Slashdot Mirror

← Back to Stories (view on slashdot.org)

Kilobots — Cheap Swarm Robots Out of Harvard

Posted by Soulskill on from the keep-wil-wheaton-away-from-these dept.

An anonymous reader writes with news of a research project at Harvard into controlling large swarms of small robots. This article describes what they call Kilobots. (Which, for clarity's sake, have nothing to do with killing. Yet.) Quoting: "They're fairly simple little robots about the size of a quarter that can move around on vibrating legs, blink their lights, and communicate with each other. On an individual basis, this isn't particularly impressive, but Kilobots aren't designed to be used on an individual basis. Costing a mere $14 each and buildable in about five minutes, you don't just get yourself one single Kilobot. Or ten. Or a hundred. They're designed to swarm in the thousands."

88 of 121 comments (clear)

  1. Welcome! by DrSlinky · · Score: 2

    I'd like to be the first to welcome out new robot overlords.

    If there's anything I can do to make this transition easier on you, you need but ask. Oh, and that neighbor I don't like is part of the resistance.

    1. Re:Welcome! by ColdWetDog · · Score: 2

      I'd like to be the first to welcome out new robot overlords.

      If there's anything I can do to make this transition easier on you, you need but ask. Oh, and that neighbor I don't like is part of the resistance.

      I'll bring the broom and dustpan.

      --
      Faster! Faster! Faster would be better!
    2. Re:Welcome! by shutdown+-p+now · · Score: 1

      Ha, you fool. I'd like to be the first to welcome our new grey goo overlord!

  2. Oblig. XKCD by Anonymous Coward · · Score: 4, Funny

    Obligatory XKCD

    1. Re:Oblig. XKCD by religious+freak · · Score: 1

      WTF is up with the XKCD hate from some people? It's weird.
      Is there a reason for this, or are people just assholes?

      --
      If you can read this... 01110101 01110010 00100000 01100001 00100000 01100111 01100101 01100101 01101011
    2. Re:Oblig. XKCD by drolli · · Score: 1

      Also see: Mantrid Drones (The Lexx).

    3. Re:Oblig. XKCD by sco08y · · Score: 2

      WTF is up with the XKCD hate from some people? It's weird.

      It's karma-whoring and contributes nothing to the discussion. It's usually the same comics over and over again. The comic itself is smug and pretentious.

    4. Re:Oblig. XKCD by religious+freak · · Score: 1

      +1

      --
      If you can read this... 01110101 01110010 00100000 01100001 00100000 01100111 01100101 01100101 01101011
  3. Ok.. by errandum · · Score: 2

    Haven't braitenberg vehicles kind of simulated this kind of behavior for a while now?

    1. Re:Ok.. by Joce640k · · Score: 1

      Dunno....but that looks like a really fun project to work on. Anybody know where you can buy those for $14?

      --
      No sig today...
    2. Re:Ok.. by RespekMyAthorati · · Score: 1

      They could probably be mass-manufactured in China for a dollar or so.

      I'd buy that for a dollar!

  4. Why? by Cedarbridge · · Score: 2

    Its a cool idea and all but designed for the thousands? What on Earth would I want a thousand little vibrating bots that jostle around in circles and blink at eachother for?

    1. Re:Why? by Matheus · · Score: 1

      Call me old fashioned, but I refuse to call these a swarm until they can fly. :)

    2. Re:Why? by SomePgmr · · Score: 4, Insightful

      Good question. Also, you'd have to find a well-cleaned aircraft hangar or something with enough perfectly flat, obstacle-free space for them to do anything. I don't think the idea is really about making these particular ones practical... but more about programming a whole swarm all at once and having them ready to go off and do whatever they're supposed to on their own. And managing that in a way that's cheap and effective. The article says they're looking to get up to a thousand of them and have them work out "self healing" and "collective trasport". These taks have been done individually, but baby steps towards a more impressive whole, I guess.

    3. Re:Why? by Anonymous Coward · · Score: 2, Funny

      My wife could think of a few good reasons..

    4. Re:Why? by InsectOverlord · · Score: 1
      To keep cockroaches off.

      And if you step on one of them by accident, you might kill it but the result won't be nearly as gross.

    5. Re:Why? by martin-boundary · · Score: 1

      What on Earth would I want a thousand little vibrating bots that jostle around in circles and blink at eachother for?

      The warm glow from spending $14,000 to help a future Harvard startup monetize its slashverts? And a free set of steak knives!

    6. Re:Why? by religious+freak · · Score: 1

      Why not?

      --
      If you can read this... 01110101 01110010 00100000 01100001 00100000 01100111 01100101 01100101 01101011
    7. Re:Why? by RespekMyAthorati · · Score: 1

      What I don't get is what they could possibly demonstrate with a thousand of these things that they couldn't demonstrate just as well with 100.

  5. Prey by bp2179 · · Score: 1

    Michael Crichton wrote a book about how we thought we could control thousand of nanobots using swarm theory. I wonder if they are using algorithms derived from bees. As long as we don't develop AI anytime soon, I think we are safe.

    1. Re:Prey by martin-boundary · · Score: 1

      I wonder if they are using algorithms derived from bees.

      Gosh I hope not! There's a documentary with Michael Caine that shows this would be a bad idea.

    2. Re:Prey by Jeremi · · Score: 2

      Michael Crichton wrote a book about how we thought we could control thousand of nanobots using swarm theory.

      Also, L. Ron Hubbard wrote a book about how we are all haunted by thousands of evil alien ghosts. Fortunately, both books are works of fiction, and therefore have little relation to reality.

      --


      I don't care if it's 90,000 hectares. That lake was not my doing.
    3. Re:Prey by PopeRatzo · · Score: 1

      Michael Crichton wrote a book about how we thought we could control thousand of nanobots using swarm theory.

      Michael Crichton was a crypto-fascist crank who was wrong about almost everything.

      Sort of like an even less talented Ayn Rand who read Popular Science.

      --
      You are welcome on my lawn.
    4. Re:Prey by TapeCutter · · Score: 2

      Michael Crichton wrote a book

      Let me guess, it's about "science gone mad", right?

      --
      And did you exchange a walk on part in the war for a lead role in a cage? - Pink Floyd.
    5. Re:Prey by Hal_Porter · · Score: 2

      Please Cease and Desist from violating the Centre for Religious Technology's Intellectual Property.

      --
      echo -e 'global _start\n _start:\n mov eax, 2\n int 80h\n jmp _start' > a.asm; nasm a.asm -f elf; ld a.o -o a;
    6. Re:Prey by Hal_Porter · · Score: 2

      Still pissed about this, eh?

      --
      echo -e 'global _start\n _start:\n mov eax, 2\n int 80h\n jmp _start' > a.asm; nasm a.asm -f elf; ld a.o -o a;
  6. Re:Kilobot by Intrepid+imaginaut · · Score: 2

    Indeed, they could have chosen a less ominous name.

    Like fluffyboppers.

  7. We will lose. by lw54 · · Score: 1

    You could kill a robot. Maybe even 10. But, in the thousands, even if they are tiny and weak, you will lose.

    1. Re:We will lose. by artor3 · · Score: 1

      I'm pretty sure a decent pair of boots is all you'd need to defeat a swarm of these things.

    2. Re:We will lose. by Kjella · · Score: 1

      No. The only reason it's a remotely good science fiction toy is because the robots self-replicate or assimilate - usually by magictech. Built what's essentially a machine gun turret on wheels and it'd be much more effective than a million toy soldiers.

      --
      Live today, because you never know what tomorrow brings
    3. Re:We will lose. by davester666 · · Score: 1

      ...except for the few thousand that manage to climb up the side of your boots and then burrow inside your legs...

      --
      Sleep your way to a whiter smile...date a dentist!
    4. Re:We will lose. by SomePgmr · · Score: 1

      Well the article says one of the next steps is "self healing", whatever that means. I'd guess that just means bringing in new bots to replace dead ones. But hey, it's something. :)

    5. Re:We will lose. by artor3 · · Score: 1

      They can't climb or burrow. They have all the mobility of my phone in vibrate mode.

    6. Re:We will lose. by Jeremi · · Score: 2

      They can't climb or burrow. They have all the mobility of my phone in vibrate mode.

      Well sure... These are just the research prototypes. v2.0 will add spring-loaded legs, and teeth.

      --


      I don't care if it's 90,000 hectares. That lake was not my doing.
    7. Re:We will lose. by shutdown+-p+now · · Score: 1

      Imagine being buried in thousands of phones, which all vibrate in sync.

    8. Re:We will lose. by delirium28 · · Score: 1

      Yeah, so did the Daleks, but that didn't stop them!

      --
      Who is John Galt?
    9. Re:We will lose. by NoNonAlphaCharsHere · · Score: 1

      And even before they all link together and form a MegaZord that crushes cars and buildings underfoot...

    10. Re:We will lose. by elrous0 · · Score: 1

      Even if I send wave-after-wave of my men at them?

      --
      SJW: Someone who has run out of real oppression, and has to fake it.
  8. D'awww by Windwraith · · Score: 2

    They are so cute!
    I don't even know why but I was "awww"ing when all the robots started to blink their lights in unison.

    1. Re:D'awww by Shemmie · · Score: 1

      Totally opposite reaction to me. I got the creeps when I saw the little imps talking to each other to synchronize their flashing lights.

      It was like watching the Borg boot up!

    2. Re:D'awww by Windwraith · · Score: 2

      Really? That was one of my favorite parts...I can imagine them making cute beeps while flocking around.
      Then again I happen to find most "real" robots cute. with those huge shiny camera eyes and unreliable movement. It's like watching metal puppies tripping while taking their first steps.
      I'd be willing to get myself a dozen and have it move around my desk all day. What would they do if they go berserk anyway? Vibrate all over me until I die of old age?

    3. Re:D'awww by SanityInAnarchy · · Score: 1

      Second obligatory XKCD.

      --
      Don't thank God, thank a doctor!
    4. Re:D'awww by elrous0 · · Score: 1

      It's cute until you realize they're blinking "kill" in Morse.

      --
      SJW: Someone who has run out of real oppression, and has to fake it.
  9. Vancouver? by drwho · · Score: 1, Offtopic

    A swarm of creatures with low intelligence? You mean, that Harvard manufactured the Canuks rioters/fans?

    1. Re:Vancouver? by guruevi · · Score: 1

      At least those Canuks fans get something done (even though it's destruction/economic stimulus). I believe they're more like a group of IT managers at a Microsoft convention - they can all start communicating with the latest buzzwords but in the end it doesn't mean anything to anyone observing and trying to get an intelligent answer.

      --
      Custom electronics and digital signage for your business: www.evcircuits.com
  10. Batteries by BradleyUffner · · Score: 2, Funny

    The battery life is only 3 hours, and is non rechargeable.. I'd REALLY hate being the intern at a company using a swarm of 1000s of these guys after the first experiment.

    1. Re:Batteries by AmberBlackCat · · Score: 1

      You fix that problem with another robot...

    2. Re:Batteries by shutdown+-p+now · · Score: 1

      See, that's why they'll have to start using humans for power eventually. You've been warned.

  11. Re:Kilobot by tripleevenfall · · Score: 1

    I'm not too worried, as my gigantic robotic stamping boot is in alpha testing

  12. Not what I'd like to see come out of Harvard... by RobinEggs · · Score: 1

    I'm afraid I misread the headline as something much, much cooler.

    Did anyone else read that Harvard had invented a "cheap swarm of killbots?"

    Because that would be awesome.

    1. Re:Not what I'd like to see come out of Harvard... by Mogusha · · Score: 1

      I'm surprised noone has said it yet, but do they have a predefined kill limit? So, if they get out of hand we can send wave after wave of men at them?

    2. Re:Not what I'd like to see come out of Harvard... by Dahamma · · Score: 1

      Yep. Once I realized what it really said I was pretty bummed.

      Ah, but we can always build more killbots!

  13. When will this lead to something useful? by c0d3g33k · · Score: 2

    I understand that small steps (no pun intended) need to be taken to advance the state of the art, but this remains an academic novelty until these little guys can do something useful. Doesn't have to be terribly complicated. There are plenty of simple yet highly repetitious and tedious tasks that would be perfect for a cooperating swarm of little worker bots working in parallel. Like carrying the leaves off of the lawn and depositing them in the woods (or a recycling bag) in the fall or similar. Then I would be impressed (and would be the first in line to buy the kit).

    1. Re:When will this lead to something useful? by 1+a+bee · · Score: 3, Informative

      For you and I further down the food chain, it'll probably be a while. For researchers, though, it's arguably already useful. FTA:

      Generally people who want to experiment with large swarms have had to be content with computer simulations, which is fine, but at some point you have to try things out in the real world (or as close as you can get in a lab), and Kilobots can make that happen. .. at $14 each, a thousand robots is actually an achievable number with a modest grant, which is something that probably has not been possible before.

    2. Re:When will this lead to something useful? by c0d3g33k · · Score: 2

      Agreed. I made a weak attempt to acknowledge this in my opening sentence, but in my haste to get to my main point, I was overly dismissive without meaning to be. Moving beyond simulations and into the real world is definitely a necessary and laudable achievement.

    3. Re:When will this lead to something useful? by RandCraw · · Score: 1

      Exactly. Clean out my gutters. Paint my house. Weed my garden. Stop just writing papers.

    4. Re:When will this lead to something useful? by innerweb · · Score: 1

      From a military or industrial perspective, this is useful. This is one step away from a self-organizing mine field, or sensor array. By allowing them to choose the proper *scatter*, you can very easily release a set of bots in the field to do anything from look for oil in the water to finding illegal aliens crossing the border. True, these little guys are not going to be the used solution, but they are providing the conceptual testing that will lead to more expensive and larger machines doing real work. Once you have the proven behaviors, it is a a simpler matter to add them to other machines to give the machines we currently use the same behaviors. I don't see them doing lawn work anytime soon, but in the military and industrial world, there are many applications where these behaviors would be ideal.

      --
      Freud might say that Intelligent Design is religion's ID.
  14. The first robot war begins in 5...4...3... by zill · · Score: 2

    How long until they divide into two camps and fight to the death over whether kilo = 1000 or 1024?

    1. Re:The first robot war begins in 5...4...3... by jabberw0k · · Score: 1

      The Kilobots and the Kibibots? (Cue Prokofiev's "Montagues and Capulets")

    2. Re:The first robot war begins in 5...4...3... by mysidia · · Score: 1

      Maybe so... The string Kibi is really close to Kibo. Perhaps the folks on alt.religion.kibology could have some insight into what the strategy of Kibo's kibibots would be :-)

  15. The size of a quarter by safetyinnumbers · · Score: 1

    That's about the size of 15 pence, for readers in the UK.

  16. But the're incapable of folding my fitted sheets.. by RNLockwood · · Score: 1

    and I think they won't do so well on my carpets. But how about swarm wars? THAT'S cool!

    --
    Nate
  17. What we have is a new measure of automation by mysidia · · Score: 5, Insightful

    1024 kilobots = 1 Megabot

    1024 Megabots = 1 Gigabot (aka 1 Decepticon)

    1024 Gigabots = 1 Terabot

    1024 Terabots = 1 Petabot (A sufficient number of bots to enslave humanity)

    1024 Petabots = 1 Exabot (A sufficient number of bots to enslave the the planets in our solar system)

    1024 Exabots = 1 Zettabot (A sufficient number of bots to enslave our galaxy)

    1024 Zettabots = 1 Yottabot (A sufficient number of bots to enslave 25% of the known universe)

    1024 Yottabots = A sufficient number of bots bots to replace all interesting objects in the known (and unknown) parts of the universe with Kilobot swarms.

    1. Re:What we have is a new measure of automation by tebee · · Score: 1

      So if each kilobot takes 1 day to make a duplicate of itself - we have 90 days until we get to the "A sufficient number of bots bots to replace all interesting objects in the known (and unknown) parts of the universe with Kilobot swarms" stage?

      --
      N.B. this user is far too lazy to write a witty and intelligent sig.
    2. Re:What we have is a new measure of automation by mysidia · · Score: 1

      we have 90 days until we get to the "A sufficient number of bots bots to replace all interesting objects in the known (and unknown) parts of the universe with Kilobot swarms" stage?

      Unless they are extremely versatile in their reproduction and operation, they run out of essential easily-obtainable resources on earth to harvest for rapid reproduction long before 90 days.

      And have to start either mining resources themselves or utilizing enslaved species to do work to extract resources for Kilobot reproduction :)

    3. Re:What we have is a new measure of automation by charlieo88 · · Score: 1

      When you open the box and count out only 1,000 ONLY THEN do see the little asterisk where the manufacture says 1,000 kilobots to the megabot because for marketing purposes they use SI prefixes.

    4. Re:What we have is a new measure of automation by 91degrees · · Score: 1

      Yes, except Kilo means 1000. Not 1024. RAM manufacturers use an approximation, but the official IEC term is .

    5. Re:What we have is a new measure of automation by mysidia · · Score: 1

      Yep, but 24 out of 1000 is less than 2.5%, and 24^8 out of 1000^8 is less than 0.00000000001%, so you keep the box anyways :)

  18. Re:Killbots by lostfayth · · Score: 1

    The editors were rather prescient then.

  19. My Kilobot by stinerman · · Score: 1

    My Kilobot has Lotus Notes and a machine gun. It is the finest available.

  20. "nothing to do with killing" by Arancaytar · · Score: 1

    So they're working on robot swarms that will totally not try to wipe out humanity?

    Even as a pun, that's a Suspiciously Specific Denial.

  21. Re:Killbots by nurb432 · · Score: 1

    Why yes, yes you were the only one.

    --
    ---- Booth was a patriot ----
  22. Re:What we have is a new measure of automatons by 1+a+bee · · Score: 1

    Pew! And I was worried about kilobots turning on us. It's them yottabots we should be worried about. Oh, wait a minute.. I think they've so throughly won we call their collective reality.

  23. neato... by Thraxy · · Score: 1

    I was very impressed by that synchronized blinking action there. That will come in handy... Endless possibilities there, my friend... endless

    1. Re:neato... by hazem · · Score: 1

      It's very similar to a "Firefly" simulation in Netlogo (http://ccl.northwestern.edu/netlogo/). It's an agent-based simulation language/system that allows you to explore similar behaviors as the ones seen here, but virtually.

  24. OMG RepRap is the new skynet by cultiv8 · · Score: 1

    Only now you have to buy parts off ebay

    --
    sysadmins and parents of newborns get the same amount of sleep.
  25. Calling CleverNickname... by RyuuzakiTetsuya · · Score: 1

    ...please pick up the shaped like a galaxy class starship courtesy telephone please...

    --
    Non impediti ratione cogitationus.
    1. Re:Calling CleverNickname... by MichaelSmith · · Score: 1

      Will hasn't posted here since 2009 but you could try reddit...

      --
      http://michaelsmith.id.au
  26. Killbots by hoborg1 · · Score: 1

    I misread and was hoping for cheap swarms of killbots...

  27. Been seeing them for years by SuperKendall · · Score: 1

    Kilobots â" Cheap Swarm Robots Out of Harvard

    We call them politicians. They come straight of of Harvard with hair gel and sense of entitlement already applied.

    --
    "There is more worth loving than we have strength to love." - Brian Jay Stanley
  28. Ian McDonald detailed what these could become by Basilius · · Score: 1

    He called them "BitBots" and they feature prominently in his recent book "The Dervish House."

    I want some.

  29. Legion - Geth by AleXelA · · Score: 1

    http://masseffect.wikia.com/wiki/Legion

    Legion is a unique geth mobile platform, designed to operate outside the Perseus Veil and interact with organics directly. To that purpose, it houses 1,183 geth programs, as opposed to the one hundred in other platforms, enabling it to operate independently and speak....
    http://www.youtube.com/watch?v=rMPeG3pQlPw&NR=1 ...If questioned about specifically using Shepard's N7 armour to repair itself, Legion becomes evasive, first rationalizing with "there was a hole" and then states "no data available" after being pressed, suggesting that it was the result of an irrational action, which goes against the concept that every action the geth take is the result of calculation, unaffected by emotion.
    http://www.youtube.com/watch?v=wF7KYQ1_jKo

    Waking up the Geth: (there is a new faction inside the Geth (that break apart from the earlier less evolutioned ones), that builds their own future, no one will be armed ... unless they involve themselves...)
    http://www.youtube.com/watch?v=VMzBSRityaU&feature=related

    Geth enters the Quarian ship: (Geth meet their creators, creators meet their killers, the Geth keep the homeworld and the Quarians had to scape and live exiled)
    http://www.youtube.com/watch?v=QDemlIaPv18&feature=related

    Of course the Geth views's of the whole story is just the opposite, they were just defending themselves .... which may not be false.

    Did you even know you could get in love with them?
    http://www.youtube.com/watch?v=2IdIhFXcRl8

    I love Legion's mind machine calculations all the time:
    http://www.youtube.com/watch?v=UJI66t-B83g&feature=related

  30. This is not swarm behavior at all by popo · · Score: 1

    Swarm behavior requires a mutual awareness between devices. This is simply a remote control which affects a large number of units. The problem with this approach is that it only works initially, but random differences in movement become magnified over time. Since each bot movement includes a random margin of error (think: drunkard's walk) the "swarm" will dissipate over time and show less cohesion. In true swarm behavior, each individual actor (bot, in this case) is aware of the greater swarm.

    --
    ------ The best brain training is now totally free : )
    1. Re:This is not swarm behavior at all by stevelinton · · Score: 1

      The article says they can communicate with one another.

  31. Kilobots, Harvard, okay, buzzwords, we got it, by arisvega · · Score: 1

    .. but what do they DO ?

    --
    The three laws of thermodynamics:(1) You can't win. (2) You can't break even. (3) You can't even quit.
  32. Re:Replicators??? by IonOtter · · Score: 1

    You aren't the only one, I assure you.

    --
    [End Of Line]
  33. what is the purpose? by hesaigo999ca · · Score: 1

    What is the purpose of these bots other then just run around and talk amongst themselves???

  34. if (Kilobots == Zerglings) by optymizer · · Score: 1

    My life for Aiur!