Slashdot Mirror

← Back to Stories (view on slashdot.org)

Cockroach-Like Robot to Help Explain Animal Movement

Posted by CowboyNeal on from the on-all-sixes dept.

neutron_p writes "A cockroach-like robot named RHex is the starting point for a major project to understand animals' most distinguishing trait: how they move without falling over. Researchers from several universities will focus on RHex, a short, six-legged robot that scampers like a cockroach, as a working model of the principles they're seeking to uncover. By tweaking the robot and using it as a physical model, they hope to tease apart the complex neural and muscular networks in insects."

49 of 124 comments (clear)

  1. Easy... by Anonymous Coward · · Score: 2, Funny

    They don't fall over because they aren't drunk.

    1. Re:Easy... by Anonymous Coward · · Score: 5, Informative

      Already been done!

      The best way to study their movement is to mount railguns on them, and fire them at random, confusing the hell out of the little guys...don't believe me?

      http://jeb.biologists.org/cgi/content/full/205/18/ 1803/i

      Check out the picture!

    2. Re:Easy... by ZeroPost · · Score: 3, Funny

      What they should really study are the effects of falling and intentionally missing, thereby attaining a state of midair suspension.

  2. How long by Dorsai65 · · Score: 5, Funny

    before one of the humanoid robots tries to squish it?

    --
    --- Asking inconvenient questions for over 30 years...
    1. Re:How long by commodoresloat · · Score: 2, Funny

      These will be the only robots to survive a nuclear war.

  3. duh... by niteice · · Score: 3, Insightful

    They don't fall over because they (usually) aren't missing a leg where one is needed for proper balance.

    --
    ROMANES EUNT DOMUS
  4. I work on this... by feelyoda · · Score: 4, Insightful

    on the perception side...

    While the ability of the bot to go over hard terrain is amazing, the point is that your relinquish direct control.

    The basic problem in perception is dealing with the drastic motions.

    The computer vision methods needed are quite complicated, requiring complimenting sensors like inertial measurement devices. Also extremely wide-angle cameras are excellent because things stay in view, but difficult because the pin-hole model fails.

    Go here for some work that is now a bit dated, from a 180degree camera strapped to rhex:
    http://www.frc.ri.cmu.edu/projects/buzzard/ rhex/

    --

    Robo-Blogs of the world: UNITE!
    1. Re:I work on this... by feelyoda · · Score: 4, Informative

      also, you can get more info here:
      http://www.rhex.net/

      look for the great video of the tumble from a pile of boulders, which doesn't seem to be a problem.

      I wish I could see ASIMO take a fall like that...and watch the subsequent execution of the grad student who let it happen.

      --

      Robo-Blogs of the world: UNITE!
    2. Re:I work on this... by plebius · · Score: 3, Interesting

      Could it be possible that a computational approach isn't necessarily the way to go?

      See, for example, the work of Mark Tilden

      http://www.exhibitresearch.com/tilden/

      --
      Psychology News
    3. Re:I work on this... by feelyoda · · Score: 4, Insightful

      "Could it be possible that a computational approach isn't necessarily the way to go?"

      As Brooks showed, you can get very complicated behavior from reactive or even semi-stateful robots. BUT, I would question the scalability to something more application specific and useful.

      For instance, imagine such a bot making a sandwich, and then cleaning your toilette...

      Also, as some point, you're going to want to give the bot an order, like go from A to B to C then back. Rhex would be unable to do that currently without a very engineered environment, which goes against the entire point of such a bot which moves skillfully in all environments.

      Adding a robust perception loop around the sense-response robot is the way to go, as far as I'm concerned.

      --

      Robo-Blogs of the world: UNITE!
    4. Re:I work on this... by jwgoerlich · · Score: 5, Interesting

      "Adding a robust perception loop around the sense-response robot is the way to go, as far as I'm concerned."

      Agreed. In fact, that was one avenue the BEAM folks and Mark Tilden began exploring. Their take was to develop a solid sense-response sub-layer, and then layer on the computing systems.

      The BEAM name for the architecture was Horse-Rider.

      J Wolfgang Goerlich

    5. Re:I work on this... by celeritas_2 · · Score: 2, Interesting

      I've read something [in Wired probably] about movie makers using evolutionary software to make their creatures walk without having to hard code every movement [think LOTR: kings, the whole orc army was digital for a big chunk of the fight] I for one am not suprised that evolution came up with balance, just a feedback system of a motion dector and legs.

      --
      -- Checking emails and kicking cheats `till the day I die.
    6. Re:I work on this... by Danny+Rathjens · · Score: 2, Informative
      The basic problem in perception is dealing with the drastic motions.

      I saw a show on animal channel the other day that was about the fastest runners. Number one was the tiger beetle. What struck me is that the reason it runs in short bursts is that its perception system can't keep up with all the input. So it has to keep stopping to get its bearings. Roaches are very fast too, and they use this same method of short bursts and stops. (which has the added benefit of making them harder to stomp. :)

      Another example of this inability to perceive too much movement input is the funky neck movements many ducks make while moving to hold their eyes still.

    7. Re:I work on this... by rawket.scientist · · Score: 3, Funny

      For instance, imagine such a bot making a sandwich, and then cleaning your toilette...

      And be sure to imagine it completing those actions in exactly that order. No one's been able to teach good kitchen hygiene to cockroaches yet, be they 'bots or bugs.

      --
      John Hancock wuz here.
    8. Re:I work on this... by BiggerIsBetter · · Score: 2, Interesting

      Why does it have to be either/or?

      Part of repatition training for humans is to transform a computational movement into a reaction movement. If you repeat something enough, it eventually gets hardwired (my lower brain is an FPGA!?!) and you no longer have to think/compute about it.

      I think a "tri-layer" approach is good, with a sense-response layer (step back - ouch, that's hot) deferring to a trained-response layer (walking) then to a computed-response layer (walking over a rock garden). And if you walk over a rock garden often enough, the evolved and successful patterns used get transferred to the trained layer to make it "second nature".

      But I'm just pulling this out of my ass, you'd need to find a hot robot expert to 2nd my opinion...

      --
      Forget thrust, drag, lift and weight. Airplanes fly because of money.
    9. Re:I work on this... by feelyoda · · Score: 2, Informative

      actually roaches are probably cleaner than you...

      most poisons work because they touch them, and then clearn their feet constantly...

      that said, they do carry disease

      --

      Robo-Blogs of the world: UNITE!
  5. Practical Applications by ScArE2100 · · Score: 5, Interesting

    This could have immediate application for the disabled. Imagine a personal moving device like the segway that walks around with technology derived from insects. It'd be pretty cool I'd say.

    Maybe a mars rover that doesn't faller over or get stuck

    There are lots of possible uses of data from this research.

  6. This is good news by Anonymous Coward · · Score: 5, Funny

    ... for real cockroaches. When Armageddon comes, the cockroaches will have robot versions of themselves for slaves.

    1. Re:This is good news by mblase · · Score: 2

      When Armageddon comes, the cockroaches will have robot versions of themselves for slaves.

      I, for one, welcome our new robotic cockroach overlords.

  7. $30M for more insect robots? Sounds like pork. by Animats · · Score: 4, Insightful
    There's way too much work on insect-level locomotion. Brooks was doing this stuff twenty years ago, and took it about as far as it's going to go. Reactive systems reached their ceiling years ago.

    We realy should be doing better than this. We should at least have Aibo-type robots running (or at least trotting) over real terrain by now. It's embarassing.

    The trouble with this insect stuff is that you can do crap work and get published. If you do work on robots that really balance, you look stupid if your control system doesn't work. Everyone can see you failed. With insect robots, failure is less obvious. Some people think this is a feature.

    1. Re:$30M for more insect robots? Sounds like pork. by gl4ss · · Score: 2, Insightful

      cockroaches are basically reactive animals.

      and uh, what is the "we should have.." based on? wishes? according to scifi we should have flying cars, that however doesn't make them technically feasible(or possible at right price) at the moment.

      --
      world was created 5 seconds before this post as it is.
    2. Re:$30M for more insect robots? Sounds like pork. by Simonetta · · Score: 2, Interesting

      There's way too much work on insect-level locomotion.

      First spend tens of million studying insect motion. Then spend hundreds of millions researching motion of higher lifeforms. Then billions to develop a factory manufacturing system to make copies of moving animals.

      Why?

      Every year we generate many millions of the most perfect and adaptive biological being the world have ever seen... babies...humans. Yet most of them get nothing but shit and are doomed to live on a dollar a day for their entire lives.

      Why spend billions to create synthetic robotic psuedo lifeforms when the actual humans themselves are so absurdly cheap, Cost next to nothing, and are self-replicating? Give 'em burgers, Allah, and heroin; they'll do whatever you need...you don't need to spend billions creating robots. Not today, not in the 21st century, not on earth.

    3. Re:$30M for more insect robots? Sounds like pork. by Indras · · Score: 2, Insightful

      Brooks was doing this stuff twenty years ago, and took it about as far as it's going to go.

      Yes, how insightful. Everything that could ever be discovered in this particular field has been. It's time to close the books on this forever. Every time I hear this kind of comment, it sets off my "idiot" alarm.

      My favorite was from my professor in Applied Electricity. We were in the second week, going over Ohm's Law, I asked how the equations worked when the resistance dropped to zero, like in a superconductor. He said that it was impossible to have no resistance in a wire, "super" conductors cannot exist. (He even used his fingers to make quotation marks in the air when he said "super," because he'd never heard the word before). This was two months ago.

      Basically, any time you think that any single topic has been completely explored, and there's nothing left to learn is probably when you're the most wrong. It's the sign of a closed mind, you might as well sell everything, quit your job, and move into a cabin in the mountains, you're definitely not going to contribute any more to society.

      Until I see a robot that is entirely indistinguishable from a real insect, there's still more left to explore. In fact, even then, there's more left to do, like miniaturization, self-replication, and so on ad infinitum.

      --
      The speed of time is one second per second.
    4. Re:$30M for more insect robots? Sounds like pork. by jamesh · · Score: 2, Insightful

      Why spend billions to create synthetic robotic psuedo lifeforms when the actual humans themselves are so absurdly cheap

      Because if you send them to another planet they'll either explode (no atmosphere) or disolve (corrosive atmosphere).

      And if you put them in spacesuits and send them there, they won't necessarily do what you tell them to.

    5. Re:$30M for more insect robots? Sounds like pork. by maxpublic · · Score: 3, Interesting

      you're definitely not going to contribute any more to society.

      Riiight. As if contributing to society is what we should all be striving for, sacrificing self-interest on the altar of saintly altruism. They teach you that crap in school nowadays?

      In any event, our friend here might be 'contributing to society' by pointing out that the $30 million dollars could be better spent elsewhere. Especially if it's $30 million in TAX dollars, in which case I agree with him.

      Max

      --
      My god carries a hammer. Your god died nailed to a tree. Any questions?
    6. Re:$30M for more insect robots? Sounds like pork. by CGP314 · · Score: 2, Informative

      People don't explode in a vacuum


      -Colin

  8. Yes but... by Anonymous Coward · · Score: 2, Funny

    How long does it live if you lop off its head?

  9. Forget moving like a cockroach... by MooseByte · · Score: 2, Funny


    I want to see robots that *survive* like a cockroach.

    Well, until they turn evil anyway.

  10. Hannibal and Attila? by BrewerDude · · Score: 5, Interesting

    Does anyone know how this differs from the insect-like robots (like Hannibal and Attilla) developed by Rod Brooks' group in the MIT AI Lab? It's been a while since I took his class, but I remember that they found that remarkably simple distributed control systems could be used to generate adaptive legged locomation patterns without requiring complex centralized control.

  11. RHex Web Page by bluewee · · Score: 2, Informative

    RHex Page

    --
    [blue] - The Ministry of Information approved this message...
  12. Rodney Brooks said it best by lawpoop · · Score: 4, Informative
    When an ant walks, it's falling over all the time. The trick is that it catches itself with its legs.

    You can get a load of his work from the documentary Fast, Cheap, and Out of Control.

    --
    Computers are useless. They can only give you answers.
    -- Pablo Picasso
  13. Huh? by bobobobo · · Score: 4, Funny
    how they move without falling over.

    Because they have six legs? Am I missing something here?

  14. RHex is also cool because it swims by exp(pi*sqrt(163)) · · Score: 4, Interesting

    I discovered this by reading the story about RHex here.

    --
    Doesn't it make you feel good to know that our freedoms are protected by politicans, lawyers and journalists.
  15. Biggest application: NASA by mblase · · Score: 4, Insightful

    There's way too much work on insect-level locomotion. Brooks was doing this stuff twenty years ago, and took it about as far as it's going to go.

    I don't think that's true. There was an article in Discover a month or two ago (can't find it online, sorry, but I believe it parallels the linked article) where a researcher was trying to tease more information out of a cockroach's walk, discovering that it doesn't actually use a three-feet-down-all-the-time approach but wobbles side to side, remaining dynamically stable as it walks. This is not what you might intuit by simply watching insects walk.

    As for "too much" being done, I must disagree. Walking robots aren't as good as they can be or it'd be perfected by now. Wheels are faster, but only over ideal terrain; complicated terrain that would confound the best wheels can often be navigated by legged animals. NASA's interplanetary rovers all use wheels, and all of them eventually encounter situations where they're useless, so if they could deploy a robot lander that could walk effectively (and efficiently), it'd be of tremendous value to them.

    1. Re:Biggest application: NASA by Dachannien · · Score: 4, Interesting

      A lot of work in the walking robot field is going into dynamic stability - at least in U.S. academia. The various bipedal robots produced by the Japanese corporate world (Asimo, for example) are all statically stable - meaning that the center of mass is always kept directly over the support base. If you watch videos of those robots carefully, you'll see that this is the case.

      Animals don't work like that. In fact, human walking gait is often described as continually falling forward, saved only by the swinging foot meeting the ground before you face-plant. As for insects, some gaits are statically stable simply by virtue of having so many legs, but the info posted by the parent concerning cockroaches using dynamic stability in tripod gait is really interesting.

    2. Re:Biggest application: NASA by awtbfb · · Score: 2, Insightful


      ... robot lander that could walk effectively (and efficiently)...

      Efficiency is the crux of the problem. Legs are incredibly inefficient compared to wheels and until the on-board power problem is handled in an acceptable manner, you won't see a lot of legged robots too far away from an outlet.

      Having said that, this is where RHEX shines. Since the legs spin like wheels they have a real advantage compared to traditional walking robots in power savings. As an aside, the thing also moves pretty fast. Again, an advantage over most walking robots and a lot like a wheel oriented solution.

      Oh yeah, kids and dogs love it too. It was like the pied piper at a demo day earlier this summer.

    3. Re:Biggest application: NASA by bar-agent · · Score: 2, Funny

      I know why the Japanese are designing robots with static stability. See, it all comes down to anime.

      Giant robots are masters of martial arts action. The key to most martial arts is keeping balanced at all times. So to built giant martial arts robots, the Japanese are specializing in static stability.

      --
      i'd hit it so hard, if you pulled me out you'd be the king of britain [bash.org]
  16. More on Bob Full by jwgoerlich · · Score: 3, Informative

    This is really old news. RHex has been around for at least a few years now.

    Bob Full is one of the lead scientists on the RHex project. His biomimetic approach is amazing. See the following link for one of his lectures.

    Robert Full: "Bipedal bugs, galloping ghosts and gripping geckos: BioInspiration for Rapid Running Robots"
    http://www.princeton.edu/WebMedia/lectures/

    J Wolfgang Goerlich

    1. Re:More on Bob Full by MrEd · · Score: 2, Informative
      Definitely old news, the legs on RHex have evolved a long way from the 'bar' style shown on the article photo.

      The new, new, new design is now a semicircular length of rubber-treaded fibreglass, which means they have spring to them. In fact when one leg finally snaps they have to replace all six as the robot depends on them to be balanced in stiffness.

      Using these legs they get some great dynamic stability as shown in 'turbo mode' and other showoff moves plus pronking, etc.


      The coolest are the round legs with the adjustable radius - stand up, sit down, roll, jump...

      --

      Wah!

  17. Re:Cockroaches and robots by XaviorPenguin · · Score: 2, Interesting

    "
    Is it because if they use cockroaches not many people will yell at them for animal cruelty, because cockroaches are "evil"?
    "

    Why do you think that the show FearFactor can get away with doing the Cockroach bit on national TV?

    My opinion only, if scientists have studied that we can balance ourselves with only 2 feet and while walking, with one foot up and one foot on the ground, why can't they apply that information to insects?

    --
    Friends help you move...
    REAL Friends help you move dead bodies... ^_^
  18. most important feature by kxmas · · Score: 5, Funny

    Will it flip on its back when runs out of batteries?

  19. cockroach to explain animal movement by SimonInOz · · Score: 2, Funny

    Well heck they certainly explain some movement.

    When my wife sees one of those little buggers she runs away - she hates them.

    --
    "Cats like plain crisps"
  20. Not the only people doing this sort of thing by Dachannien · · Score: 4, Informative

    Read about competitive work here.

  21. Not a chance by lmuk · · Score: 3, Funny


    I've seen Robot Wars and the walkers never stand a chance...

  22. Re:Incredible by achurch · · Score: 2, Insightful

    I'm truly amazed, that with all the modern science we have today, that we don't know the answer to this question.

    And we still may not be getting it. All they've built is a robot that coincidentally can also move without falling over--there's nothing (at least as far as I can tell from the article) to say that it works the same way real insects do.

    In all fairness, though, the question "how do animals move" is probably less important than "how can we get robots to move". While learning how the biological systems work can certainly provide insight, we don't have to exactly replicate those systems in mechanical robots, and in fact the optimal movement system for a robot may be different from that for an animal. It's sort of like emulating hardware: if you wanted to you could emulate a CPU down to the logic-gate level, but it's much more efficient to just re-interpret instructions into equivalent operations on the host CPU.

  23. BEAM dead? by LinuxInDallas · · Score: 2

    Whatever happened to BEAM robotics? Anyone involved with that? I first heard about it on a Science Channel robotics series and it seemed interesting. I have been to the main page by the inventor Mark Tilden but I have seen nothing new in the past couple years. Does anyone know if this area of research has died off?

  24. Political Science by Doc+Ruby · · Score: 3, Funny

    We might finally learn where all these politicans come from, by studying cockroaches and other vermin.

    --

    --
    make install -not war

  25. Yes, that is my most distinguishing trait! by bedford · · Score: 2, Funny

    Animals' most distinguishing trait is that they don't fall over? You gotta admit that's a little funny. "Hi, I'm Greg, an animal, and my most distinguishing trait is that I don't fall over".

    Anyway, I just got a kick out of that. I'm sure once I read on, the point will be well made.

  26. Most japanese bipeds use ZMP, not static stability by chr1973 · · Score: 2, Informative

    I know for certain that not all japanese bipeds use static stability. Bipeds from Honda and Sony (and most of them IMHO) use something called zero moment point (ZMP). The japanese biped robot WL-10RD used this as early as in 1984. Here's a reference I just found: ZERO-MOMENT POINT THIRTY FIVE YEARS OF ITS LIFE by Vukobratovic and Borovac. Should be good since Vukobratovic introduced the concept in the 70s. PS. I did my PhD thesis on control and balance of legged locomotion.