Japan nuclear crisis: Where are the robots?


Today, Reuters asked the questions we've all been thinking. Namely, "This is Japan, right? They've got lots of robots, right? So why aren't there robots helping out at Fukushima?"

The answer they got is somewhat unsatisfying. Mostly, there's a lot of, "Nope, no robots being used right now. Sorry." But there are a couple of interesting facts in the article.

• Robots are a part of the wider nuclear industry, capable of doing tasks like detecting radiation, and scaling walls. And robots were used in cleanup at Three Mile Island and Chernobyl.

• There are no robots currently working at Fukushima Daiichi. It's unclear why. Although, there is, apparently, one radiation-detecting robot on site.

• There is one possible reason why robots aren't being used at Fukushima. It all comes back to power plant design. A South Korean nuclear official told Reuters that robots and power plants have to be designed with each other in mind. Fukushima Daiichi, which dates to the 1970s, may simply not be navigable to newer nuclear helper 'bots.

Reuters: Japan a robot power everywhere, except at nuclear plant

Via Jer Thorp

Image: Some rights reserved by ChelseaWa


  1. More than that, why haven’t they taken one of those exoskeletons they’re working on, wrapped it and the operator in a lead lined spacesuit equipped with a camera to make up for the visor, and sent him in to toss girders around and shove the fuel back underwater by hand? As I told my wife last night, this is Japan! They should be all over this.

    (Obviously, 99% joking to reduce the stress of the situation, and not trying to make light of the real risks and people who are quite probably working themselves to death. But 1%…yeah, why hasn’t someone built that yet?)

    1. They did. But you know, there was this thing called an earthquake, followed by this thing called a tsunami, and surprisingly enough a lot of stuff got busted and buried.

      California needs to send their robots, OK? Get ’em over there. You can send Dweezil Zappa too.

  2. The robotics-related question that has bothered me is: where are the aerial surveillance drones?

    I mean, hell, my local Best Buy sells the ARDrone. You’d think something like that would exist in Japan. Sure, it only gets 12 minutes of flight time, but the video footage it could provide would be invaluable to diagnosing issues – or even surveying the tsunami damage/rescue operations.

    But who knows, maybe the radiation is causing too much EM interference for any of that wireless tech to work right. :)

  3. Wait… The “newer nuclear helper ‘bots” are designed to work at newer nuclear power plants, but not at older nuclear power plants? Because accidents are only going to happen at NEWER nuclear power plants? Really?

  4. This EXACT question has been bugging the superheck out of me since I first heard of the nuclear plant difficulties over there. Maggie, I always love your posts, and thank you again!

  5. as a side note: high levels of radiation can be crippling to electronic circuits as well as biological systems. designing a radiation hardened robot requires additional considerations.

  6. Suppose hose-wielding robots had been stored on site. Would they still be in working order after the tsunami? How would they be getting their power in the event of a long-running emergency? They don’t have power at the plant, which is the heart of the problem, and at the rate that water is boiling off, they’ll need to operate hoses for a long time.

    It might make more sense to have back-up pumps powered by the heat difference, Stirling engines or something similar.

    1. I believe newer plants are designed just that way (there was one article, which I can’t find now, that said newer plants are set up so that gravity drives the water through the system–outlet in the bottom of the pool, steam goes up, condenses, and rolls back down into the holding tank, presumably.) But the plants in trouble were built in the 70s, and don’t have that design.

  7. “Robots are a part of the wider nuclear industry, capable of doing tasks like detecting radiation, and scaling walls.”

    Wait, what? Scaling walls?

  8. There are Japanese robots available, but they’re all scheduled up with awkward-dancing and creepy-smiling duties.

  9. So what happens if the structure of a newer plant changes can the robot still function there? sayyyy like in and earthquake or post tsunami situation? or if Homer Simpson spills coffee?

    As for lead suit guy… Suit would probably have to be too thick to be practical. Also there are secondary effects that come from the shielding. For example when a high energy beta particle is stopped by a dense material such as lead you get secondary electromagnetic radiation from the abrupt slow down. This is why you see plexiglass used in labs since the deceleration of the particles is not so abrupt and secondary radiation much less energetic. So if you have awesome lead suit to protect from from gamma radiation you might get fried from it stopping the beta particles. ( my knowledge on the subject is fairly out of date and has probably had song lyrics from the past decade or 2 over writing it). eg http://en.wikipedia.org/wiki/Bremsstrahlung_radiation

    As for Robotics – I would not be surprised if these events really kicked the japanese robotics industry forward in the coming decades as people ponder questions like the ones raised in the article.

  10. It’s pretty sad that there are no robotic Chinooks dumping water 24/7 over the reactors, since the technology is pretty much there–it’s just a matter of money. If only the militaries of the world could be convinced to spare a dime on technologies with peaceful applications…

  11. IIRC, the lead designer of Russia’s moon-rover, the Lunakhod, put together a similar robot so they could look into Chernobyl.

    1. Yes, they did it. Used it on the top of the building to cleanr it. And it stopped working pretty soon because of the radiation – which is propably the answer for the question.

  12. What’s even more puzzling to me is that we keep hearing how the danger of radiation exposure is making it difficult for workers to deal with the problems at the reactors. Don’t they have radiation suits for them to wear?

    1. Many of the workers (if not all the ones left) have been in radiation suits for days. The discussion of radiation being too high to approach assumes that a suit is being used.

      As an aside on robots, the radiation won’t be nice to most off the shelf robots either. Not that they wouldn’t be of some use, but it’s something to think about.

      And getting a robot to carry a hose over there isn’t the only issue. It’s that the pumps themselves have been a problem, along with the power to the site. It *seems* like it wouldn’t be that hard to drop in some generators and pumps, but my guess is that it’s more difficult than we think.

      Tepco hasn’t proven to be completely forthright about things, but I doubt that they, or all the other people working on this, are complete morons.

    2. Radiation suits (which one assumes the workers have been wearing from the start) are nowhere near as effective as TV and science fiction portray them as being.

      They’re generally intended to prevent the wearer from inhaling radioactive particles/gas. While some suits have shielding, it’s minimal and won’t do much against beta and gamma radiation.

  13. “Where are the robots?” is an unfair question in my opinion. Since when does japan have a robot that can perform any task we ask of it (in this case: fire fighting, high pressure plubming, high voltage electrician, heavy debris removal, all within an electronically hazardous space designed for humans), and is already sitting on the shelf? Perhaps a team of highly specialized robots? I feel like the question, while not intentionally, is slightly condescending and ignorant of robotic technology (I guess that could explain why the question was asked).

    I’m also guessing that the robots used in the prior nuclear disasters were not first responders, but instead were modified robots designed to perform specific tasks, and were utilized in later phases of the disaster like cleanup and monitoring.

    As Tribune mentioned, the new awareness this disaster is creating will push development into new areas that until now have not been the robot designers primary goals.

  14. I was wondering about that this morning. There are, as a standard on-the-shelf thing, fire-fighting robots. These can carry water hoses. Why the heck can’t they get one of those there and use them to get a hose to the spent fuel pools? A fire-fighting robot should be able to get inside the building and roll up to the pool, or at least get a high pressure nozzle a lot closer than people can.

  15. I was thinking to myself last night: someone should call up Grant Imahara from Mythbusters and have them R/C up a couple of fire engines to pour water into these buildings. What could it hurt (saying they don’t have any issues where they drive a firetruck into the side of a reactor, but the reactors are tough, so imho it’s worth it)?

  16. I131—is both a beta and gamma emmiter—so in a lab typically a beta emmiter gets plexiglass shielding while a gamma emmiter requires 12 inches of lead. Beta particles actually generate gamma when they hit lead so a lead apron (.5 inches thick) which does not fully block gamma will actually cause more gamma exposure when the beta particles hit it. Working with these isotopes, you normally limit time of exposure and increase distance with tongs etc, shielding is remarkably ineffective for a human, you shield the source and samples with lead but that’s about all you can do. Unfortunately radiation also has a tendency to screw up unshielded electronic devices, so they can’t just pull something off the shelf unless they want to add 12 inches of lead to it. Also keep in mind that they are the electric plant so when there is no power, there’s also no hope of restoring power until a reactor can be brought back on line. It’s really a horrible situation because I can tell you that the people they use to maintain generator function are typically outside contractors and not nuclear specialists so everything needs to be shut down, decontaminated and carefully monitored before they can come in to fix any equipment at the site. At least that’s what they do in the states.

  17. Yeah, that’s a great idea guys.

    So when the robots inevitably rise up against us, not only will they be unstoppable killing machines, they’ll also be radioactive.

    Real bright, Einsteins.

  18. Integrated electronic devices are susceptible to radiation damage. it’s one of the problems with space flight, where radiation is a real risk. You would have to have space qualified or higher electronics in the robots (very special devices).

  19. Oops, error in my comment about the shielding, iodine is a weak gamma emitter and does not need 12 inches of lead(that’s for sodium, which I work with, which would be much worse, due to the 12 year half life as compared to 8 days) (Re:Maggie’s comments about experts)

    Still it’s important to note that shielding is not a perfect solution, reductions in time and distance (especially distance) are also required.

  20. I’m appalled at how so many seem to oversimplify: “Some solution was know to exist to some vaguely similar problem at some other time under some other conditions – therefore it should work here and now”. Sigh.

    “Radiation” refers to all sorts of different particles, of differing types and different energies. The calculations of what is required to completely stop all of the radiation from any widely mixed source such as this is quite simple: a nearly infinite amount of material. Surprised? That’s because I used the absolute “completely stop”. The much more practical question is how much shielding is necessary to reduce the level to some acceptable (comparable to natural background) levels. – Now there the answer gets amazingly complex. As others have pointed out above, as you put in more material to stop high energy gammas, that in turn produces scattered radiation of lower energies, and additional x-ray emissions as well as the aforementioned “braking radiation”. Suffice it to say that the thick steel and concrete is a reasonable start, so shielding people and electronics becomes difficult.

    To answer Mister44, ionizing radiation will cause unexpected (often tiny) groups of electrons to be released inside the electrical circuits at unexpected times and places – that gets them all “confused” since usually everything is so orderly inside computers and such. So its not short circuits but more like bits changing from 1 to 0 or vice-versa. Robots don’t work well when confused. Yes there are ways that satellites and high-reliability systems deal with the inevitable bit-flips that occur all the time – but those are not the usual robot applications. Bomb-disposal robot makers would never invest in extreme rad-hard systems and architectures because the market is (was?) nearly zero dollars for that specialization.

    1. Thanks for the info. Would a hard wired robot, with no electronics, just gears and motor and controlled via wire have similar problems?

  21. In Chernobyl the robots were fail because the radiation knocked their bits out — seriously, even the teleoperated ones the chips were just toast.

  22. There was an article about that yesterday in Atlantico (in French): http://www.atlantico.fr/decryptage/robots-francais-pour-centrale-japonaise-en-panne-57041.html

    The interviewee works for Aldebaran Robotics (the company that makes the Nao robot, among others). He states a few interesting points. The Japanese have worked a lot on entertainment and service robots, as that’s the highest potential markets. But robots designed for working in hostile irradiated environment fit in niche markets and they are mostly being designed in Germany, USA and France. He also says that without specific shielding, robots would fail in a matter of minutes in Fukushima because CPUs can’t stand radiations.

    So, robots are currently being shipped from USA and France to Japan.

    Examples of such robots (and interesting data about intervention techniques in case of nuclear accidents) can be seen there: http://www.groupe-intra.com/pages2/intervention/moyens1.htm

    1. One would think that there would be a number of useful items in transit to the work site as we speak.

  23. I work for a robotics company here in Canada called Inuktun Services that specializes in remote robotic inspection equipment for the nuclear industry. Our Versatrax 450 is currently used in a number of power plants in the US. Since this past Monday, we have been trying to find the right people to send equipment to in Japan , however because of the language barrier, it has been nearly impossible to contact the right authorities that would be able to possibly utilize our equipment. Anyone with information that may of be of use is urged to contact me cross@inuktun.com.

    1. My $0.02. Google a Japanese translator for hire. Call them. Hire them. Pitch that person your product, have them pitch it back to you in English a couple times, to your liking. Then, make a plan how to look up the first few cold calls & have them start making those calls to Japan. Good luck.

    2. Hi

      Have you found a contact person yet? I found The Center for Robot-Assisted Search and Rescue and they may be able to help


      Hope you are able to find a contact to get your robots to Japan.

      All the best,

      C.T. Kirkpatrick

  24. I saw a talk at Oak Ridge National Lab years ago about using robots to do tasks at nuclear reactors to save humans from radiation. The example they gave was when they needed to put a radioactive substance into a lead jar and screw on the lid. There were robotic hands that could put the substance in the jar, but to screw on the lid, pre-robot, a worker would run into the room and turn the lid a fraction of a rotation, and then run out. Then a 2nd worker would do the same thing. Etc.

    Now we have robots that can do very specific tasks like this in easy-to-navigate environments. But the environment at Fukushima is radioactive, hot, and probably full of debris. I don’t think our robots are that “smart” yet. Technology solves some problems (and creates others), but technology can’t solve problems that we haven’t foreseen.

  25. See, this is why all nuclear reactors should be below sea level and on the coast line. That way, if they need to cool the plant down, they can just open the floodgates and let the seawater pour in.

    bonus: scoop up the fish and other sea creatures that die from the radiation for use in pet food

  26. Are robotics combined with radio controlled functions being discussed?

    How about limited functions for radio controlled robotics? Remote control the water truck close with water hoses ready. Then drag the truck back with chains and refit to go forward again.

  27. If electronics are incredibly difficult to rely on, the only other option for a machine, would be to use purely mechanical parts. But you’d have a lot of tubes running out the back of the robot. Even if the tubes were bundled in one giant sleeve, that’s a lot of parts that could fail. Still, it’s an idea.

    The “camera” would be a bundle of fiber optic filaments. The wheels would be spun by pneumatic fluid inside. All of the parts would be actuated by fluid or air.

    I can’t even back up my vacuum cleaner without it becoming tangled in the power cord.

  28. The challenges involved in modifying some existing system to make it *quickly* able to perform remotely are daunting. The use of “direct control” without any (susceptible) electronics is not unthinkable – just not likely to be off the shelf for a land based system. An agile engineering team given a supportive shop and a stock of analog subassemblies might put something together in a week, but would it be sufficiently reliable? The underwater ROV systems are cabled rather than RF control, but they use digital signaling.

    I wonder how long it would take a team of engineers to modify a conventional truck with a water cannon to make it remotely controllable. Could be a couple of weeks if one was willing to accept high risk operational safety (will it lose control and back over me?).

    The basic hobby style of radio control has a long history of pure analog signaling, and therefore not highly susceptible to the radiation induced misbehaviors as digital electronics are, but they suffer from other forms of low reliability and difficulty in assuring safety. Even the hobbyists are moving more and more to digital RC. There are few commercial quality systems that don’t currently have at least some components that are digital.

    The current situation requires an existing “robotic” (perhaps better teleoperated) system with some applicable capabilities (stair climbing, door opening, motion on uneven floor surfaces, etc) that has been tested and shown to be reasonably safe – else a robot could just add to the complexity of the problem. No need for autonomy, just a robust mechanical system that has some dexterity and can carry in some cameras and a few radiation detectors and link that data back. In an environment that was never intended for a robot, and now likely strewn with debris and flooded with water that is a mighty tall order.

    I’m sure there are a few out there with systems that have some useful capabilities – and they are all eager to get them into the scene. We can hope that some truly well-designed and capable systems are able to get in there and help, but I can understand that those who are currently tasked with managing the disaster are going to have trouble sifting through the options being presented – and its decidedly NOT a good idea to just let anyone come in and try.

  29. I really am baffled and somewhat upset by these official or responses or those responses by “authorities” on the subject. Having worked in the automation industry building robotics and automated factory lines, etc., #1 there is no excuse for not having robotics in these types of potentially hazardous areas given the high degree of automation found in Japan (we’ve seen robots 15 years ago in Japan that were able to weld joints together on radiators for air conditions, which is a very difficult process considering each joint is unique and requires unique calculations and approaches. It was truly amazing to see one human “babysitter” for an entire factory full of robotic workers doing such complex tasks.).

    #2, that is simply not true what this fellow from S. Korea is saying. You 100% do not have to mutually design robots and their facilities at the same time. I have a friend with whom I worked who builds robots for custom situations such as exhaust scrubbers that are capable of climbing irregular smoke stacks, etc. Each robotic creation is an innovation and consequently it takes effort, yes, however that is what this is for.

    At any rate, this is an embarrassment and folks are just trying to sweep it under the rug. For Japan this is a serious blunder.

Comments are closed.