Please create an account to participate in the Slashdot moderation system


Forgot your password?
Earth Robotics Hardware

MIT Unveils Oil-Skimming Robot Swarm Prototype 123

destinyland writes "Today MIT reveals a swarm of autonomous floating robots that can digest an oil spill. The 16-foot robots drag a nanowire mesh that acts like a conveyor belt to soak up surface oil 'like paper towels soak up water,' absorbing 20 times its weight and then harmlessly 'digesting' the oil by burning it off. Powered by 21.5 square feet of solar panels, the 'Seaswarm' robots run on the power of a lightbulb, and with just 100 watts 'could potentially clean continuously for weeks' without human intervention, MIT announced. The swarm uses GPS data and communicates wirelessly to move as a coordinated group to 'corral, absorb and process' oil spills, and MIT researchers estimate that a fleet of 5,000 could clean up a gulf-sized spill within one month."
This discussion has been archived. No new comments can be posted.

MIT Unveils Oil-Skimming Robot Swarm Prototype

Comments Filter:
  • Yeah! (Score:4, Insightful)

    by somersault ( 912633 ) on Saturday August 28, 2010 @03:10PM (#33404692) Homepage Journal

    Burning oil is well known for being harmless!

    • Re: (Score:2, Insightful)

      by FooAtWFU ( 699187 )

      What, is this a global-warming remark? All that oil was destined for burning anyway. It's doing a hell of a lot less harm being burnt than it is choking off marine life.

      But nuance is dead in the climate-change "debate"; go figure.

      • by Icculus ( 33027 )
        Well the how of the burning is significant. Think of the smoke produced burning oil in a Diesel engine des vs lighting a pan of it on fire in your garage. The combustion could be just slightly bad instead of extra bad. Otherwise, agreed. Better to turn it into smoke and CO2 than kill the sea critters.
    • by aliquis ( 678370 )

      It was bound to happen anyway. Better than having it in the ocean.

      Only problem I assume is that the oil isn't at the surface.

      Well, that and they don't have 5000 robots, and I don't see how they would "catch 'em all"(it).

    • Why would we want to burn the oil, when we could use it for our cars, especially when the OIL industry tried to make us believe there was a shortage starting...we could use this oil from the spill to actually help us instead of aimlessly burning it up! I prefer Kevin Costner's centrifugal machine where it splits the oil from the water and then can be processed further later on....I think we could figure a way out to build a fleet of these and not only be able to use them to filter out oil from water, but ma

  • by larry bagina ( 561269 ) on Saturday August 28, 2010 @03:13PM (#33404716) Journal
    sure, maybe, if you exclude all the sub-surface oil. And there's a lot of subsurface oil.
    • Indeed, and with the case at hand in the Gulf, thanks to those dispersants, quite of bit of the oil is now a sort of emulsified goo. It has a specific gravity just about equal to that of the ambient seawater (so it hangs in suspension) and even if it could somehow be made to surface, it won't burn. AFAIK, there are no credible plans to get rid of that stuff. I'd imagine that it prolly clogs up fish gills quite well, too.
  • Isn't that a lot of oil to burn? If we don't actually mind burning oil you could power the robots with some/all of the oil they collect instead of using solar power.

    Next thing you need is an electronic sensor that can smell and taste oil in the air and water. Then working as a swarm they can find oil spills, move to them autonomously and consume them.
    • Re: (Score:3, Interesting)

      by vadim_t ( 324782 )

      Solar is much simpler. You just need a panel and a battery.

      Burning oil is more complicated. You need an engine that can burn crude mixed with whatever it mixes with in the water (some rugged diesel engine maybe?), a pump, a tank to have the ability to move through clean water, an electric generator to power the circuitry, a battery as well (for starting the motor for instance). It's going to be heavy and complicated, more prone to failure, and harder to keep afloat. And definitely more expensive. And it'll

      • by Dr Max ( 1696200 )
        solar is nice and simple, if it wasn't for night time it would be perfect.


        I don't think you could get a very good explosion out of the crude oil, but it should burn hot enough to heat and turn a steam turbine. It has plenty of water when its in the ocean. it doesn't have to burn very hot. there aren't many moving parts. it makes sense to use the oils energy if your burning it anyway, but it wont always be on top of oil.

      • by TheLink ( 130905 )

        > Burning oil is more complicated.

        How so? MIT have allegedly already figured out how to collect oil to burn it.

        You could use a Stirling or Ericsson cycle engine or others. The heat sink is the ocean.

        They claim to only need 100 watts of electrical power, and they're willing to have 2 square metres of PV panels.

        There's about 37 megajoules of energy in one litre of crude oil. Assuming your engine has a 10% efficiency from oil to electricity, you need to burn 0.027 millilitres per second. 0.027 millilitres p

  • by Animats ( 122034 ) on Saturday August 28, 2010 @03:27PM (#33404798) Homepage

    OK, each unit sops up some oil, using "nanowires". Then what? The oil then has to be transferred to some collection boat. That part isn't implemented.

    A fleet of semi-autonomous skimmers that deliver oil to a collection ship or a shore station would be useful. Operations like that are risky for small boats, as are operations near shore, near rocks and reefs, and such. So it's a good robot application.

    The "nanowires" just sound like the usual hype from MIT's PR operation (which has gotten out of hand enough to be an embarrassment for MIT.)

    • by hipp5 ( 1635263 ) on Saturday August 28, 2010 @03:35PM (#33404844)

      Then what? The oil then has to be transferred to some collection boat. That part isn't implemented.

      The way I read it was that each bot disposed of the oil by burning it on-site. No need for central collection.

    • Re: (Score:3, Insightful)

      by tomhudson ( 43916 )
      The math doesn't work. 16'x7' is 112 square feet. 33 barrels of oil (which is how much would have to be removed each day by a fleet of 5,000 skimmers) covers that to half a meter thick, and weights 4 tons. You won't be going very fast towing that with a 1/8 hp (100 watt) motor.

      So, skip the self-propelled aspect, just attach floaters to the absorbent, toss them in the sea, pick them up, pass them through a wringer to squeeze out the oil instead of heating it, rea-attach the floaters and toss them back in.

      • As a few other people have pointed out, the idea is that each robot burns the oil as it collects it, it doesn't tow it around. Whether or not that is feasible is still up in the air, but your math is irrelevant as far as their design concept is concerned.
        • by hey! ( 33014 )

          Right. So the key here is downward scalability. In the proposed system concept, the oil/water separation process doesn't take energy input. You only have to put energy into the process when you've collected enough oil to make it worthwhile to bootstrap a process whereby you burn the oil you recover to power further separation.

          The alternative would be to build somewhat larger robots that had some kind of centrifuge separator. That'd work too, but only if you could drop the robot right on a nice thick oil s

          • power can be had, something like the SEADOG [] or the OPT [] would work well near shore, booms could channel any oil to the centrifuge for processing on a cheap semi-permanent basis. Something like the Pelamis WEC [] looks like it wound be towable and able to power processing on the open sea.

        • And I would suggest you get some facts - yo can't just burn waterlogged crude oil by putting a match to it, any more than you can burn a puddle of diesel fuel by throwing a lit match on it.

          So forget the idea of burning it on site - even if you divert ALL the energy to heating the mat, 100 watts won't do it - that's 300 btus. Nowhere near enough to get ignition started. The heat will just be conducted away as the crude turns to semi-liquid state before re-congealing.

          The design doesn't work.

          • I never said burning it on site would or wouldn't work, I just pointed out that that is what their design calls for. Towing it somewhere else was never part of the design and therefore whether or not it can do that is irrelevant.
            • by shawb ( 16347 )
              The article says the oil will be burned locally... basically all these robots would keep bringing oil to a collection point where it is removed by heating the fabric, and then burned by traditional means. I assume at this point the collection point will be a larger ship of some sort where it would be feasible to use the heat from burning the oil to remove oil from other skimmers. The numbers they used don't say that each skimmer would carry 33 barrels of oil at a time; what is realistically meant is that
              • Re: (Score:1, Troll)

                by tomhudson ( 43916 )

                No, it's 33 barrels a day to arrive at the 5,000 robots can clean up 5,000,000 barrels in one month. Do the math. It's not hard - unless you're an MIT "scientist".

                Otherwise, each robot would have to collect 1,000 barrels at one time. That would represent a layer 45 feet high on their 16'x7' fabric. Do you really believe that a 1/8 hp electric motor could tow that much?

                Their math sucks. Their thinking does too,

          • by aliquis ( 678370 )

            If it has already started burning then?

            Because I can't see how anyone can claim how much energy 100 watts is.

            What if it got enough solar cells to produce 100 watts for 24 hours before using the stored energy within 0.2 seconds to ignite the oil?

            • It's not "oil" - its more like tar. It's only self-sustaining under certain conditions - like you need to be able to enclose it with a chimney-like chamber so that the heat of combustion isn't dissipated too quickly, but heats up the rest of the lump. Otherwise you get very incomplete combustion, your mat turns into a lump of asphalt-like glop, and you need to incinerate it to get the tar out. Look at how the ships that burn it off are set up.
        • by aliquis ( 678370 )

          The robot moves around on the skimmer and not the other way around? Moving the robot is probably much easier =P

    • Re: (Score:1, Troll)

      by slick7 ( 1703596 )
      Why bother? According to BP the oil is gone and everything is hunkey dorey.
    • by hey! ( 33014 )

      The "nanowires" just sound like the usual hype from MIT's PR operation (which has gotten out of hand enough to be an embarrassment for MIT.)

      *sigh*. I wish that I was back up in the 'Tech on Boylston Street...

    • by gweihir ( 88907 )

      Exactly my point. They cannot burn the oil, that would not work and damage the fabric. The video says you "heat up" the fabric and it releases the oil. So what is missing from the video is an expensive, energy hungry tender that these things have to go back to whenever they have soaked up the, maybe, half barrel they can hold.

      Suddenly this does not look too good anymore. It certainly does not look fast and "autonomous" goes right out the window with the tender.

      • by khallow ( 566160 )

        Exactly my point. They cannot burn the oil, that would not work and damage the fabric. The video says you "heat up" the fabric and it releases the oil. So what is missing from the video is an expensive, energy hungry tender that these things have to go back to whenever they have soaked up the, maybe, half barrel they can hold.

        Once the oil is released from the fabric, then you burn it. They don't need to hold the oil.

        • by gweihir ( 88907 )

          They do need to hold the oil until they get back to the tender Ship and it is removed there. Burning the oil directly destroys the collector fabric.

          • by khallow ( 566160 )

            They do need to hold the oil until they get back to the tender Ship and it is removed there. Burning the oil directly destroys the collector fabric.

            They plan to remove the oil from the fabric first, then burn it. Since the oil is no longer in contact with the fabric, the fabric won't be harmed.

            • by gweihir ( 88907 )

              Indeed. Bu the whole implementation of removing the oil is missing and will not nearly be as easy or "cool" as suggested.

              • by khallow ( 566160 )
                My take is that they'll probably heat up the fabric (since my impression was that the fabric is less absorbent at higher temperature) and wring it.
      • I'm not understanding why the Hive-ship tender has to be "Energy Hungry" it just follows the swarm. When a worker skim-bot is full it just returns to the hive,
        1. climbs the ramp and dumps it's oil into a tank,
        2. does diagnostics and reports for repairs if necessary,
        3. gets any batteries or fuel tanks topped off,
        4. enter the queue for redeployment.
        The hive ship is getting dry oil that's not even that crude anymore, the heavy residuals sank all ready and the volatiles evaporate

    • You stick around slashdot, you'd think MIT is this giantic grant-sucking corporation.
    • Then what? The oil then has to be transferred to some collection boat. That part isn't implemented.


      The fabric, developed by MIT Visiting Associate Professor Francesco Stellacci, and previously featured in a paper published in the journal Nature Nanotechnology, can absorb up to twenty times its own weight in oil while repelling water. By heating up the material, the oil can be removed and burnt locally and the nanofabric can be reused.

      Yeah, we need someone to point out "FTFA" replies on /.

  • Sounds like an advertisement for the next generation of vaporware projects. So what's its carbon footprint?
  • by Anonymous Coward

    5000000 barrels and 5000 robots, that gives 1000 barrels for each robot and month??
    so in one day one single robot will take up approx 30 barrels: that is more than one barrel per hour
    day and night?
    Seems to me as some MIT miscalculation or am I missing something?

  • Wow! (Score:5, Funny)

    by ironnation ( 1888780 ) on Saturday August 28, 2010 @03:56PM (#33404942)
    SeaWow holds twenty times its weight in liquid, doesn't drip, doesn't make a mess, you burn it off. Made by MIT, you know MIT makes good stuff. Okay, here's some oil stains. Not only is that damage going to be on top, there's your plumes underneath, that's gonna get into your sand, see that. Now we're gonna do this in real time, look at this, it goes on a spill, I don't even have to control it, it just does the work. You following me, camera guy? It acts like a vacuum. SeaWow -- you'll be saying WOW every time. And if you call right now, cause you know we can't do this all day, you get 4999 more SeaWows, that'll clear up a spill in a month. Here's how to order!
  • by 93 Escort Wagon ( 326346 ) on Saturday August 28, 2010 @04:01PM (#33404974)

    Let's see... they're tiny robots, they consume raw materials...

    Mark my words - pretty soon some bright lab jockey will come up with the idea of giving them the ability to build more of themselves using those raw materials. And we all know what'll happen next.

  • Ready the electromagnets!
  • It seems this thing (and the video) is pure fantasy: It does not show anywhere how in practice the oil is going to be removed and how much oil each robot can carry. In my view, that leaves an absolute critical component out and makes this whole thing a publicity stunt and nothing more. Especially its power consumption is a pure lie, as the stated ratings are never going to be enough to remove the oil from the mat. Think of a swarm of these things and a large tender with huge energy needs that they have to g

  • Collecting and burning oil but for power? Seems odd. Maybe a mech eng can explain.

    • by Raptoer ( 984438 )

      The oil being collected is generally very heavy, but can vary in composition. Oil of that type cannot be used in any normal type of combustion engine. That limits you to engine that rely solely on difference in temperatures rather than ones that try to control combustion. You could run a sterling engine off of this, but you have to consider what happens between oil slicks in the same area. If the slicks are too far apart, your robots run outta energy and just sit.

      Solar energy on the other hand will always b

  • oh, wait... I have an oil powered solar collector. It seats seven and even has cup holders. It uses collected solar radiation to warm the seats and the beverages in the cup holders, but that feature only works during the day time.
  • As shown, there's no provision for navigation lights.

    I believe under the maritime Rules of the Road [] [PDF], one of these would be classified as a ``vessel not under command'', in which case it should display two red lights, one above the other, at night, and two black balls ditto during the day. (Rule 27 (a)) I don't know whether these are large enough to require such displays, but hit one of them in a sailboat at good speed, and you could be in real trouble.

    Additionally, a radar reflector would b

  • I don't think the designers appreciate the difficulties what they are proposing.

    First they suggest someone could/should have thousands of these autonomous vehicles sitting around (in an operational state) waiting for oil spills (with no auxiliary purpose).

    Second they ignore the sheer chaos that would ensue as thousands of small, low-profile vehicles travel in and around other vessels necessary to actually stop/control an oil spill. These things wont show up on radar. They probably can't be seen at night, an

  • Remember that any technology looks best a) when in development, and b) to its originators. Take any and all numbers and promises and scale them back from 50 to 80%.

    Random thought: Put one in the water upside down---can it right itself? Because it's guaranteed that in heavy seas they'll be flipped over every so often.

    On the other hand, given that the waves could flip it back up, as well, on average you'd have 50% of them rightside up at any time. It might be easier to double the number of gizmos t

  • A superfluous invention. Why not just rely on Alcanivorax to clean up the mess? It's already there, it replicates itself in direct proportion to the amount of oil, it self-destructs when there's no oil left to clean up, is 100% bio-degradable and it costs nothing, doing the job in half the time.
    • by vadim_t ( 324782 )

      Well, some googling suggests that it's aerobic, and needs extra nitrogen and phosphorus.

      So, yeah, it'll happily munch on the oil, but it consumes oxygen, and it requires nitrogen, which is a recipe for an algal bloom for yet more deoxygenation. So not only everything gets to get poisoned by the oil while it's there, they'll also suffocate when the bacteria get to it. That sounds like an awesome plan.

      • Don't promote your hypothesis too widely. There are real-world observations in the Gulf of Mexico that falsify it.
        • by vadim_t ( 324782 )

          Glad to be proven wrong, then.

          Still, I don't see why not clean it up even faster. Not like the oil is doing any good floating there. And some googling suggests that the bacteria only eat hydrocarbons, so whatever remains after that still would need to be dealt with.

  • Can MIT come up with something else than "solar powered nano technology autonomous robot swarms". Is this the "Build a really cool solution, and then spend a decade looking for the problem it solves"
    Would be nice to see some innovation there. Are they not supposed to have the skills and intelligence to think outside the box and go in new directions?
    - a NON-autonomous robot
    - a FOSSIL powered something
    - some new MACRO technology.

  • 3%? I guess it's the best that MIT could do, given their comprehension of the problem to solve. And people still wonder why I think the Cardinal Red 'S' looks pretty good.
  • The interesting aspect in the MIT approach is that they approach the problem with a set of relatively small devices. Due to the wavy water, the oil patches become smaller and distributed. With the relatively small MIT robots they can scent out oil patces the same size. Other approaches involve large machines which become ineffective for small patches.

I've got a bad feeling about this.