the roof every day."

    • The second thing I thought: "Do not give this to my children. I'll be climbing on the roof every day"

      I don't know- looks like what they've invented here is a toy that can get itself down from the roof. And land on children's heads like the reincarnation of lawn darts, but that's neither here nor there....

  • If the future of /. is flash videos, then I'm going to have to leave.

    Not gonna watch video, not going to load flash. I prefer articles and discussions.

    Thanx in advance.

  • They cut the video pretty much before any single jump. Plus, they don't have a good shot of the last flight. The jumps are cool but any of us could do a jumping machine that "just jumps". The hard thing is to make the whole package.

    Come on, don't tell me they didn't had time for it. They are a big company and should watch what they release.
    This reminds me of this: http://www.youtube.com/watch?v=VXJZVZFRFJc

    Maybe Boston Dynamics is just taking a piss out of DARPA.
  • That thing is Awesome!

  • Legged locomotion is incredibly difficult to engineer with weakly defined benefits. Use some wheels, and when the obstacle is to big, just jump. I really like that, that might be the one way to do it. That is: for small drones, operating outdoors. However, it may not translate so well to indoors environments, or to larger robots, or load-carrying ones.
  • It looks really cool, but there is a cut between every time the device is posed to jump and it jumps. A camera change at 0:09. A really weird one at 0:18 where the camera doesn't move but light changes so you know there was a cut. The jump at 0:35 doesn't look like it has a cut but it is a smaller jump too. The next jump has one at 0:51. So 4 jumps with three cuts between the preparing to jump and the jump. Not saying it is a fake, but it has many attributes that a faked video would have.

  • Afghanistan walls have been increase to 31 feet.
  • The spindly protrusions off of the wheels, are those somehow helping to stabilize the flight trajectory after launching? How come they don't cover the plastic wheels with some kind of rubber tread to increase traction and decrease the scratching/scrambling noise the plastic makes. I would imagine a bit of stealth would be necessary for the environments this will be used in. Or would that not actually help in the dirt/sand?

    That said, launching off of a roof and landing on concrete and still working perfectly

You know you've landed gear-up when it takes full power to taxi.

Working...