Globally Distributed Sensor Net Monitors Air Quality and Radiation

Radu Motisan has been building a global environmental surveillance network which first monitored radiation levels, and since has added the ability to measure air quality. He believes that people need to be more aware of the environment around them in a similar way that society has awakened to issues about personal fitness and health. We can’t do this without a simple and reliable way to measure the environment.

He discussed the project at length during his presentation at the 2015 Hackaday SuperConference. Watch that talk in the video below, then join us after the break for more details on the hardware and infrastructure that collects and presents the data publicly.

Beginning with uRADMonitor

uRADMonitor PCB shown on top of its metal case
uRADMonitor PCB shown on top of its metal case

The uRADMonitor was the original hardware which Radu used to develop the network. To the user these are simple devices that need only be plugged into Ethernet and power in order to function. They then enumerate on the network and begin phoning home with background radiation measurements.

With the known location of each node, this data is graphed over time and made available on the website. Radu did a lot of work to grow the number of nodes in the wild, and both the concept and execution were recognized with a place in the semifinals of the 2014 Hackaday Prize.

Growing to Measure More

Environmental Monitor pcb next to its enclosure
Environmental Monitor PCB next to its enclosure

With a proven data network and many nodes already in the field, Radu started looking for ways to improve the hardware. His goal was to leverage the data network with better nodes that are portable, position aware, and with an increased number of sensors. Portability is handled by adding a rechargeable battery and WiFi to replace the hardwired Ethernet connection. The radiation sensor from the previous edition of the hardware is joined by an air particulate sensor, as well as a CO2 and VOC sensor. The new model also includes a small touchscreen which provides an on-device user interface that is new to the system. The work was recognized as both a finalists in the 2015 Hackaday Prize and a finalist for Best Product.

Interesting Data

Each pentagon represents several sensors deployed in that geographic area
Each pentagon represents several sensors deployed in that geographic area

Radu shares a couple of very interesting stories about the data collected by his hardware. The addition of particulate, CO2, and VOC sensors were an immediate success. Even in his relatively small city of 300,000 people, the hardware recorded clear pollution changes which correlated with heightened traffic patterns.

On the radiation side of things he noticed a spike in one node located in the United States. His first thought was that there may have been a malfunction. But when inquiring with the node’s owner he found that was not the case. The owner’s daughter had been conducting school experiments using depleted uranium and the sensitive equipment registered a higher background radiation level due to the proximity of these samples. Along similar lines, Radu has noticed that turning on the equipment immediately after it has been on an airplane registers a spike in radiation from more cosmic rays in the upper atmosphere.

Perhaps most intriguing is his recollection of a sensor he built and strapped to the outside of his car. Driving around his country he detected several big radiation spikes. This data has been withheld from the reporting network. He believes he might have detected areas in his country rich in uranium. This data may have something to do with the nation’s natural resource which is why it has been held back. Other than this, all data being reported is available and can be viewed on his website.

54 thoughts on “Globally Distributed Sensor Net Monitors Air Quality and Radiation

      1. It’s external magic that put all of that crap on that site?

        It’s very simple, if you do a global good-for-everybody site you don’t put all that advertising and tracking stuff on it.and then ask volunteers to supply the data and buy the sensors.

        and then there is this:
        “Driving around his country he detected several big radiation spikes. This data has been withheld from the reporting network. He believes he might have detected areas in his country where uranium is being enriched. This data is most likely protected by national security laws which is why it has been held back. ”

        So he already censors the data before even being asked? Yeah this is a reliable project alright.

        1. Ugh. I get it, you want everything to be 100% ideologically pure. Real life isn’t quite that black and white. If Radu revealed the location of those radiation spikes and the Romanian government got pissed, he could get bogged down in legal bullshit that shuts the project down entirely–he’s right to keep that bit quiet. In the US, we deal with the same crap when our projects butt up against ITAR regs–unless you want a painful fight, sometimes it’s best to choose your battles and move on.

          You take issue with advertising? He’s got a product to sell. He’s running a crowdfunding campaign. He’s trying to move _units_ to support his _business_. He needs to get paid to support his work–not everyone wants to just do side projects. Go ahead and question his methods–do it politely, maybe in private (he’s quite accessible)–but don’t question his motivation.

          Or just be an asshole, I suppose. Your call.

    1. How about you enlighten us as to why you take offence? What about google do you have an issue with? Otherwise your post is useless and your use of offending language begs one to wonder why anyone should take you seriously.

      1. They’re concerned about pervasive tracking everywhere. The concern has merit, but they’re doing a good job of making people not want to listen. *sigh*

        Use or something similar and tracking ceases to affect you. Just silently watch everyone else walk off the cliff.

        1. Sites that use googleapis won’t work with it blocked.
          And sites that use google maps won’t show a map unless you allow google to run scripts.
          And once they run their scripts they can get enough data to ID you across the web BTW. regardless of any VPN.

          And sites with cloudflare can’t be visited without going through the (US) transparent proxy.

  1. I’ve found the idea of this pretty interesting for a while now, but I’ve got to admit to occasional paranoia. I tend to never stick devices on my network that I don’t control. It would be neat if another design could be created which uses a SIM card to connect to his servers, assuming the amount of data being sent isn’t extravagant. After some of the data logging I’ve done in the past I’m assuming there wouldn’t be too many MB sent per month. That would truly allow for a portable design.

    1. While I relatively sympathize with your concern and don’t like the idea of having no control or feeling of control over this device… I think this is a slam-dunk example of “let it go, man”. The SIM/cellular idea would be wasteful and expensive.

      What we NEED is a device that people can deploy that can be the foundation for a shadow-internet that costs nothing but the maintenance of each private node. ;-) (video and large files absolutely NOT-IN-SCOPE and QOS-limited heavily – more intended for email, web, chat protocols, OS cellular network back-haul, etc)

    2. You aren’t being paranoid. Luckily, there is a solution. But first, a digression!

      When I managed networks professionally, I dedicated a sub-network/VLAN with default deny rules in place (ie. unless it was white-listed, all traffic, in or out, was dropped) for this kind of thing. That was great when we were forced to install devices on our network that needed Internet access, but which we couldn’t control. I would configure the firewall to allow just what was needed and drop everything else. The important point was that nothing on that network could access anything but the Internet. When possible, I limited the addresses that a device could access to what was required. Fairly run-of-the-mill firewall stuff.

      And then wireless happened. Users wanted access to the Internet and many internal services for their devices. We put wireless on its own network, allowed access to the Internet, but that’s it. That was fine, until users’ wireless devices started attacking one-another. The problem was that firewalls only apply to traffic travelling between networks. Traffic travelling between devices on the same network is a free-for-all. The solution was obvious: The firewall should behave as if each device was on its own network. It has a succinct name which I have forgotten. My recollection is that most wireless access point hardware supports this feature and that OpenWRT can expose it if the manufacturer’s firmware doesn’t.

      I really wish I could remember what that feature is called. But it is really nice as it allows one to define rules that allow fairly permissive access to the Internet, yet protect devices from one-another. In an Internet Of Things world, it will be a necessity.

      1. I’ve always enjoyed messing around with networking devices and even did a (very) little bit of VLAN work for a time. Unfortunately I lack a very detailed understanding of too many networking pieces. Setting up a well controlled wired network is one of the projects on my list for when I buy a house that:
        -has a basement for moderately loud networking/server/audio equipment
        -I can run wires through the walls without the apartment complex throwing a fit
        -I plan to stay in more permanently

        Until then I just live with what I can hack together on the weekends and doesn’t bother the family too much.

    3. I completely agree. I’m very paranoid about such things. I have a sandboxed network just for this reason. However…

      I also have a uRadMonitor (you can see me on the map above! I’m the one in Atlantic Canada) and I’m not worried about this device in particular. It’s a very simple little device and it moves almost no traffic across my network. Now no one can ever be 100% certain, but I’m reasonably confident that it’s benign. It’s just got a teensy tiny little microcontroller in there. It barfs out a very small packet once a minute to a central server. I think the source code is even kicking around somewhere if you want to have a look. I know I’ve seen parts of it (I asked for calibration data for the tube, Radu posted it to his website right away).

  2. Interesting San Paulo has twice the level of CPM as a location near me, yet only a few years back (in the middle of a drought and water restrictions) we were allowed to hose down our properties because there had been a dust storm and the source of the dust was an area where above ground nuclear tests had been carried out.

    Also would a set of graphs that showed the averaged values for each hour of the day be meaningful? Would it show particles originating from the sun and allow us to subtract them from our readings (treat them as background noise) in another graph to show other sources more clearly?

    1. Here is one explanation, and one potential problem with the design of the device.

      The above indicates that there would be location, seasonal and time-of-day variations in CPM due to that one natural source. So how do you cancel out such sources of noise? Also have lightning detector/locator networks and feed that data into a Kalman filter? Perhaps these radiations sensors need to be incorporated into the sensors used in this network?

      To measure radiation from just dust, fall-out, you need to use a dust filter on a continuously moving roll system and measure the filter as it passes near an otherwise completely shielded sensor. The airflow through the filter and the filter movement rate needs to be constant and calibrated too.

      1. I don’t think you could filter it out, unless you have a network of sensors covering an area the size of a state, then you could simply ignore sporadic events.

        A much more common event is gamma ray showers.

        1. If I get a click on a gamma-ray detector at the same time as I get a click on a broadband RF receiver I could flag it as suspect. They both travel at the same speed don’t they?

          Otherwise you have to use something like the other method I described, using a moving band of air filtering material. The advantage of the filter method is that the band can be time stamped and once passed the detector, and into it’s own shielded area, it is rolled up between two plastic layers so that you can go back and sample it later with a mass spectrometer to determine exactly what type of isotopes are in your fall-out.

  3. I met a guy here in Salt Lake City the other day who is creating his nown network of air quality sensors across Utah. He uses a nicer active sensor than what this guy is using. He showed me several instances where the numbers Utah is reporting to the EPA are completely innacurate (always in Utah’s favor) and where the data has been manipulated. Anyone can go look at the EPA’s numbers and see Utah reporting negative concentrations of PM2.5 and unrealistic patterns compared to other cities.

    I’m glad to see the availability of independent measuring networks increasing. We need to hold our government accountable and have accurate information about our air quality.

    1. Oregon does the same thing with their air quality stations. They consider extreme highs and lows as “abnormal” and remove selective “high” days and leave the “lows” in so that their numbers don’t freak people out. This always happens around field burning sessions when the farmers torch their fields. Some days a low pressure zone hits and entire communities have less then 1/4 mile visibility yet some how the air-quality says its just fine. When you can taste the air its not fine! In addition the air quality measurements for radiation are done with air samples that are on filter paper. There are stacks and stacks of samples from all around the state still waiting to be tested years later. So Oregon doesn’t have accurate radiation monitoring, certainly not real-time over most of the sampling stations they use. I am sure the cities like Portland and Salem have some real-time stations but I think a system of citizen scientists with this sort of tool would at least be able to cross check the “public” record and start holding people to higher standards.

  4. So why would anyone need this? Measuring background radiation is nice, but what do you need that data for?
    Fukishima was a strange event wrt to fears: instead of fearing the cause of ~20.000 deaths (tsunami/earthquake), people seem to fear that thing that killed no-one at all (radiation). No offence to the maker that went after this market though, fools and money are easily parted after all.
    If you want a sensor that really saves lives: a speedometer that is set 10k too fast (5 above the usual ‘error’) will easily do that. Or just remove the airbags and replace them with a few visible spikes.

    1. You think radiation doesn’t kill people? What?

      And the governments almost always hide reality. When Chernobyl blew up for instance it was the swedes who reported it through their detectors, while the rest of the world and all authorities of all countries tried to hide it. And there are countless examples. The same in Japan, if people didn’t have access to their own detectors we’d not even know there was a ‘small incident’ today.

      But of course this project will only help with hiding it, since the management is on-board with censoring data.

      But you are a person who thinks radiation doesn’t harm people so you probably won’t even understand what I wrote. oh well. Just call me an asshole to fit in with the good citizens and go your jolly way I’d say.

        1. Correct methinks.

          But in this article’s thread it’s hard to keep track you have to admit. But yeah, I realized it after a while but was tired of the whole shit and didn’t care anymore and went to sleep instead.

  5. “Driving around his country he detected several big radiation spikes. This data has been withheld from the reporting network. He believes he might have detected areas in his country where uranium is being enriched”
    Actually what I said in the presentation is that I ran into some areas rich in uranium, like high content of minerals in soil :) And that could have (or not) something to do with national resources.

    Also, I saw there was some concern about some Google components that I am using: regarding Google Analytics, I never thought using it could alarm anyone, for me it’s just a simple web tool and I use it to understand traffic and interest, quite a normal thing to do, to help the project grow. But I’m happy to know about better tools for that.

    Also, there is no Google Maps on uRADMonitor portal, what I am using is OpenLayers.

    Thanks for the comments and feedback, let’s try to channel it to something constructive, as I said , I take any good suggestions if there are better alternatives to some of the tools I am using.

    1. Hi Radu,

      Sorry about the “enriching uranium” mix up… I misunderstood what you were saying. I’ve updated that in the post.

      As far as [Whatnot’s] comments, this had enough people click the ‘Report comment’ link that it was kicked back to moderation and I deleted it. Sorry you had to see all that negativity here. It was a great talk and it’s a wonderful project.

      1. Thanks Mike. Indeed it sounds so similar :)
        Don’t worry about negativity, I’m not taking it personal in any way, instead just focusing on learning what I can improve from the various feedback.

        1. Glad you understand what I mean when I use a term like ‘fuck that guy’, which is “I personally reject that person’s way of doing things’ in the ‘internet shorthand’. (and as we now also in personal political communication shorthand from ambassadors and the like. As well as communications from CEO’s and bankers. Which I ad to prevent the old ‘we must blame internet manners’ nonsense.)

      2. Yeah right, censor data, censor critique, all to appease the fucking morons..
        Or is it to appease google who also runs their stuff on this site?

        Either way I hope you are proud of yourself.

        Meanwhile that site STILL has google/cloudflare/googleapis/google-analytics/doubleclick running on it, and I think that still ruins the project and makes me say fuck that guy.

      3. And we have to believe there’s some valid reason to hide observed traces of natural resources? And that that is the only thing censored out of the data?
        Luckily the crowd here has degraded their mental capabilities enough that they might buy it.

  6. About the not-released-data. That part of the data was an experimental setup, for the unit which was hooked up to my car, so that was a mobile detector, asserting various locations over a wide area of the country. That data is not part of the uRADMonitor network, its purpose was to test various sensors in a field test. I wrote about it on my blog:

    And the subject was also featured on hackaday here:

    Hope this clarifies any remaining issues.

    1. At least that 3rd try explanation finally makes sense somewhat.Combining moving data with a stationary network might not be a good way to run things. And lead to messed up interpretations.
      Glad to see you at least address it and went through the trouble fixing that confusion. Where I assume you are telling the truth, based on it making sense (finally).

  7. “The owner’s daughter had been conducting school experiments using depleted uranium”
    I don’t know what country this person is in, but possession of isotopically processed source material requires NRC licensing in the US per NRC 10 CFR 40. Unless of course this child happens to own a company, then she may posses 7Kg of the solid material.
    Perhaps the school owns the material?
    Isotopically pure Uranium has very low activity, due to it’s incredibly long half life. So the explanation of “and the sensitive equipment registered a higher background radiation level” sounds dubious to me. The sensor would need to be sitting directly on top of the material. And the Geiger tube he is using has poor sensitivity to alpha particles, so any events detected would be due to daughter nucleotides.

    1. I think he is in Romania, so CNCAN law #’s 14, 111, 353, and 366. And their “norms” as they are called- NSR, NSN, NPF, NTR, NDR, and NRN would all apply. The laws are very similar to US law.

      1. So if you pick up a bullet from a heavy gun like on the Bradley (BFV) or A10 you are suddenly an international nuclear criminal? Interesting. And I’m surprised they didn’t use that to justify some war yet. Just strafe with an A!0, then say you spotted depleted uranium in the enemy’s possession, and bam, you got your approval fro a full scale war.

        1. Yes. They are illegal to posses in the US without license from the NRC. If you are part of a government group that uses the material, then you are covered by their license while on the job.

  8. “Radu has noticed that turning on the equipment immediately after it has been on an airplane registers a spike in radiation from more cosmic rays in the upper atmosphere”

    More likely the geiger tube or something in the counter was not happy with the flight. Cosmic radiation does not linger. Something else is creating the spike in counts when you get back.

    1. Agreed. And because a tube running in Geiger avalanche mode is extremely sensitive to the voltage of it’s power supply, the more likely cause is the high voltage supply is producing a higher voltage immediately after the flight.
      And because the cargo hold of an airplane is not well temperature regulated, this is the most likely culprit.

      1. But can’t the radiation from cosmic sources enervate natural occurring material to temporarily have an increased output? Which would then be detected..
        Would be interesting to get to the bottom of this.

          1. We know that, but the article says “turning on the equipment immediately after it has been on an airplane”.
            Which says the radiation was noticed after it was on the ground again. Or is this also a mix-up and did it just record the data while in flight? Because we are wondering why it would notice more when back on the ground again.

        1. Particles of sufficient energy to cause transmutation are quite rare even at 37,000 feet of altitude. The detector he is using is a hollow tube with a rarefied atmosphere of quench gas. It is highly unlikely to capture a charged particle in its walls that would cause the required transmutation.

  9. You are an amazing designer and an inspiration. I followed your project from the early stages of HAD Prize and had you pegged to win for sure as your tech, skills and documentation are amazing. Glad you are keeping momentum and moving forward towards progress.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s