Slashdot Mirror


Robot Balances on a Single Spherical Wheel

dalangalma writes "Researchers at Carnegie Mellon University's Robotics Institute have developed a new kind of robot called the Ballbot that balances on a steel ball. Using a mechanism similar to a ball mouse, the Ballbot uses rollers to drive its single, spherical wheel and balance in place or glide around the room. The promise of such dynamically stable robots is that they can be much taller without having to have a wide base, making them much more suitable for working with humans. They are also much more agile, since they can be pushed out of the way easily without falling over. You can read the press release or check out the project's web page when it recovers from traffic."

211 comments

  1. translation by McDrewbie · · Score: 5, Insightful

    "single spherical wheel" = a ball

    1. Re:translation by jboker · · Score: 4, Funny

      thanks, i was confused

    2. Re:translation by Anonymous Coward · · Score: 0

      Maybe the meant that the robot has other kinds of balls.

    3. Re:translation by quakeroatz · · Score: 1

      Even with the AC, misspelling, that is quite funny.

    4. Re:translation by camg188 · · Score: 1

      as opposed to a single rectangular wheel.

  2. Step one completed by rudeboy1 · · Score: 1

    I want my Tachikoma!

    --
    Raging in an online forum won't do anything for the world around you. To see change, you must take action.
    1. Re:Step one completed by LionKimbro · · Score: 4, Interesting

      Tachiko-maaaa!

      My daughter's obsessed w/ Tachikoma, as well. She's buying blue everything, to look like one. Her biggest dilemma right now in life, is whether to die her hair purple, (and thus resemble the Major,) or to die her hair blue (and thus resemble a Tachikoma.)

    2. Re:Step one completed by heffeque · · Score: 1

      Me want a Tachikoma too! ^^

    3. Re:Step one completed by gbjbaanb · · Score: 1

      I say purple. When she gets older, she'll realise there's absolutely nothing wrong with looking like Motoko.

    4. Re:Step one completed by daft_one · · Score: 1

      ZOMG! Your daughter wants to die? Get her some therapy now, I... Oh wait, context tells me you just need to learn how to spell the word 'dye'.

  3. meet the new dalek by mgabrys_sf · · Score: 5, Funny

    re:"The promise of such dynamically stable robots is that they can be much taller without having to have a wide base, making them much more suitable for working with humans. "

    Until said humans goes down a flight of stairs.

    *Oof*
    *Bounce*
    *Expensive crashey noises*

    At this rate we'll have-em screaming "exterminate" in no time.

    (and yes I know post 2005 Daleks can do everything and go everywhere - I'm talking pre-video-game dalek)

    1. Re:meet the new dalek by Anonymous Coward · · Score: 0

      Yea but will this said robot know not to piss in the wind?

      ~kalinga

    2. Re:meet the new dalek by Kennric · · Score: 4, Insightful

      With a suitable suspension I don't see why they would not do fine on stairs. If the balancing algorithm is good, it should recover fine on each step. A little laser rangefinding could even allow it to know it was going down stairs and adjust accordingly. Certainly a dynamicly balanced bot would fair better than a 4-wheeler going down the same stairs. If I can ride a motorcyle down the stairs, my new robotic biroid overlords should not have a problem.

    3. Re:meet the new dalek by miro+f · · Score: 3, Insightful

      yes but can you get them to go UP the stairs?

      --
      being vague is almost as cool as doing that other thing...
    4. Re:meet the new dalek by Mr.+Freeman · · Score: 3, Insightful

      Yeah, but try going UP the stairs. That's a bit more of a problem.

      --
      -1 disagree is not a modifier for a reason. -1 troll, flaimbait, redundant, overrated are NOT acceptable substitutes.
    5. Re:meet the new dalek by mgabrys_sf · · Score: 0, Troll

      this is a lame rejoinder on your sig - but even when in an abject pissy mood where I want my karma to drop like a rock, for the hell of it all - getting bonus points for when my sense of humor sneaks out is a great thwarter of such moods. IE - I like-em.

    6. Re:meet the new dalek by theparag0n · · Score: 1

      Actuelly, the imperial daleks in Remembrance of the Daleks could also do stairs.

    7. Re:meet the new dalek by LionKimbro · · Score: 4, Interesting

      I think this technology could be adapted to human like motion.

      Turn the ball on the ground into an ankle. That is, put a foot/pod beneath the ball, and the robot balances itself above the foot/pod.

      Then it should be stable enough to take a step with the other foot/pod.

      I'm thinking: "The reason you can't put two of these side-by-side, joined at the top, and make it walk up stairs, is because there's such a sudden change in it's weight distribution when it picks up one foot. You need something to be stable."

      I started asking myself, "How do people do it?", ... and realized that we have these big flat things underneath us: our feet. And that we might balance ourselves using our various joints.

      If you stack up 3 of these things on top of each other, and synchronize their intelligence, can they stand up on top of one another? That's what I want to know now.

    8. Re:meet the new dalek by Yvanhoe · · Score: 1

      Of course the point of the article is that, if we can balance a robot the size of a man a single ball, we can probably make an extremely stable two-leged robot with small feet (and not weighting half as much as the body of the robot) of the same size.

      --
      The Wise adapts himself to the world. The Fool adapts the world to himself. Therefore, all progress depends on the Fool.
    9. Re:meet the new dalek by mattyrobinson69 · · Score: 1

      R2D2 could fly, he(?) didn't seem to have the ability to go upstairs, until he turned into a flying ninja, for a few seconds.

    10. Re:meet the new dalek by Anonymous Coward · · Score: 0

      If you stack up 3 of these things on top of each other, and synchronize their intelligence, can they stand up on top of one another? That's what I want to know now.

      It would be a better challenge if you didn't synchronize them, then they'd all have to react to the force from the guy on top of them as well as the movement of the guy below them. Can these really get a true balance, or are the robots constantly shifting in an attempt to compensate for leaning one way or the other? If it's the latter, I can easily see the feedback making them have larger and larger movements until they are all madly flailing about before finally falling down.

      I wonder how far it can lean without rolling over, which I guess would be a function of how much torque the mechanism can handle before the rollers lose grip on the ball (does the robot need to know how much it weighs/is carrying so it can calculate the correct torque to apply to the ball or does it figure it out from its balance?). I think you're right about using this as an "ankle" too, though I wonder if you can just make the "ball" flat on the bottom and limit how far it can roll in a direction, or have a separate foot that also has a roller mechanism (that probably would have to be synchronized), allowing the "ankle" to roll freely between them.

    11. Re:meet the new dalek by budgenator · · Score: 1

      I just got this real crazey image of this monopod robot jumping up a flight of stairs like a one legged man. Maybe they should put an warning devices that sounds "boink, bionk bionk" so people would be aware of the possibility of robots crashing down the stairs.

      --
      Apocalypse Cancelled, Sorry, No Ticket Refunds
    12. Re:meet the new dalek by Paolone · · Score: 1

      It's about sharing the data to make the stata and intentions known, not to trick the other segments... :)

    13. Re:meet the new dalek by mad.frog · · Score: 1

      Or until their batteries run low.

      Hopefully they'll have a "power-low" sensor that will alert them to extend training-wheels or some such...

    14. Re:meet the new dalek by cmeans · · Score: 1

      My prediction for the future: Stairs will be deprecated in favor of ramps.

      Or maybe there'll be a ramp on each stair-case. Stairs at that inclination will be easier for humans, but a ramp would be easier for Ballbot and it's brethern.

      Of course, as someone has already invented the Elevator, maybe the problem is already solved.

    15. Re:meet the new dalek by Anonymous Coward · · Score: 0

      And don't forget the old joke:
        How does a Dalek get up stairs?

        Answer: It doesn't. A REAL Dalek just levels the building.

    16. Re:meet the new dalek by Anonymous Coward · · Score: 0

      Well, maybe. Does the hypothetical stair-thwarted robot have arms? And are they strong enough to support its own weight? Are they agile and long enough to get the base up one step?

    17. Re:meet the new dalek by Iron+Condor · · Score: 1

      This guy here: http://www.bostondynamics.com/content/sec.php?sect ion=BigDog can not just climb stairs but carry 120 pounds in the process. Click on the video for an eery demonstration of the uncanny valley effect.

      --
      We're all born with nothing.
      If you die in debt, you're ahead.
  4. Brilliant!! by BarryLoper · · Score: 1

    Just don't make it drive into the wind!

  5. Segway on a ball! by toybuilder · · Score: 1, Insightful

    Cool! I'd love to see if they can adapt this to a Segway-like transporter!

    1. Re:Segway on a ball! by Anonymous Coward · · Score: 0

      Hmm 2 wheels to 1 that means my hoverboard is next woot!

    2. Re:Segway on a ball! by Jeremi · · Score: 1

      I'd like to see the human-powered, human-carrying version. Think unicycle, but with a spherical wheel. You know, for when traditional unicycling isn't enough of a challenge anymore...

      --


      I don't care if it's 90,000 hectares. That lake was not my doing.
    3. Re:Segway on a ball! by discHead · · Score: 1

      Actually, if memory serves, the Segway patents already cover a number of possible wheel/ball configurations, including a single ball. I can see the IP lawyers flaring nostrils and raising hackles already.

  6. i'm surprised this hasnt come around sooner... by surfbass · · Score: 1

    im surprised this hasnt come around sooner, i mean segway demonstrated it with their scooter...sort of

    1. Re:i'm surprised this hasnt come around sooner... by Mark+Programmer · · Score: 5, Informative

      Similar, but different.

      The overall concept is the same (in the sense that you have a system to balance and probably similar mathematics to do it), but the single point of contact with the ground makes for a different solution matrix. Segways need only balance in one dimension; this robot has to account for two, and it must solve for its balance with one manipulator. To add to the complexity, it must also navigate and motivate itself; a Segway's navigation system takes advantage of the sensors and processor installed in all of its cargo (a handy feature!) to offset the 'heavy lifting' of directional goals.

      I believe this project may also pre-date the segway by a few years, if you look at the papers published on it.

      --

      Take care,
      Mark

      There is a solution...

    2. Re:i'm surprised this hasnt come around sooner... by Chirs · · Score: 1

      If you'd read the paper you'd know that the math is identical to that in the Segway. They break the problem down into X and Y axes and solve them separately.

    3. Re:i'm surprised this hasnt come around sooner... by m874t232 · · Score: 1

      To add to the complexity, it must also navigate and motivate itself; a Segway's navigation system takes advantage of the sensors and processor installed in all of its cargo (a handy feature!) to offset the 'heavy lifting' of directional goals.

      Segways are already shipping a robotics platform, and it's being used in mobile robots.

      Maybe you can explain what you think the advantage of this system over a Segway-based mobile robot is.

    4. Re:i'm surprised this hasnt come around sooner... by Mycroft_VIII · · Score: 1

      "Maybe you can explain what you think the advantage of this system over a Segway-based mobile robot is."

      Changes in travel direction w/o rotation of the whole bot.
      In fact all sorts of sudden directional changes and steering become much simpler.

      Mycroft

      --
      https://signup.leagueoflegends.com/?ref=4c3ed6600b6ea
    5. Re:i'm surprised this hasnt come around sooner... by veganboyjosh · · Score: 2, Funny

      In fact all sorts of sudden directional changes and steering become much simpler.

      at first glance, maybe.
      but think about this: i'm a bicycle mechanic, and tool around with various art bikes in my spare time. one idea i had was for a dual steering bike, where both wheels turned at the same time, synched up with rods and bearings. the goal was to make a bike that travels diagonally, or even sideways, once the balance issues are worked out. as i thought about this, it occured to me that if i was riding along at a steady pace on a northerly heading--say 10 mph, and turned the bars to the west a little bit--turning both wheels in the same direction--the bike and me aren't gonna all of a sudden gonna be travelling in a northwest heading. the bike and i have momentum built up, as well the wheels act as gyroscopes, which aren't all that easy to turn while they're spinning. as i go to turn the bars, i'll be tweaking the bike and me in the direction/amount i've tried to turn the bars. the bike and i will continue heading north, but we'll be sort of doglegged, riding sideways... i couldn't figure out how to get past this, so i abandoned the project...

    6. Re:i'm surprised this hasnt come around sooner... by m874t232 · · Score: 1

      Yeah, that's the obvious answer, but I don't think it works. The Segway can turn in place, and I don't see any need to be able to move sideways. In fact, for a robot intended to interact naturally with humans, sideway motion is probably a bad idea.

    7. Re:i'm surprised this hasnt come around sooner... by SnowZero · · Score: 2, Interesting

      Preface: My lab has four Segway-RMPs (RMP = robotics mobility platform). I don't work with them directly, but I know quite a bit about the work being done.

      The main advantage of the ballbot is that it can be narrower to fit in tight spaces. The footprint of an RMP isn't really any better than the other robots it replaces. The Segway's main advantages are that it is fast, it works outdoors, and it can carry fairly heavly loads. The ballbot looks like a work in progress, and the results in the paper do not indicate that the control is anywhere near as stable as the current Segways.

      In the paper they state the inevitable demise of statically balanced robots, though I'm not sure I believe that. Balance controllers are not very friendly when they trip over objects on the floor - they drive faster and faster into the object until the object yields (i.e. breaks) or the robot cannot keep up, and eventually fall one way or the other. An angry Segway is thus not very pleasant. Balanced platforms do react nicely to disturbances at the top end, however. Ballbot will have to have a smooth cover so as not to get hooked on anything, which is another problem. Ultimately, I would not be suprised to eventually see semi-balanced bases: A statically stable robot that can hinge near the bottom for dynamic actions but can come to rest while holding a position.

      Regarding the speed issue, I think I'd be happy to race an RMP against the ballbot. Even if directional changes are required for the Segway, it would be hard to beat it over a traversal of any length. Our lab is also at CMU, so a race is actually possible :) The current owner of Xavier should throw him in there too; He looks very lonely now as he has been consigned to work on a robotic crane for several years now instead of roaming the halls of Wean or NSH.

      Finally, it is quite a shame the paper doesn't even seem to acknowledge the existence of RMPs. Yes, they are not the original self-balancing robots, but they are the only one which is commercially available, and famous enough that they likely form the inspiration for this work. We've had the RMPs since 2003 (which the Segway HT obviously pre-dates), and the earliest ballbot paper was 2005.

    8. Re:i'm surprised this hasnt come around sooner... by Da+Fokka · · Score: 1

      Everyone who has played Quake knows sideway motion is the ONLY way for humans!

    9. Re:i'm surprised this hasnt come around sooner... by saider · · Score: 1

      i couldn't figure out how to get past this, so i abandoned the project...

      How about crosslinking the wheels so that when you turn the handlebars left, the front wheel turns left and the back wheel turns right. That's how the 4 wheel steering cars work.

      --


      Remember, You are unique...just like everyone else.
    10. Re:i'm surprised this hasnt come around sooner... by saider · · Score: 1

      However, for a robot that has to retrieve things from shelves for humans, moving sideways can be an advantage. Think about how a Segway would work in a library with narrow aisles (not very well).

      Being able to interact with humans does not mean acting human in every way. If I want a human assistant, I'll hire an intern.

      --


      Remember, You are unique...just like everyone else.
    11. Re:i'm surprised this hasnt come around sooner... by budgenator · · Score: 1

      actuatly I'd think connecting a drive chain to a wheel that steers would be the real show stopper. the only way arround that would be to use a drive shaft feed down through the forks, like a handelbar stem. You'd almost need a split handlebar so each hand could steer each wheel independantly, that means you can't use the handlebars for body support; now we're talking recombent. What you'd have would definately be unique.

      --
      Apocalypse Cancelled, Sorry, No Ticket Refunds
    12. Re:i'm surprised this hasnt come around sooner... by veganboyjosh · · Score: 1

      this was an option, but it wouldn't really do much in the way of changing the steering, except for making the turning radius shorter. plus, if you check out swing bikes, these are actually sort of along this line. i wanted something new...hence the goofy steering idea.

    13. Re:i'm surprised this hasnt come around sooner... by Anonymous Coward · · Score: 0

      However, for a robot that has to retrieve things from shelves for humans, moving sideways can be an advantage. Think about how a Segway would work in a library with narrow aisles (not very well).

      The Segway can turn in no more space than an equivalent ball robot requires. Furthermore, the ball robot has to turn anyway to orient its sensors and manipulators (which will be fixed to its torso because of weight/cost).

    14. Re:i'm surprised this hasnt come around sooner... by veganboyjosh · · Score: 1

      the drive chain was certainly another big obstacle...but i had a somewhat convoluted way to make it run worked out on paper, but once the "this bike doesn't really drive sideways" issue came into play, i was no longer interested... independent steering with l/r bars is an interesting idea. i've built some things where you couldn't weight the bars--like i you say--and it's more akin to riding a unicycle, in that you use your feet, legs, and butt to balance. stopping and starting are tough, but once you get going, it's alright. imagine starting to ride a bike with no hands.

    15. Re:i'm surprised this hasnt come around sooner... by jefu · · Score: 1

      If I remember correctly (from a long excursion some years back into the physics of bicycles) the gyroscopic effect is minimal at most bicycle speeds. More important is the angles in the bike's geometry (especially fork angle). As the bike moves faster though the gyroscopic effect increases, but so do other factors.

      I can't find it now in a quick google search, (though wikipedia has some interesting informaion but there was a physics professor somewhere who investigated the physics of bikes for a class - working out the differential equations and all that, and who proved that the ideas worked by scavenging lost bikes on campus and constructing bikes to prove various points.

    16. Re:i'm surprised this hasnt come around sooner... by Iron+Condor · · Score: 1

      Changes in travel direction w/o rotation of the whole bot.

      So now the robot starts bumping because it has sensors only in one side of the system.

      Or maybe you're willing to quadruple the sensory complexity for the marginal gain in mobility of an omnidirectional system? But then things like "pick up the thing in front of you" are going to become a nightmare and the whole question of control programming explodes in complexity. Contrary to your assumption, steering becomes much much harder. For a little RC car I have two sticks that say "forward/backward" and "left/right". What sticks should I have for the omnidirectional system? And when it finally arrives at the destination, you will still have to turn it to bring whatever arms or other actuators it has within reach of the object to be picked up (placed down, turned, manipulated etc).

      If the robot has a preferred sensory direction and "turns it's head" when it changes direction, then it might as well turn the whole torso.

      Not to mention that you always beat any omnidirectional system in speed with one that is optimized for speed in one particular direction. If you want fast, optimize for one-directional speed and accomodate turns.

      Really -- I consider this an expensive gimmick at best.

      --
      We're all born with nothing.
      If you die in debt, you're ahead.
    17. Re:i'm surprised this hasnt come around sooner... by SnowZero · · Score: 1

      A correction on the above post: Both the papers do mention the RMP. Looks like I scanned the papers a bit too fast. One of the authors emailed me to set the record straight.

  7. Familiar? by The+MAZZTer · · Score: 3, Informative

    It's been a loooong time since I watched the show, but didn't the Jetson's have a robot like this?

    1. Re:Familiar? by ScrewMaster · · Score: 1

      It's been a long time for me too, but I think there was a robot called "Uniblab" that rolled around on a ball.

      --
      The higher the technology, the sharper that two-edged sword.
  8. Will it have... by Tavor · · Score: 1

    ...a beak and a visor, and really shiney metal?

    --
    Windows has detected an undetectable error.
  9. Weebles... by RyanFenton · · Score: 4, Funny

    Weebles wobble, but they don't take over humanity.
    Other jokes could involve roll-on deodorant, or those kid's inflatable boxing dummies that flip back up when you hit them.

    Neat design, really. I think a lot of people's first urge would be to try and push the robot over though, which like the old Weebles, may produce that same herky-jerky roll back to stability. I wonder how much gunk would build up in the mouse-wheel of those things though - be hard to turn it over to clean with that production model. Have to have some kind of door in it to clean without taking it out.

    Ryan Fenton

    1. Re:Weebles... by Crazyscottie · · Score: 2, Interesting

      In a world that brought us self-cleaning ovens among other gadgets, I would assume that mankind is intelligent enough to build a self-cleaning spherical-wheeled robot. Then again, we all know what happens when you assume things...

      --
      Just because it can't be explained doesn't mean it isn't true. Science fits into reality... not the other way around.
    2. Re:Weebles... by dalangalma · · Score: 1

      You could always watch the videos of it being shoved on the project web site...

    3. Re:Weebles... by SubliminalVortex · · Score: 1

      They'll just go the same way mice did... use a laser. After a few laser blasts in the eye, you'll either quit knocking them over or just be too blind to find out exactly where they are to do it again.

    4. Re:Weebles... by nilky · · Score: 1

      To clean it... brushes where the wheel/sphere enters the cowling. For gunk not able to be cleaned with a brush... simply remove the wheel/sphere and replace it with a clean one. I can see a franchise in this... similar to replacing gas containers for your gas grill... let me know if you want to participate in a start-up. Names for the start-up... ball replacement services, nilkys' balls, Cum-n-clean....

      --
      "Dishonesty is one of the ugliest possible human characteristics. Being dishonest and proud about is about the only poss
  10. Killer app? by ChozCunningham · · Score: 4, Funny

    Shrink this down and fill it with ink. An automatic pen! Well, it sounds good on paper...

    1. Re:Killer app? by Anonymous Coward · · Score: 0

      A very large piece of paper, at that. I wonder what it sounds like on stairs?

  11. Ooops by GeoSanDiego · · Score: 3, Funny

    Help. I am a robot and I have fallen and I can't get up.

    1. Re:Ooops by I+Like+Pudding · · Score: 5, Funny

      Incorrect. Weeblots wobble, but they don't fall down.

    2. Re:Ooops by Short+Circuit · · Score: 4, Funny

      That's an amusing mental picture:

      A robot ends up dragging and whipping its body all over the lab, smashing and breaking things in an effort to right itself.

      Of course, I assume they have it give up if it the torso passes a certain inclination.

    3. Re:Ooops by tomhudson · · Score: 2, Insightful

      FTFA: "Ballbot has an onboard computer that reads balance information from its internal sensors, activating rollers that mobilize the ball on which it moves -- a system that is essentially an inverse mouse-ball drive. When Ballbot is not in operation, it stands in place on three retractable legs."

      One word: lint. This sucker will fall down as soon as the rollers get clogged up, same as your mouse stops working.

      Or try spilling some oil from your salad on the ground. Or leave a few pieces of duct tape, sticky side up, lying around,

      I'll welcome our robot overlords to follow me ... down this flight of stairs, you motherf*ckers! :-)

      "Hollis is working to prove that dynamically stable robots like Ballbot can outperform their static counterparts."

      Humans are also dynamically stable. Even when we're standing still, we're dynamically balanced, as muscles alternately contract and release to maintain our balance. He doesn't have to "prove" this - its something evolution worked out long ago.

    4. Re:Ooops by WilliamSChips · · Score: 1
      I'll welcome our robot overlords to follow me ... down this flight of stairs, you motherf*ckers! :-)
      I only have one word for you: LEV-I-TATE!
      --
      Please, for the good of Humanity, vote Obama.
    5. Re:Ooops by AGMW · · Score: 1
      Or try spilling some oil from your salad on the ground.

      Salad! On slashdot? The very idea!

      We like MEAT! MEAT I tells ya, MEAT!

      Often mechanically recovered and made into patties.

      --
      Eclectic beats from Leeds, UK
      handmadehands.co.uk
    6. Re:Ooops by tomhudson · · Score: 1

      Which is why you won't mind spilling some salad oil.

      Now I could understand you being upset if I had suggested something sacreligious, like spilling beer ...

      Of course, the best would be a spot of crazy glue. One time, as a practical joke, some people at Dorval's Dept of Transport office put some crazy glue under another woman's desk right before she got in. She sat down, and her shoes got "welded" to the floor. 5 minutes later, a couple of RCMP guys (who were in on the prank) came in with a phonied-up arrest warrant for her. That's when she discovered she couldn't stand up ...

  12. Slow News Night by MSTCrow5429 · · Score: 1

    Um, ok, so what? Is there a lack of robotic performers at the circus?

    --
    Slashdot: Playing Favorites Since 1997
  13. ok, it can stand but by jboker · · Score: 1

    will it be compatible with my ipod?

    also, when will my college get one i can play with?

  14. I for one... by Anonymous Coward · · Score: 5, Funny

    welcome our big balled overlords

    1. Re:I for one... by Telvin_3d · · Score: 1

      I see my fame is spreading

    2. Re:I for one... by Scaba · · Score: 2, Funny

      And who doesn't enjoy a sack of Schweaty balls during the holidays?

    3. Re:I for one... by cno3 · · Score: 1

      The Japanese call 'em Tanuki.

  15. A boon to senior citizens by grammar+fascist · · Score: 4, Funny
    They are also much more agile, since they can be pushed out of the way easily without falling over.

    You know, I've always hated that whenever I try to push a robot out of the way, I fall over.

    This feature will be a boon to senior citizens especially.
    --
    I got my Linux laptop at System76.
    1. Re:A boon to senior citizens by philo_enyce · · Score: 3, Funny

      better make sure they buy their robot insurance...

    2. Re:A boon to senior citizens by Otto · · Score: 1

      Yes, because robots take old people's medicines and eat them for fuel.

      --
      - Give a man a fire and he's warm for a day, but set him on fire and he's warm for the rest of his life.
    3. Re:A boon to senior citizens by paulthomas · · Score: 1

      Pak.

      Chooie.

      Unf.

    4. Re:A boon to senior citizens by P3NIS_CLEAVER · · Score: 0

      wouldn't it be easier for the robot to zap them if they try to push it?

      --
      Please sign petition to restore sanity to our banking system!!!

      http://financialpetition.org/
    5. Re:A boon to senior citizens by Kamineko · · Score: 1

      Bernard, Oozo isn't a robot: he's an inflatable clown.

    6. Re:A boon to senior citizens by Anonymous Coward · · Score: 0

      Yeah, I know what you mean. I am constantly pushing robots out of the way. I mean, I'm always surrounded by these stupid robots, and they're always getting in my way, so I have no choice but to push them over. This new invention sounds great, because now I can push my robots over without destroying them, which is what usually happens.

  16. How is this "much more suitable"? by SuperBanana · · Score: 4, Insightful

    The promise of such dynamically stable robots is that they can be much taller without having to have a wide base, making them much more suitable for working with humans

    Huh? How does thinner = autmatically "much more suitable"? I just don't get it. We're used to working with people, and that extends to their personal space, which is substantially larger than the person. Who says that human/robot 'synergy' has anything to do with how fat the robot is?

    Nevermind that all that active movement to keep standing upright means lots of battery drain, and irritating/distracting movement to said humans. We find coworkers that tap pens on their desks annoying, but they think this will 'fly'?

    Oh, when the battery runs out of juice, or a motor (or its controller) goes dead- hilarity ensues...

    1. Re:How is this "much more suitable"? by catbutt · · Score: 1

      I suspect the movement to keep standing up can be minimized to being indetectable. The quicker and more accurately it can compensate, the less the motion has to be.

    2. Re:How is this "much more suitable"? by dalangalma · · Score: 1

      Importantly, a six-foot-tall conventional robot would have trouble fitting through doors because of the large base it would require to stay stable.

    3. Re:How is this "much more suitable"? by aug24 · · Score: 1

      Suitable because human shaped, so able to navigate our environment (better han a 'traditional' short, wide, wheeled 'bot anyway). J.

      --
      You're only jealous cos the little penguins are talking to me.
    4. Re:How is this "much more suitable"? by Idarubicin · · Score: 2, Insightful
      Huh? How does thinner = autmatically "much more suitable"? I just don't get it. We're used to working with people, and that extends to their personal space, which is substantially larger than the person. Who says that human/robot 'synergy' has anything to do with how fat the robot is?
      Find a wheelchair. Try to navigate through your daily routine. Ever notice how most human spaces aren't designed for a wide chassis? Imagine if ''everybody'' was in a wheelchair. Try passing the 'slow walker' (you know who I'm talking about) in the hallway without running into the guy coming the other way. Tall and skinny--in other words, more human-shaped--is an advantage in working with humans, because humans work in human-shaped places.

      For that matter, you're less likely to trip over a robot that's five feet tall than one that's two feet tall; tall robots are easier to see. Wide, low robots would be a nuisance to work with. It would be like having dozens of short, fat children always getting in the way.

      --
      ~Idarubicin
    5. Re:How is this "much more suitable"? by budgenator · · Score: 1

      when the battery runs out of juice, or a motor (or its controller) goes dead- hilarity ensues... hummm inherently unstable computer controlled ballbot ballancing on an insulated urethane covered ball, dry winter day and static electricity; the possibilities for amusement are endless.

      --
      Apocalypse Cancelled, Sorry, No Ticket Refunds
    6. Re:How is this "much more suitable"? by sm62704 · · Score: 0, Flamebait

      Huh? How does thinner = autmatically "much more suitable"?

      Good point, I know some fatasses that can barely fit through a door.

      --
      mcgrew's razor: Never attribute to stupidity that which can be explained by greedy self-interest
  17. Piers Anthony alien race from Cluster? by andykuan · · Score: 2, Interesting

    I vaguely recall there being an alien race in Piers Anthony's Cluster series that balanced and moved on a single large ball. Anyone remember the name of the race? It'd be wonderfully nerdy if the CMU folks renamed their robot after these aliens. I also vaguely recall these aliens having sex by going around and around in a circle rolling continuously over a trail of slime. Maybe the CMU robots can do that too...

    1. Re:Piers Anthony alien race from Cluster? by slughead · · Score: 1

      Anyone remember the name of the race? It'd be wonderfully nerdy if the CMU folks renamed their robot after these aliens. I also vaguely recall these aliens having sex by going around and around in a circle rolling continuously over a trail of slime.

      You remember how they have sex but you can't remember their name?.. ?

    2. Re:Piers Anthony alien race from Cluster? by NinjaFarmer · · Score: 5, Funny

      You remember how they have sex but you can't remember their name?.. ?

      I can't count the times I've asked myself something similar in the morning.

    3. Re:Piers Anthony alien race from Cluster? by William_Lee · · Score: 2, Informative
      Actually, there is a race of aliens called mulefa in Philip Pullman's 'The Amber Spyglass' trilogy that roll around on a single large ball. I think they should name them after these. BTW, It's a pretty good read.

      http://en.wikipedia.org/wiki/Mulefa

    4. Re:Piers Anthony alien race from Cluster? by andykuan · · Score: 1

      It's always easier to remember what something does rather than what something is named. Besides, at the time, my hormonally-activated 12 year-old mind picked up on anything sex-related no matter how, uh, alien.

    5. Re:Piers Anthony alien race from Cluster? by Maru+Dubshinki · · Score: 2, Informative

      The Mulefa (from what I remember of reading the trilogy and having me memory refreshed by the Wikipedia article) don't use a single ball or wheel; they use several wheels and stick their limbs through them. I don't think the Piers Anthony ones used a single ball either (didn't they use pillars as rollers?).

      --
      Enquiring minds want to know!
    6. Re:Piers Anthony alien race from Cluster? by Anonymous Coward · · Score: 0

      I think it was called Polarian. Tallish black critter. It's in Barlowe's guide.

    7. Re:Piers Anthony alien race from Cluster? by winkydink · · Score: 1

      This is slashdot. Unless you are left-handed, you call it Righty.

      --

      "I'd rather be a lightning rod than a seismometer." -Ken Kesey

    8. Re:Piers Anthony alien race from Cluster? by smoker2 · · Score: 3, Informative
      The Polarians.

      "It was a tear drop shaped thing with a massive spherical wheel on the bottom and a limber tentacle or trunk at the top"

      They communicated by vibrating a smaller spherical ball on the end of the trunk against another object (ie the ground). AFAICR

      As for the sex, they chase each other round in an ever decreasing circle, incensed by each others trails, until they touch. Then the male polarian releases his large wheel and joins onto the females large wheel so that they both join at the seam. Then they spin the wheel until it springs into life and forms a child. After that, the female pinches the males communication wheel to use as a transit wheel, leaving him temporarily mute, and he gets to use his old transit wheel again.

      I can't believe I just described that !
      I do have the trilogy on the shelf though. It is the Cluster Trilogy by Piers Anthony, consisting of Vicinity Cluster, Chaining the Lady, and Kirlian Quest. Weird books, based around the transference of a beings aura across galactic distances and that aura inhabiting living "hosts" once transferred. Lots of tarot references too.
    9. Re:Piers Anthony alien race from Cluster? by cvas · · Score: 1

      I can't count the times I've asked myself something similar in the morning.

      Instead of Roman numerals you should count using a more modern number system, specifically one that can represent zero.

    10. Re:Piers Anthony alien race from Cluster? by RossumsChild · · Score: 1
      Oblig. Slashdotter's-get-no-sex joke:

      Of course not! Zero is an uncountable number!

    11. Re:Piers Anthony alien race from Cluster? by budgenator · · Score: 1

      Piers Anthony is such a pervert that the sex is the only reason most people read his stuff. I'm surprised that a lot of his stuff isn't sold in adultbook stores only. Read "Firefly" and tell me he's not out there.

      --
      Apocalypse Cancelled, Sorry, No Ticket Refunds
    12. Re:Piers Anthony alien race from Cluster? by ediron2 · · Score: 1

      (thanks for the best laugh in a week... I put up with a lot of useless crap readin' slashdot just for this sort of wanderin' humor. From robots with one ball to piers anthony's wierd sexual proclivities to GP to your comment... priceless)

  18. Obligatory by eosp · · Score: 1

    "I'm not sure this floor is entirely stable."

  19. Not quite by XanC · · Score: 1

    Rosie was perched atop a single roller skate, from what I can tell and remember.

    http://upload.wikimedia.org/wikipedia/en/a/aa/Jets ons.jpg
  20. NEW technology?!? by snafu109 · · Score: 4, Funny

    Using a mechanism similar to a ball mouse

    Pfft... Let me know when they've developed a mechanism similar to a laser mouse.

    1. Re:NEW technology?!? by mblase · · Score: 1

      Pfft... Let me know when they've developed a mechanism similar to a laser mouse.

      Didn't the Quintessons have something like that?

    2. Re:NEW technology?!? by Ant+P. · · Score: 1
    3. Re:NEW technology?!? by Anonymous Coward · · Score: 0

      Pfft... Let me know when they've developed a mechanism similar to a laser mouse.

      Yeah, I'd hate to have to clean the mouse ball gunk off the rollers every month on that bad boy.

    4. Re:NEW technology?!? by Anonymous Coward · · Score: 0

      That's insane! That's completely skipping the necessary optical-but-not-laser step!! Madness, I say!

  21. Videos! by Anonymous Coward · · Score: 0
    1. Re:Videos! by sdriver · · Score: 1

      They call that a push? More like a tap.... Let me give it a push and that thing will be on the ground. :P

    2. Re:Videos! by dalangalma · · Score: 1

      You could probably jump onto this robot without it falling over. I have never managed to shove it until it falls.

  22. Power issues, thinness by Mark+Programmer · · Score: 5, Informative

    FTA:

    "When Ballbot is not in operation, it stands in place on three retractable legs."

    So 'nervous balance' motions won't be necessary, one could assume.

    As for the thinness issue: it is precisely the personal space issue that makes a thin robot useful in a crowded public space. Our perception of personal space factors in the personal space of the other person. So a robot that is as wide as one's 'personal space bubble' causes people interacting with it to give it even more room. The thin chassis on this robot alleviates that problem by only taking up an amount of space roughly equivalent to the human torso, so that a person's "personal space guess" comes out accurately.

    --

    Take care,
    Mark

    There is a solution...

    1. Re:Power issues, thinness by Breakfast+Pants · · Score: 1

      After a few people get smashed by these things, when they break down, people's "personal space guess" will begin take into account the height of the damn thing.

      --

      --

      WHO ATE MY BREAKFAST PANTS?
    2. Re:Power issues, thinness by quakeroatz · · Score: 1

      "When Ballbot is not in operation, it stands in place on three retractable legs."

      A huge ball with THREE retractable appendages. Don't leave this robot at home with your GF.

  23. Movie by acu_gumby · · Score: 1

    They already did this in Robots

  24. Re:Not So Strange by DumbSwede · · Score: 1

    Maybe I'm not the average Slashdotter, but I have had sex with lots of women (many of whom seemed liked aliens) and can't remember most of their names.

  25. Re:Not So Strange by slughead · · Score: 1

    Did they leave a slime trail too?

  26. Technological Threat by MsGeek · · Score: 2, Informative

    "Technological Thret" is an 1988 animated short directed by Bill Kroyer. It centers around the introduction of an office robot that rolls around on a single ball in the bottom of its structure. One of the interesting little details about the short was that the robots were first animated in a computer, then rotoscoped in 2D to fit with the rest of the 2D short. The bot in the short didn't look *that much* like the Ballbot, but just enough to bring it to mind.

    --
    Knowledge is power. Knowledge shared is power multiplied.
  27. Re:Not So Strange by DumbSwede · · Score: 3, Funny

    Only the ones without legs

  28. Note to guy in red shirt who works on robots by TheModelEskimo · · Score: 1, Offtopic
  29. Ballmer's been replaced with a bot? by Anonymous Coward · · Score: 0

    "Researchers at Microsoft's Redmond Institute have developed a new kind of robot called the Ballmerbot that balances on a steel ball. Using a mechanism similar to a ball mouse, the Ballmerbot uses rollers to drive its single, spherical wheel and balance in place or glide around the stage. The promise of such dynamically stable robots is that they can be much taller without having to have a wide ass, making them much more suitable for shouting "developers, developers developers" at humans. They are also much more agile, jumping around the stage better without falling over. You can read the press release or check out the project's web page when it recovers from sweaty armpit syndrome."

  30. Wow by nascarguy27 · · Score: 1

    A dupe within the summary. Two of the links point the exact same press release. Anyway, I'd like to see a humanoid robot balance on a ball while trying to fix a problem with windows. Some people cannot do either.

    --
    Funny createSig(Witty remark, Odd reference)
    {
    return (Funny)remark + (Funny)reference;
    }
  31. Ballbot in action by novus+ordo · · Score: 4, Informative

    There are some videos available on youtube, in case you want to spare them some bandwidth.

    --
    "You're everywhere. You're omnivorous."
    1. Re:Ballbot in action by OldManAndTheC++ · · Score: 1

      I notice the first video in the list is titled, "Ballbot Balances Itself On One Ball"

      Ouch!

      --
      Soylent Green is peoplicious!
    2. Re:Ballbot in action by snafu109 · · Score: 1

      I'd actually be more worried about sparing YouTube some bandwidth.

      Millions of dollars of capital -> YouTube -> ? isn't really a viable business model.

  32. supermarket shopping cart syndrome by zogger · · Score: 2, Insightful

    first wad of gum that thing rolls over might bring it to a stop, or make it list to port or something ->
    roll THUNK roll THUNK. And stairs? Or going outside?

      I like old school, traditional sci fi robots, humanoid looking, bipedal walking type. If they need to get fancy, drop in the jackie chan mod chip or something.

    1. Re:supermarket shopping cart syndrome by KlaymenDK · · Score: 1

      I can't tell if it was intentional of you, but I'm laughing my bum off.

      Grad student: oohh yeah, that old geezer *did* say there were downsides to those old 'balled mice'. I s'posse this is what he meant.

      On a different note: what exactly is 'port' on a (presumably non-seafaring) cylindrical omnidirectional robot????

  33. Well, at least the robot will be safe... by djeca · · Score: 1
  34. Slashdot losing its touch? by bitspotter · · Score: 1, Interesting

    "You can read the press release or check out the project's web page when it recovers from traffic."

    Wait, so you mean it was slashdotted //before// it was posted to Slashdot?

    Slashdot must be sliding into irrelevance...

    That's right: SLASHDOT IS DYING...

    1. Re:Slashdot losing its touch? by Anonymous Coward · · Score: 0

      Wait.... /.ing the article before it's posted is WAY more efficient than making us all LOOK at the article before it goes down...

    2. Re:Slashdot losing its touch? by Anonymous Coward · · Score: 0

      HAS NETCRAFT CONFIRMED IT?

  35. Who named this??? by I+Like+Pudding · · Score: 2, Funny

    This thing should star in a 1960s Japanese movie.

    *speaking into wrist communicator* Ballbot! TRANSFORM!

  36. Very nice by Animats · · Score: 5, Insightful

    That's a nice system. There are some annoying limitations, though. It's tough to change the orientation of the robot, although you can handle that if you have a rotational joint further up so you can rotate the torso. Small diameter balls have the same problems as small diameter wheels - it's easy to get stuck in small depressions. That's why the Segway has such big wheels. And driving a sphere is always a tough problem mechanically. Most of the solutions have trouble with dirt accumulating on the drive wheels, which is why optical mice have replaced ball mice. It's possible to build a spherical electric motor, and that might be the way to go if this concept turns out to be useful.

    It's good to see all this activity in self-balancing systems again, having worked on this around 1994-1995, and seen others working on it in the 1980s. Today, you can buy so much of what you need off the shelf, like good INS units. We used to waste too much time building custom stuff.

    1. Re:Very nice by Jeremi · · Score: 1
      It's tough to change the orientation of the robot, although you can handle that if you have a rotational joint further up so you can rotate the torso.


      Hmm, how about a couple of internal counter-rotating flywheels.... spin them both up simultaneously, and then whenever you want to turn the robot, apply a bit of brakes to just one of them.

      --


      I don't care if it's 90,000 hectares. That lake was not my doing.
    2. Re:Very nice by BasilBrush · · Score: 1

      A biped with one of two of these balls instead of feet might be interesting.

  37. Problems by DigitAl56K · · Score: 1

    What happens when the robot tilts forward and so needs to move forward to correct its balance, but runs into something before it has moved far enough to correct itself? Does it simply fall over?

    1. Re:Problems by Tablizer · · Score: 3, Funny

      What happens when the robot tilts forward and so needs to move forward to correct its balance, but runs into something before it has moved far enough to correct itself? Does it simply fall over?

      My kids would probably take it down with a single Hotwheel toy. They took me down a few times that way. Too bad they are too young to sue.

    2. Re:Problems by wkitchen · · Score: 1
      What happens when the robot tilts forward and so needs to move forward to correct its balance, but runs into something before it has moved far enough to correct itself? Does it simply fall over?
      Of course. Short of having something to push against or grab onto, and an appendage with which to push or grab it, falling over is the only option in that situation. Humans sometimes fall over for the same reason. That's what happens when you trip over something. Some object impedes the motion of your foot, which prevents you from correcting your balance. Another example that I've personally experienced is when I got my bicycle wheels in a rut going down a rather steep hill. I could no longer correct my balance because the bike could not freely move side to side. The result was me lying injured (all minor, fortunately) on the ground next to my damaged bike. If you need to correct, but can't, you fall.
  38. Link to videos by krunk4ever · · Score: 1
    1. Re:Link to videos by Meltir · · Score: 1

      When those links turn the server into a smoking pile of plastic, here is a google video of it in action: http://video.google.com/videoplay?docid=9117167457 978282229

  39. how far can it fall? by krunk4ever · · Score: 1

    I'm curious to how far can it fall but still be able to not recover itself. Look at the robot, it seems to be built where it can't be tilted too much, but having one of those bust up that Michael Jackson move (where he tils and awesome touches the floor and comes back up) would be neat! Or it could be like one of this bobbing clowns where you can punch it down, but it'll always stand back up.

    1. Re:how far can it fall? by dino213b · · Score: 1

      I am guessing the recovery depends on its ability to accelerate the ball...and looking at videos, it doesn't look like much. This reminds me of the first year's DARPA challenge where a motorcycle team entered a contest and failed when the poor thing couldn't right itself after a spill.

      As if autonomous navigation and imaging wasn't hard enough..

      Also reminds me of the futurama episode where Bender walks over a rope..and to tempt fate, he whips out a unicycle out of his storage compartment and rides it instead of walking over the rope.

      Just because we can do it doesn't mean we should. Except when it comes to research..then it's perfectly acceptable to be absurd.

  40. Segway? by RandySC · · Score: 1

    How about making a Segway type vehicle with a single ball. This would be awesome for cornering, as it could bank like the Carver 3 wheeler http://www.carver-europe.com/

    I almost fell off a Segway during my first few hours of riding during some high speed cornering.

    --
    Organization: alphabetical, sometimes numerical or messy
    1. Re:Segway? by Amouth · · Score: 1

      doing better than the GWB the leader of the free world..

      he fell off it sanding still in under 20 sec

      --
      '...if only "Jumping to a Conclusion" was an event in the Olympics.'
  41. Nothing new: simple Linear controller by Anonymous Coward · · Score: 1, Interesting
    This is engineering 301 and nothing new. Here's a paper that shows the control algorithm and details of the Ballbot. Here's the PDF version. They're using a simple linear model; nothing new.

    They could make it really robust and easier to develop had they used a fuzzy logic controller. But that would be a little too imaginative for American engineers. The Japanese have been doing stuff like this for more than 20 years. Some old American guy rolls a bowling ball across the floor, calls it a "Ballbot", and everyone jumps up and shouts.

    1. Re:Nothing new: simple Linear controller by Anonymous Coward · · Score: 0

      This is engineering 301 and nothing new. Here's a paper that shows the control algorithm and details of the Ballbot. Here's the PDF version. They're using a simple linear model; nothing new.

      That's fine in theory. They made the damn thing work, which is why they made it to a peer-reviewed conference. The paper doesn't claim "the most advanced controller ever", and why should it, when a traditional controller suffices? They simple said "look, you really can make a robot that works like this".

      They could make it really robust and easier to develop had they used a fuzzy logic controller. But that would be a little too imaginative for American engineers.

      What the hell is the fascination with fuzzy logic? There is nothing in fuzzy logic that stastics has not known for 50 years. "fuzzy" is a pretty much a way of saying "I don't feel like working out the true probability distributions". At CMU we don't do a lot of fuzzy logic (nor in the US in general anymore), but that's because we have a statistics department. We try learn what we can from other disciplines, rather than reinventing the wheel. If you wanted better control of a robot, I'd suggest an adaptive controller of some sort, rather than a late 1990s fad.

      The Japanese have been doing stuff like this for more than 20 years.

      Refernce to peer-reviewed paper, please. And no, an externally driven inverted pendulum doesn't count, that's been done since the dawn of modern control theory.

      Some old American guy rolls a bowling ball across the floor, calls it a "Ballbot", and everyone jumps up and shouts.

      And what have you added? The Robotics Institute has excellent contacts with Japan (and Germany). Robotics is a truly international field, regardless of where you go. I know at least 5 people in CMU's School of Computer Science that work on robotics and have spent at least 3 months in Japan working on robots, and two of them have spent over a year there. The prominent RI scientist Takeo Kanade has helped to encourage numerous Japanese students to study in RI. Many students from India and China come to study as well. Put that all together, and combine it with the fact that RI and SCS produce a lot of good robotics work, and I think the result is pretty damn reasonable. Feel free to prove "us Americans" wrong.

      P.S. I'm not affiliated with the ballbot project in any way, but I work on robotics at CMU.

    2. Re:Nothing new: simple Linear controller by m0nstr42 · · Score: 1

      They could make it really robust and easier to develop had they used a fuzzy logic controller. But that would be a little too imaginative for American engineers.

      What the hell is the fascination with fuzzy logic?

      There isn't any, or at least there hasn't been any with serious control theorists in the last 10 years since the fad went away. This guy is full of crap. Fuzzy control is just linear control with a non-smooth control surface.

    3. Re:Nothing new: simple Linear controller by chiph · · Score: 1

      Perhaps if they'd used a heavier ball (think un-drilled bowling ball covered in urethane) that would provide the dampening they need. I noticed in the YouTube videos that a couple of shoves got it oscillating, and it took a while for it to return to a (more or less) stationary position.

      Chip H.

    4. Re:Nothing new: simple Linear controller by Anonymous Coward · · Score: 0

      There isn't any, or at least there hasn't been any with serious control theorists in the last 10 years since the fad went away. This guy is full of crap. Fuzzy control is just linear control with a non-smooth control surface.

      Oh, I didn't mean it to sound like controls people are currently obsessed with it; I should have said "I don't understand some people's obsession with fuzzy logic". No doubt these are the same people who think Neural Networks and Genetic Algorithms are the most advanced machine learning methods, and can solve all AI problems.

      The only thing I'll fault controls people with is that IMO they don't do a good enough job explaining the more advanced methods so that people outside of the field can understand it. That's my theory as to why everyone who isn't an ME ends up using only PID controllers (or PD controllers in the case of graphics people, who call it a damped spring).

    5. Re:Nothing new: simple Linear controller by m0nstr42 · · Score: 1

      It's OK, I was trying to agree with you :)

      And for the record, I'm not an ME (sort of... it's complicated) and I've only used a PID once. I do like the ME approach to controls, though... mmmmgeometry.

  42. Terrible Secret by Wilson_6500 · · Score: 4, Funny

    With the advent of robots that are immune to pushing, the eternal question becomes more relevant than ever: do you have stairs in your house?

    1. Re:Terrible Secret by Hadlock · · Score: 2

      As long as you've got a pusher robot, you should be protected.

      --
      moox. for a new generation.
    2. Re:Terrible Secret by icepick72 · · Score: 1

      This robot is only immune to pushing when the external environment is of the proper type. For example, if you push the robot when it's inside a space that is too close for recovery, or if you push it into a wall, then it cannot recover and simply falls over. A power glitch, a software bug ... and you had better hope there are backup systems or else you have yourself and hole in the drywall and a bowling ball. I assume the inner working of that thing are for the purpose of balancing itself, which is a neat concept, but the robot doesn't do anything useful ... yet. The big balancing system will have to become smaller to be fit into robots that do other things.

    3. Re:Terrible Secret by SnowZero · · Score: 1

      Please see this for background information.

      PAK CHOOIE UNF

  43. They must have had a few bugs in the software by viking2000 · · Score: 4, Funny

    In the videos they have threaded a big innertube over the structure, presumably to cushion the fall every time it tipped over.

    Since the purpose of this structure is to interact with humans, They should just install big inflatable boobs and ass instead.

    1. Re:They must have had a few bugs in the software by anakin876 · · Score: 1

      yes - but a single uni-boob and a uni-ass that wrapped all the way around would loook kind of weird.

  44. It would have been easier... by 70Bang · · Score: 1


    ...had they used an earth ball. Not the dinky ones which are startig to pop up, but the ones which showed up at jugging festivals (alongside torches, unis, antique big wheeled bikes, etc..

    The balls would have much higher centers of gravity (making the balance extremely easy) and fine tune the machinery required to remain balanced. From there, they could work their way into smaller and smaller balls, tweaking their materials (hardware & software) to a point of where it could handle pretty small pieces.

    Starting large and migrating to smaller balls would mean you have fewer variables to deal with at the beginning and make the process simple[1] along the way.
    _______________________________
    [1] "Make things simple, not simpler." -Erasmus

    1. Re:It would have been easier... by Anonymous Coward · · Score: 0

      If the ball's big enough, the bot can have enough mass below the horizontal equator of the ball to be intrinsically stable. Not that this one does that. You have a funny idea about stability if you think 'higher is better'.

      Besides, as has been mentioned elsewhere, the larger the wheels the better they deal with surface deformities. Too small balls, and they'd be unable to run over a grain of sand.

      AC

  45. A Jetson's Ripoff by Jehosephat2k · · Score: 1

    Then there is Rosie http://mrl.nyu.edu/~perlin/experiments/rosie/ . Does this mean he can't patent the idea due to prior art?

  46. Angular momentum could help? by aschoeff · · Score: 1

    Stability is an issue, as some people have commented here. I myself can't imagine how you can realistically balance what is essentially a pencil on its sharpened point for any length of time, especially with batteries draining away. The robot must look like some sort of drunk date, listing to and fro with some rather large amplitude.

    If it isn't in there already, how about some sort of flywheel/gyro (axis of rotation vertical, along robot chassis) built into the robot, at a height in the chassis that globally minimizes the effective lever distance any lateral force it may receive along its height? All that inertia would do a lot towards keeping the thing upright, just like a top. With good bearings you would drain very little, and it could even be used as a method to rotate the whole robot quite quickly, albeit in one direction unless you have some sort of intermediating transmission, by having a variable clutch to transfer some of the angular force from the flywheel to the chassis.

    Hey, you could even make the flywheel out of materials so it also works like one of those static energy storage devices. Here is a somewhat decent link to a patent for one, but it's been covered on our beloved Slashdot before here, here, and here. This would give you a combination source of both electrical and direct mechanical energy for the robot to do with as it would. Theoretically, it could even be the primary power source, eliminating the need for heavy and expensive batteries, but the flywheel would have to spin really fast to store enough energy, and wouldn't have a way to spin back up during quiescent moments.

    By the way, the robot building (and fighting!) community explored and exploited such dynamics thoroughly, spinners anyone?

  47. How does this thing turn in place? by mad_minstrel · · Score: 0, Offtopic

    While this certainly is impressive, there is at least one problem I see: How is this thing going to turn in place? It doesn't need to do that to move, but it will have to if it needs to, say, pick up an object. Especially on a flat, hard surface, like a wooden floor, where there's virtually no friction between the floor and the ball. One solution is that it could throw a weight around, but that's hardly gracefull. Also I wonder if it could overcome a simple doorstep, or curb.

    --
    May the source be with you.
    1. Re:How does this thing turn in place? by stunt_penguin · · Score: 1

      One of these maybe?

      --
      When the posters fear their moderators, there is tyranny; when the moderators fears the posters, there is liberty.
    2. Re:How does this thing turn in place? by Anonymous Coward · · Score: 0

      Since the robot has to continuously move back and forth ever so slightly to keep it's balance, isn't this going to work in pretty much the same way as Foucault's Pendulum where the Earth will rotate beneath the robot.

      I guess a better question would be... How do you prevent the robot from turning in place?

  48. Duck Tales by nilbog · · Score: 1

    Blithering Blatherskite!

    --
    or else!
  49. Amazing. by tomato · · Score: 2, Insightful

    This is the first time I've posted to Slashdot in nearly 5 years.

    That robot in the videos is truly amazing - graceful, simple, and harmonious.

    This is what we need - systems that co-operate with humans, can be pushed around, and stick to their own personal space.

    We need more beauty and grace in the world, not less.

    I work in technology, but I also study feng shui, and that robot has it in abundance.

    1. Re:Amazing. by Anonymous Coward · · Score: 0

      Yeah, until it tries to go over even the slightest bump and gets stuck. This thing couldn't even make it over the transition from hard floor to carpet in my house.

      This is a very-flat-surface-only mechanism.

    2. Re:Amazing. by 4D6963 · · Score: 1

      This is the first time I've posted to Slashdot in nearly 5 years.

      The first? According to your user page it's more like your 14th.

      --
      You just got troll'd!
    3. Re:Amazing. by tomato · · Score: 1

      It feels like 5 years :)

      I looked at my userpage, but the year is left off all the datestamps and I'm not leet enough to work out how to get Slashdot to tell me which year the comments were made.

    4. Re:Amazing. by 4D6963 · · Score: 1

      You just roll over the link to the article your comment is attached to, and you look at the URL. The first number is the year.

      --
      You just got troll'd!
  50. A Balancing Act.... by SubliminalVortex · · Score: 1
    While it's nice that they have a robot that will eventually right itself on a sphere, it would be even better if they can have one that will do the same on a single wheel (like a unicycle). I'm sure in the spherical case, it is considering the gravity on all sides for which it can mostly compensate with several servos and rollers.

    In that for the wheel, it would need to know that it would have to shift its own weight to make a turn (as well as shift its weight again to go straight) as opposed to just moving a few different rollers.

    It's like learning how to ride a bike; we're just watching others ride bikes that don't exist. <grin>

  51. Didn't someone do this with a pogo stick? by ishmalius · · Score: 1

    I seem to recall this from several years ago. If so, then the balancing would be even tougher, as the bouncing motion would need to find convenient points of opposition on the floor, and plan the bounces to try to hit them.

  52. inappropriate conduct? by angelwalkwithme · · Score: 0

    It's pretty cool. But my only beef is that the robot will suffer from the fate of all computer mice. Dirt and debris clogging up the sphere. This will eventually lead to the robot to begin complaining and eventually it will urge the human to: "Please clean my balls." And I'll be damned if I will be cleaning anybody elses balls, robot or human!

  53. good 1, made me spit coffee by supercrisp · · Score: 1

    thanks for making the /. cliche pay

  54. PAK CHOOIE by WantTheHumpus · · Score: 1

    ANSWER: ITEMS PUSHED - GRANDMOTHER

    TARGET - STAIRS

    IT WAS THAT I DID

    PAK CHOOIE

    a b c d e f g h i j k l m
    n o p q r s t u v w x y z
    a b c d e f g h i j k l m
    n o p q r s t u v w x y z

  55. Call it Marvin by jagdish · · Score: 1

    Yes Marvin the depressingly sad robot with the big single spherical head. You must name him Marvin, you hear me. You must. Do you know whats depression? I'm sure you can tell me the dictionary definition but have you ever felt it,experienced it? I have. It gnaws at you with everything its got till you feel raw inside. It lashes deep within you till you experience the same pain and wanton lust. It can drive a person crazy. It can drive a person insane. It can be at once the most wonderful feeling and the most terrible. Wonderful cause you know what you want for a change, and terrible because you are afraid of what might happen if you dont satiate the feeling. Not that it matters anyway, Cause we are all going to die soon. (3am here,too much coffee)

  56. All this robot talk... by quakeroatz · · Score: 1

    All this robot talk is making my dual spherical wheels itchy.

  57. Note to Photographer by eMbry00s · · Score: 0

    Use a tripod so you don't have to use such ridiculously high ISO settings, man! (really, wtf)

  58. Cool.... by Nonillion · · Score: 1

    A real life 'Bigweld'

    --
    "I bow to no man" - Riddick
  59. Gyro by bobbuck · · Score: 1

    Or you could use a gyro to turn.

  60. These lil' robots by Automat · · Score: 1
    Practically roll out the sales floor themselves.

    --
    Automation applied to an inefficient operation will magnify the inefficiency.

  61. Efficiency, dynamic energy need by A+nonymous+Coward · · Score: 1

    Seems to me there are two energy disadvantages. One, a regular wheel can have bearings and coast quite well. A single ball -- don't see how. Seems to me that you want the weight to be supported as close to top dead center as possible, because as the support points move towards the edge, you have more stress of the weight trying to slide down and outwards, which means more friction against the ball. Imagine worst case, where the weight is supported at .0001 degrees above the horizontal point ... it would take a tremendous force to keep it from sliding that last fraction of a degree and falling off. Maybe you could have the weight supported at the top with a separate load bearing, but how could one bearing adjust to movement in both directions? Anyway, all that extra friction means that much more power required to just move. I wonder how far it can coast?

    Second, a legged robot, even just two legs, needs no energy to stay upright when not moving, unless it's windy. Considering the current state of battery tech, this seems impractical. I wonder how much extra energy it takes to stay upright when motionless.

    1. Re:Efficiency, dynamic energy need by Chandon+Seldon · · Score: 1

      For a two legged robot, you probably don't want it to be able to stand stable with zero power. Human's sure can't do that, and humans are what you'd be trying to copy. I'm betting you lose a good chunk of control for very little gain.

      --
      -- The act of censorship is always worse than whatever is being censored. Always.
    2. Re:Efficiency, dynamic energy need by A+nonymous+Coward · · Score: 1

      Sure, I know two legs don't mean shut down all control. But you need a lot less than a single wheel, which is always off balance, whereas legs can more or less keep corrections to a minimum.

  62. Created in his Master's image by Anonymous Coward · · Score: 0

    Somebody open a PayPal contribution fund -- this poor little robot's only got one ball.

  63. Fudd's Law by ec_hack · · Score: 1
    They are also much more agile, since they can be pushed out of the way easily without falling over.

    Don't they teach elementary physics anymore?

    "If you push something hard enough, it will fall over!" - Fudd's First Law of Opposition.

    Brought to you by the Department of Redundancy Department.

    What did I do with that boxcar of industrial coke?
  64. Giant ball crushes people? by bugg · · Score: 4, Interesting
    Ah, Carnegie Mellon. My alma matter.

    Knowing the folks at the robotics institute it'll be about 2 years before we see the army testing out a prototype robot that balances on a ball .... and then uses the ball to crush infantry while launching missiles.

    CMU has a lot of great projects like that. Gladiator, Crusher... I know that most individual robot builders mean well, and I have friends who are anti-war at the RI, but seriously, look at who is actually demanding this technology and where it's being used. It's not being used to house the poor.

    --
    -bugg
  65. Yes by ishmalius · · Score: 1

    Here is a paper about it.

  66. Do you have to run it on a robotpad? by Anonymous Coward · · Score: 0

    What happens when it gets clogged with rubber bits and lint?

  67. Krang by Anonymous Coward · · Score: 0

    Didn't they have robots like this on the Technodrome?

  68. Only a robot... by sm62704 · · Score: 1

    ...can stand on its own ball. Ouch.

    --
    mcgrew's razor: Never attribute to stupidity that which can be explained by greedy self-interest
  69. Re:Not So Strange by Paolone · · Score: 1

    If you really strive for it, even ones with legs.

  70. CARRYING ALONG THE MOUSE ANALOGY... by way2trivial · · Score: 2, Funny

    we'll make it optical, that'll solve the problem!

    --
    every day http://en.wikipedia.org/wiki/Special:Random
  71. This will be awesome by Anonymous Coward · · Score: 0

    ... when the robots get taller and on friday nights, when really drunk, we can go ball-bot tippin' instead of hurting those poor cows.

  72. arrrr.. by zogger · · Score: 1

    ..t'would be to the left of the forrard motion of travel, matey!

    I just thought the image was amusing in me mind, a wad of gum, zillion buck rolling robot..buncha guys in white coats standing around scratching their heads..heh heh heh. I think that's why we as society thought it might be a good idea to have BOTH scientists and engineers.

  73. Sticking to the Floor by Doc+Ruby · · Score: 1

    I remember seeing a DSP demo in probably 1992 where two upright sticks were joined in the middle by a pair of hinges, one allowing the "z axis" stick to bend in half in the xz plane, the other to bend in half along the yz plane. The DSP got feedback from a mercury switch in each of the x and y directions, indicating whether the top stick was balanced in that direction, and opening/closing the respective hinge to compensate for being off-balance.

    That stick just stood up straight on its end, wobbling ever so slightly. Slap it and it would bend like a knee, until it stood up again. Simple, effective, and it looks magic.

    I've been waiting to see some kind of home robot that does that effect. Now this robot does it with a ball. I'd prefer to see the much cooler stick hopping around my house. It might have an easier time getting across the debris that clutters my floor. I know I do.

    --

    --
    make install -not war

  74. The poor are just like the rest of us by snowwrestler · · Score: 1

    I wouldn't want to live inside a giant robot. Why would the poor feel any differently?

    --
    Build a man a fire, he's warm for one night. Set him on fire, and he's warm for the rest of his life.
  75. segway ip by prurientknave · · Score: 1

    I have no idea how segway plans to defend their ip about balancing on two wheels, i.e. if they have any. People have been doing two wheel balancing robots for a long time. I remember seeing a few two wheeled robots in a micromouse maze solving robot competition in the 90's. They were the fastest speed run robots around. Granted it wasn't designed to balance a human but the principle is the same. The benefit of these robots is obviously the turning radius and the speed of the turns themselves. Ball robots aren't that big a stretch but the problem has always been how to deal with gunk getting stuck to the robot, when the robot only has one wheel. On two wheel robots you can always readjust wheel spin rates to compensate for a damaged wheel.

  76. Nothing new here.. by DaMouse404 · · Score: 1

    Clowns have been doing this for years.. -DaMouse

  77. Re:GNAA unveils "New for Jew Thousand and Six" pro by fm6 · · Score: 1
    Excellent hosting provider: DreamHost [dreamhost.com]
    First, forget about spamming Slashdot with affiliate links. No money in it. I once managed to get my Amazon affiliate link into the main story — and made about $10 on the deal.

    Second, Dreamhost is at the point that every small operation comes to where they discover that the business plan that worked fine when they had 10,000 customers doesn't scale up to 100,000 customers. Some companies can adapt to success, but Dreamhost ain't one of them. Slow servers, too much down time, no emergency plan. I'm bailing as soon as my current contract expires.

    Third, the GNAA gag was never funny, and is now definitely old.

  78. video of the most advanced robots on earth by 80+85+83+83+89+33 · · Score: 1

    i wonder how long till there is a BallBot league in robocup!

    the middle size league in the robocup competition has THE most complex robots there are. this is where the latest advances in robots combine.

    here is a video of the top two teams from the Bremen round this year.

    http://www.shadow-dragon.info/upload/robocub_final e_tribots.wmv

    i know that not many will get to see this post since i'm posting two days after the ballbot story was posted... but the video is very revealing as to the current state of robotics.

    --
    i disable sigs
  79. Can you ride it? by shish · · Score: 1

    This looks like it could make a great mode of transport...

    --
    I mod down anyone who says "I will be modded down for this", regardless of the rest of their comment
  80. Down, down, down... by hicksw · · Score: 1

    is fine,if a little noisy.

    Let's see that ball bearing robot go UP a flight of stairs.

    And using the elevator/lift doesn't count.

  81. Re:GNAA unveils "New for Jew Thousand and Six" pro by Jerk+City+Troll · · Score: 0, Offtopic

    Thank you for the advice! I will think of something else to link to then, perhaps the GNAA website? Anti-Slash? I am open to suggestions! Also, how can you say that GNAA is never funny? I know I laugh out loud almost every time I see one of these things pop up. Remember, humor is very subjective.

  82. Oh Great. . . by AcidLacedPenguiN · · Score: 1

    . . . our only refuge from the impending robotic uprising was to tilt those vending machine sized robots onto their backs, and now some american scientists have decided to go and ruin that for us. I suppose we better lay down and just accept our new dynamically upright robotic overlords.

    --
    disclaimer: I've been known to store numbers in my ass for which to dig out when quantities are required.
  83. Meanwhile, over at the MIT robotic labs... by rholland356 · · Score: 1

    Let Carnegie Mellon bring on its steel ball robot. MIT will counter it with their Wheel Chock Robot. Meanwhile, Penn State robotics department has in development the magnetic iron filings robot.

    And Arizona State is working on the unlevel tile floor robot, which has two modes of attack--the missing tiles mode and the wide, deep grout lines mode.

  84. TEST: by Doc+Ruby · · Score: 1

    TEST:

    --

    --
    make install -not war