5 ms·

What are the radiation levels the failed robots experienced? Searching, I keep finding information about radiation levels in reference to human health. But I'm

What are the radiation levels the failed robots experienced? Searching, I keep finding information about radiation levels in reference to human health. But I'm interested in the total dose those robots get before they fail. I used to work on space electronics that had to survive mega-rad levels of total dose radiation. How quickly do electronics accumulate those levels inside Fukushima?

pvaldes 8y ago
Maybe I sound naive, but why do we need to move electronics there?. Why they don't just use a disposable lead tube ending in the minimum working telescopic lens (mechanical) system possible and a cable moving electrons inside to build an image in a computer at km of distance. The brain of the robot should be out of the building, not inside. Couldn't we estimate radiation by how long some well known matherial is surviving before to crash instead to use a sensor?.
- pjc50 8y ago
  Ah, the HN "why don't they just"... A good overview: https://www.wired.com/story/fukushima-robot-cleanup/ https://www.wired.com/story/fukushima-robot-cleanup/ containing the excellent phrase "Until we send the bot in, we don’t know what the conditions are. And after it’s sent, we can’t change it". https://www.independent.co.uk/news/world/robots-fukushima-nuclear-disaster-dying-probe-clean-up-tepco-toshiba-reactor-nuclear-radiation-a7612396.html https://www.independent.co.uk/news/world/robots-fukushima-nu... : "The latest attempt to harvest data on Fukushima failed after a robot designed by Toshiba to withstand high radiation levels died five times faster than expected. The robot was supposed to be able to cope with 73 sieverts of radiation, but the radiation level inside the reactor was recently recorded at 530 sieverts per hour."
  - jandrese 8y ago
    If anybody is curious how much radiation a sievert is, there is a handy chart for reference: https://xkcd.com/radiation/ https://xkcd.com/radiation/ 530 sieverts/hour would mean a person exposed for less than a minute would be a dead man walking. Even a 30 second exposure would be difficult to recover from. Electronics would suffer from constant bit flips. You would really need to keep most or all of the electronics back in a shielded box at the end of a long cable and make the robot as dumb as possible. This sadly only goes so far since you do need sensors that are made out of electronics on the robot. Cameras for example. It's a complicated system, and worse you need this robot to be quite capable if it's going to deal with uneven terrain and eventually breaking up and removing the melted fuel.
    - pvaldes 8y ago
      > a person exposed for less than a minute would be a dead man walking. Yup, fortunately robots aren't alive We know how to fabricate telezooms to take a photo to an animal at 400m distance. We use periscopes rutinely that do not need electronic necesarily. We know how to take a photo to the reflection in a mirror and zoom it later. We have the technology to make a mirror with a very finely polished metal surface (would be that sensible to radioactivity?), and we have all the time in the world to make a big photo secuentially from many small simple photos, ommatidium like. with less info encoded ... Why we need electrons when we could use photons and an optic cable to translate the info at 100 km in some fractions of second? What happens with an optic cable when radioactivity hits it?
      - nate_meurer 8y ago
        Optical materials of all kinds are vulnerable to darkening, clouding, and embrittlement by much lower levels of ionizing radiation than those in the reactor [1]. Darkening along the length of an optical fiber adds up quick, and will quickly render it opaque. Shielding won't help you at these radiation levels; you would need inches of lead cladding [2], and even then your fiber optics might not have a useful lifespan, exposed to at least tens of Sv/h along much of the length. And shielding offers nearly zero protection against neutron radiation. Also note that the radiation in the reactor can heat the internal elements of your perisccope by hundreds of degrees, so without a cooling system your optical system's tolerances are at risk. Remember to make everything waterproof, and robust against shock and vibration. After you figure all that out, you still need to get your periscope into the reactor somehow. Humans can't go closer than about a kilometer, and any robot you use must be able to navigate industrial wreckage. These are not simple problems. 1 - http://www.dtic.mil/dtic/tr/fulltext/u2/a013786.pdf http://www.dtic.mil/dtic/tr/fulltext/u2/a013786.pdf 2 - https://www.eichrom.com/wp-content/uploads/2018/01/gamma-ray-attenuation-white-paper-by-d.m.-rev-4.pdf https://www.eichrom.com/wp-content/uploads/2018/01/gamma-ray...
        pjc50 8y ago
        "Virtually unusable after lengths of as little as 10m", wow. I'd no idea they were that vulnerable.
  - geomark 8y ago
    At 530 sieverts (53,000 rads(gamma)) per hour mega-rad hard electronics could last about 18 hours. Long enough to do some work, but expensive if the robot could not be rennovated and they had to have an army of expendable robots. Such electronics have existed for quite a few years in the aerospace industry. Hard to come by and expensive for sure.
joe_the_user 8y ago
Well, that was a lesson on how crappy Google has gotten with historical, even recent historical data. All searches for Fukushima robots were events of the last two days. At least adding "2012" let me find some things. Here's a very simple article on radiation hardening, which incidentally describes the danger to electronics. https://spectrum.ieee.org/tech-talk/semiconductors/design/radiationhardening-101 https://spectrum.ieee.org/tech-talk/semiconductors/design/ra... iRobots entered the main building in 2011. What happened just now is robot actually entered or surveyed the actual reactor vessel, in the building. https://spectrum.ieee.org/automaton/robotics/industrial-robots/robots-enter-fukushima-reactors-detect-high-radiation https://spectrum.ieee.org/automaton/robotics/industrial-robo...
- shaftway 8y ago
  https://www.google.com/search?q=fukushima+robots&tbs=cdr%3A1%2Ccd_min%3A%2Ccd_max%3A2%2F1%2F2019 https://www.google.com/search?q=fukushima+robots&tbs=cdr%3A1... Limiting results to pre-2019/02/01 seems to go back to an article in 2017. Searching before that... https://www.google.com/search?q=fukushima+robots&tbs=cdr%3A1%2Ccd_min%3A%2Ccd_max%3A2%2F1%2F2017 https://www.google.com/search?q=fukushima+robots&tbs=cdr%3A1... ... seems to pick up a good variety of articles going back to 2011.

pvaldes 8y ago

Maybe I sound naive, but why do we need to move electronics there?. Why they don't just use a disposable lead tube ending in the minimum working telescopic lens (mechanical) system possible and a cable moving electrons inside to build an image in a computer at km of distance. The brain of the robot should be out of the building, not inside. Couldn't we estimate radiation by how long some well known matherial is surviving before to crash instead to use a sensor?.

pjc50 8y ago
Ah, the HN "why don't they just"... A good overview: https://www.wired.com/story/fukushima-robot-cleanup/ https://www.wired.com/story/fukushima-robot-cleanup/ containing the excellent phrase "Until we send the bot in, we don’t know what the conditions are. And after it’s sent, we can’t change it". https://www.independent.co.uk/news/world/robots-fukushima-nuclear-disaster-dying-probe-clean-up-tepco-toshiba-reactor-nuclear-radiation-a7612396.html https://www.independent.co.uk/news/world/robots-fukushima-nu... : "The latest attempt to harvest data on Fukushima failed after a robot designed by Toshiba to withstand high radiation levels died five times faster than expected. The robot was supposed to be able to cope with 73 sieverts of radiation, but the radiation level inside the reactor was recently recorded at 530 sieverts per hour."
- jandrese 8y ago
  If anybody is curious how much radiation a sievert is, there is a handy chart for reference: https://xkcd.com/radiation/ https://xkcd.com/radiation/ 530 sieverts/hour would mean a person exposed for less than a minute would be a dead man walking. Even a 30 second exposure would be difficult to recover from. Electronics would suffer from constant bit flips. You would really need to keep most or all of the electronics back in a shielded box at the end of a long cable and make the robot as dumb as possible. This sadly only goes so far since you do need sensors that are made out of electronics on the robot. Cameras for example. It's a complicated system, and worse you need this robot to be quite capable if it's going to deal with uneven terrain and eventually breaking up and removing the melted fuel.
  - pvaldes 8y ago
    > a person exposed for less than a minute would be a dead man walking. Yup, fortunately robots aren't alive We know how to fabricate telezooms to take a photo to an animal at 400m distance. We use periscopes rutinely that do not need electronic necesarily. We know how to take a photo to the reflection in a mirror and zoom it later. We have the technology to make a mirror with a very finely polished metal surface (would be that sensible to radioactivity?), and we have all the time in the world to make a big photo secuentially from many small simple photos, ommatidium like. with less info encoded ... Why we need electrons when we could use photons and an optic cable to translate the info at 100 km in some fractions of second? What happens with an optic cable when radioactivity hits it?
    - nate_meurer 8y ago
      Optical materials of all kinds are vulnerable to darkening, clouding, and embrittlement by much lower levels of ionizing radiation than those in the reactor [1]. Darkening along the length of an optical fiber adds up quick, and will quickly render it opaque. Shielding won't help you at these radiation levels; you would need inches of lead cladding [2], and even then your fiber optics might not have a useful lifespan, exposed to at least tens of Sv/h along much of the length. And shielding offers nearly zero protection against neutron radiation. Also note that the radiation in the reactor can heat the internal elements of your perisccope by hundreds of degrees, so without a cooling system your optical system's tolerances are at risk. Remember to make everything waterproof, and robust against shock and vibration. After you figure all that out, you still need to get your periscope into the reactor somehow. Humans can't go closer than about a kilometer, and any robot you use must be able to navigate industrial wreckage. These are not simple problems. 1 - http://www.dtic.mil/dtic/tr/fulltext/u2/a013786.pdf http://www.dtic.mil/dtic/tr/fulltext/u2/a013786.pdf 2 - https://www.eichrom.com/wp-content/uploads/2018/01/gamma-ray-attenuation-white-paper-by-d.m.-rev-4.pdf https://www.eichrom.com/wp-content/uploads/2018/01/gamma-ray...
      - pjc50 8y ago
        "Virtually unusable after lengths of as little as 10m", wow. I'd no idea they were that vulnerable.
- geomark 8y ago
  At 530 sieverts (53,000 rads(gamma)) per hour mega-rad hard electronics could last about 18 hours. Long enough to do some work, but expensive if the robot could not be rennovated and they had to have an army of expendable robots. Such electronics have existed for quite a few years in the aerospace industry. Hard to come by and expensive for sure.

joe_the_user 8y ago

Well, that was a lesson on how crappy Google has gotten with historical, even recent historical data. All searches for Fukushima robots were events of the last two days. At least adding "2012" let me find some things. Here's a very simple article on radiation hardening, which incidentally describes the danger to electronics. https://spectrum.ieee.org/tech-talk/semiconductors/design/radiationhardening-101 https://spectrum.ieee.org/tech-talk/semiconductors/design/ra... iRobots entered the main building in 2011. What happened just now is robot actually entered or surveyed the actual reactor vessel, in the building. https://spectrum.ieee.org/automaton/robotics/industrial-robots/robots-enter-fukushima-reactors-detect-high-radiation https://spectrum.ieee.org/automaton/robotics/industrial-robo...

shaftway 8y ago
https://www.google.com/search?q=fukushima+robots&tbs=cdr%3A1%2Ccd_min%3A%2Ccd_max%3A2%2F1%2F2019 https://www.google.com/search?q=fukushima+robots&tbs=cdr%3A1... Limiting results to pre-2019/02/01 seems to go back to an article in 2017. Searching before that... https://www.google.com/search?q=fukushima+robots&tbs=cdr%3A1%2Ccd_min%3A%2Ccd_max%3A2%2F1%2F2017 https://www.google.com/search?q=fukushima+robots&tbs=cdr%3A1... ... seems to pick up a good variety of articles going back to 2011.