Detecting Nuclear Weapons Using the Cell Phone Network

Okay, this is clever:

Such a system could blanket the nation with millions of cell phones equipped with radiation sensors able to detect even light residues of radioactive material. Because cell phones already contain global positioning locators, the network of phones would serve as a tracking system, said physics professor Ephraim Fischbach. Fischbach is working with Jere Jenkins, director of Purdue's radiation laboratories within the School of Nuclear Engineering.

[...]

Tiny solid-state radiation sensors are commercially available. The detection system would require additional circuitry and would not add significant bulk to portable electronic products, Fischbach said.

I'm not convinced it's a good idea to deploy such a system, but I like the idea of piggy-backing a nationwide sensor network on top of our already existing cell phone infrastructure.

Posted on February 1, 2008 at 12:54 PM • 39 Comments

Comments

McGregorMortisFebruary 1, 2008 1:23 PM

Combine this idea with the previously-posted New York City proposal to require a license to operate a Geiger counter. What do you get?

In New York City, you'll require a license to operate a cell-phone.

KashmarekFebruary 1, 2008 1:28 PM

By the way, what makes anybody think that we won't be required to get a license or permit, to use the internet. Or to buy groceries, water, electricity...

Timmy303February 1, 2008 2:02 PM

A small GPS-enabled solid-state detection device would be cheap enough to produce in mass quantities. Maybe posting one on over a front entrance could be a pre-requisite for opening a business. Put some Federal airplane smoke detector tampering laws in place to protect them ...

drewFebruary 1, 2008 2:26 PM

Timmy

The problem with fixed sensors is that they'd require regular testing. Anyone could disable them within a given area for week long periods without anyone noticing. It'd be much more difficult for someone to disable everyone's cell phone sensor. Plus, you're talking about whole new infrastructure vs 'clever' piggybacking.

Carlo GrazianiFebruary 1, 2008 2:30 PM

Excellent. Now all a terrorist has to do is break open a smoke detector in a major landmark or office building or airport to cause a radiation panic and an evacuation.

DerekFebruary 1, 2008 2:41 PM

Installing sensors in police cars, taxis, ambulances, and municipal buses/public transit systems would also provide decent coverage.

anonymous canuckFebruary 1, 2008 2:50 PM

Okay, I admit it's clever. But a lot would need to be worked out. Like false alarms.

Now this seems to be different from massive video camera networks. But if the objectives aren't managed properly ...

Pat CahalanFebruary 1, 2008 3:05 PM

Carlo ->

I haven't read this particular writeup, but some designs for this sort of detection network take that into account; the numerical risk analysis incorporates signal tracking and other semantic factors.

Some blip on the detection network is rated low priority; a blip that appears near the edge of the network (say, at a port) and moves geographically through the network would be regarded as a higher probability of an actual threat.

AnonymousFebruary 1, 2008 3:06 PM

The article is pretty short on details: what kind of radiation is it detecting: gamma or alpha/beta? I'm assuming gamma: is it energy specific? Small volume means low counts and high noise.
"The researchers tested the system in November, demonstrating that it is capable of detecting a weak radiation source 15 feet from the sensors."
What was the integration time for the sensor? How weak is weak? Potassium & uranium in rock can give off quite a bit of (weak) radiation. Shielded gamma sources get downshifted to low X-rays. Cosmic rays would give a lot of bursts too.
I guess if you link it with GPS you could get a background "expected" radiation map built up, but this sounds like it would create a lot of ongoing false positives.

Timmy303February 1, 2008 3:07 PM

Nah I was being sarcastic, I guess I wasn't obvious enough in my dig about airline smoke detectors. In my idea it would be too easy to engineer false positives and false negatives, which actually create a much more easily exploited framework for a "cyberterrorist" scenario than actually smuggling nuclear material into the US and presents minimal risks to the terrorists.

Man I need to call Fox, 24 does NOT have to be delayed because of some stupid writer's strike, WE'RE ALL QUALIFIED FOR THAT JOB.

gawpFebruary 1, 2008 3:09 PM

The article is pretty short on details: what kind of radiation is it detecting: gamma or alpha/beta? I'm assuming gamma: is it energy specific? Small volume means low counts and high noise.
"The researchers tested the system in November, demonstrating that it is capable of detecting a weak radiation source 15 feet from the sensors."
What was the integration time for the sensor? How weak is weak? Potassium & uranium in rock can give off quite a bit of (weak) radiation. Shielded gamma sources get downshifted to low X-rays. Cosmic rays would give a lot of bursts too.
I guess if you link it with GPS you could get a background "expected" radiation map built up, but this sounds like it would create a lot of ongoing false positives.

Alan PorterFebruary 1, 2008 3:17 PM

Right... when I worked at (large Swedish cell phone manufacturer), we would put a HUGE engineering effort into saving a couple of pennies per phone. And now we're want to add on some sensors that generate no revenue...

ShriFebruary 1, 2008 3:28 PM

@anonymous canuck -
That brings up a clever extension. Collect images from the existing still and video cameras in the cell phones to create a video or still camera network. Turn on the mics and collect voice recordings. Use Seti@home cycles to process the data for suspect conversations. Collect GPS location data to determine the location of the perps. Book 'em, Dano!

Bill Higgins-- Beam JockeyFebruary 1, 2008 3:38 PM

1. Could cellphones help track giant city-sized cosmic-ray air showers, which are the result of highly energetic but rare particles? (Probably not, but maybe.) You could opt in and volunteer your phone for science, rather like SETI-at-Home.

The press release is vague about the detection technology, and I haven't yet found any scholarly publications from these folks about their scheme, so I can't do more than guess.

For a similar idea, see Quarknet, networked radiation detectors in high schools spread over a wide area: http://quarknet.fnal.gov/

2. What are the risks, and how might the system protect the privacy of cellphone users? False positives and radiation panics have already been mentioned. Is calibration a nightmare?

3. Does the cellphone customer get to see real-time radiation data from his own phone, or does it just go to a giant data-collection system? Does the customer get to see products of the GD-CS, such as a radiation map of his neighborhood?

anonymous canuckFebruary 1, 2008 4:23 PM

@Bill Higgings

Sure, when the phone detects something bad the network can text you ...

from bigbro: UR screwed m8

RoyFebruary 1, 2008 4:44 PM

Now think of how many cell phones travel through or near Radiology at every hospital in the country.

Think of how many people are walking around with radioisotopes in their bodies.

Think of how radioisotopes get delivered to Radiology. (Most of them move on Greyhound buses on surface streets and freeways.)

Now, how does the government deal with all these false positives without having its usual hissy fits?

Matt from CTFebruary 1, 2008 5:12 PM

While it may not be for general deployment, where this technology would fit in well and still have enough sensors to work well is the new 700mhz nationwide public safety 3G (or better...4G?) network.

www.cyrencall.com gives some information on it's nascent design.

While we don't need a tactical communications system that lets a cop in Boston talk to a firetruck in Washington, there is a lot of neat stuff that can be done vehicle-to-vehicle level if they truly deploy this system correctly.

Such as from firetrucks being able to use a license plate or VIN to pull up Vehicle Safety Datasheets detail locations and disarming procedures for hazards like high voltage electrical systems and airbags.

Or, related to security, tying unique vehicle identifiers to indicate authorized vehicles to enter a security perimeter. Rather then a mere ID card, a system that relies on the network to confirm security -- the Vehicle radio ID is unique on the network, the ID was authorized to respond, and the ID is where and with whom it is expected to be.

Rather then terrorism, the primary security problem isn't people wanting to do harm but well intentioned but incompetent and undisciplined freelancers who just decide to show up. If the vehicle isn't authorized through official channels, the radio network can immediately flag it to security posts.

Anyway, if we have a system that well built (a big hope, I know...it's a Federal project...) then piggybacking features like this should be easy.

AnonymousFebruary 2, 2008 4:26 AM

Wow. So many comments, and nobody so far has addressed the obvious:

Why look for something that has never ever been an issue before, and the chance of it happening is effectively zero?

ARE YOU FUCKING NUTS?

SteveJFebruary 2, 2008 4:46 AM

"professor Ephraim Fischbach ... working with Jere Jenkins"

Sorry, is this actually a comic book? :-)

FNORDFebruary 2, 2008 4:52 AM

Now, it claims that it can be "trained" to ignore "normal" sources of radiation. How does that work? Except most normal sources of radiation are also the tempting sources for dirty bombs.

You could do it by location, except that, as noted, the sources move around. The medical devices have to be transported. And "normal" sources are everywhere.

Aaron MuderickFebruary 2, 2008 7:52 AM

If you study radiation and radioactive materials you'll learn that the materials used in nuclear weapons are some of the most difficult to detect at a distance. Nuclear weapons aren't particularly radioactive and the energy they emit is easily blocked by modest shielding. A 'dirty bomb' may be harder to shield but it isn't impossible to do.

You will also learn that radioactivity is everywhere in our environment. Both natural and man-made sources exist in-place and are transported up and down our streets.

Don't believe me? Check google for the number of times that kitty litter sets off the radiation detectors at US ports. Check google for the number of times that medical patients set off the detectors at train stations and airports.

Or, lookup the facts our the latest detector for our ports. Despite the fact that cargo containers pass directly through the detector, it is only capable of finding a 10kg lump of enriched uranium with 50% accuracy.

A cell phone detection grid isn't a bad idea, per se. But this is more press-release based science in an attempt to secure more grant funding. Its a square peg in a round hole.

AxelFebruary 2, 2008 11:59 AM

If we want to reuse the cell phone infrastructure for national security: Why not use existing cell phone cameras in combination with face recognition software? May exceed today's hardware and software capabilities, but Moore's law will change that over the next few years. The chances of a terrorist walking past someone with a cell phone should be pretty high...

stevelaudigFebruary 2, 2008 5:27 PM

a next step, perhaps not the next step, will be drug detection and on and on and on

RogerFebruary 2, 2008 8:19 PM

Aaron is almost right. Simple solid state detectors such as Prof. Fischbach is suggesting here, have very low sensitivity and only moderate discrimination. They could conceivably detect the weak radiation from a fissile core, but only if they had a long integration time (i.e., were sat next to it for quite a while), it was very close by, and/or it was poorly shielded.

It is, however, perfectly possibly to create sensors (called Mössbauer sensors) that are so exquisitely sensitive to very precisely defined gamma ray energies that they can potentially detect small amounts of a particular fissile isotope, through quite thick shielding, from a vehicle racing past at speed. In fact these sensors are so exquisitely selective that -- even though the gamma rays are a nuclear, not chemical phenomenon -- they can often determine the chemical form or crystalline lattice of the fissile material from measurements of the line broadening, spectral shape, and so on. Thus, for example, a sensor of this type could actually be told to look only at gamma rays from metallic plutonium-239, ignoring those from plutonium oxide, and thus find a plutonium based fission weapon even if it was concealed in the midst of a pile of MOX fuel rods with tonnes of plutonium.

But there is no way you could fit one of these in a cell phone, either price-wise or size-wise. TTBMK the smallest available weigh about 400 g, and to get them that small you need to pay aerospace prices.

wsindaFebruary 4, 2008 2:48 AM

So many holes:

- If you use a sensor specific to a certain radioactive material, the attacker can use different stuff. If you use a non-specific sensor, there are many false positives. (The sensors are open to analysis by the attacker.)

- You need the cooperation of cell phone owners. (The attacker can spread rumors that the sensors also detect alcohol and marihuana.)

- The system will be much less effective in sparsely populated areas. And it won't detect material that is transported by air.

- Even if you detect the bomb, you may be too late to prevent the attack. (How long does it take to drive a van -- or an ambulance -- to the center of a city and push the red button?)

- How big really is the risk you're trying to mitigate? (Found any dirty bombs yet? Caught anyone with plans to use them?)

AndyFebruary 4, 2008 3:52 AM

What about all the legitimate uses of radioactive substances? Radiotherapy and medical imaging leap to mind - there you go, radioactive substances moving across your network.

I don't think that this is economic for the given threat. You'd save more lives many other ways for less money.

DaveFebruary 4, 2008 7:01 AM

This is smart because it isn't targeting a specific threat. It has more uses than just countering terrorism. What if there's a leak at the hospital ? These could help detect it. Inadequate shielding on one wall of the radiation lab ?

What about a trucker transporting radioactive materials irresponsibly who dropped one off the back of the truck ? Or one who got in a crash ?

As far as false positives are concerned, radiation isn't a binary value, it's a scale . Also, there are three different kinds of radiation, alpha, beta and gamma. The information from the phones would all be collected at a central point where the types and levels of radiation could be collated from multiple sources (lots of phones) and an intelligent decision made. The intelligent decision would be to send a guy out there with a Geiger counter to do a proper analysis.

Someone detecting small amounts of radiation in the radiology ward of a hospital isn't going to set off alarms. But someone detecting large amounts would.

paulFebruary 4, 2008 10:32 AM

Depending on just how what radiating stuff is around creating the false alarms, this could become a pretty effective ad hoc tracking system for the few of us who don't always carry cell phones...

NickFebruary 4, 2008 12:26 PM

My first reaction to the header "Detecting Nuclear Weapons Using the Cell Phone Network" was, wouldn't the mushroom cloud be enough of a giveaway?

But seriously, how much earlier are you going to detect and locate a quiet release of radiation by having 100,000 dinky, un-secure, unreliable sensors in people's pockets compared to having 100 better quality sensors at likely targets and in the hands of trained first responders?

Plus, once the public knows about the attack, your sensor network is going to start running away from the scene as fast as possible.

Bill Higgins-- Beam JockeyFebruary 4, 2008 2:09 PM

The number of users who would pick an OPTIONAL radiation-detector feature is presumably small. (Although it might be handy in my work!) The benefits of the Purdue scheme probably depend upon GPS and radiation detectors being present in very large numbers of phones.

One could imagine a government-required or government-subsidized program for radiation phones, if it's perceived to be desirable.

Cellphone designers are going to howl if we force them to add a gadget to every handset that doesn't enhance the functionality they're selling to the consumer. It's difficult enough to squeeze battery, display, controls, and all the phone chips into a tiny package.

Another point: this kind of research may be useful in incorporating other kinds of sensors in cellphones (or other ubiquitous devices), even if it never leads to a cellphone radiation-sensor network.

I will be interested to read more, when and if the Purdue group publishes something.

RogerFebruary 5, 2008 3:47 AM

@wsinda:
"If you use a sensor specific to a certain radioactive material, the attacker can use different stuff. If you use a non-specific sensor, there are many false positives."

Not really. If the device being detected is a fission weapon, then it will contain either uranium 235 or plutonium 239. A bomb based on other fissile materials is theoretically possible but is orders of magnitude less likely still -- and in any case would almost certainly be contaminated with traces of one of the above two isotopes.

If it is a radiological bomb, then indeed it could contain any of at least a dozen highly likely materials, and maybe several dozens of other less likely but still practicable isotopes. However in order for such a device to be significantly more effective than its burster charge alone, the activity needs to be quite large, making it much easier to detect.

"(Found any dirty bombs yet? Caught anyone with plans to use them?)"

Yes, at least 1 plot and 2 actual devices so far. UK citizen Dhiren Barot pleaded guilty to conspiring to carry out such an attack, although at the time of his arrest he had not obtained any radioactive material. Chechen terrorists have twice deployed actual devices, although neither functioned. The first was reported by them to Russian authorities without being armed, presumably as a propaganda exercise. The second was apparently meant to be detonate but was located and rendered safe before it functioned.

JMSFebruary 6, 2008 10:51 PM

> "(Found any dirty bombs yet? Caught
> anyone with plans to use them?)"

> Yes, at least 1 plot and 2 actual
> devices so far. UK citizen Dhiren Barot
> pleaded guilty to conspiring to carry
> out such an attack, although at the
> time of his arrest he had not obtained
> any radioactive material. Chechen
> terrorists have twice deployed actual
> devices, although neither functioned.
> The first was reported by them to
> Russian authorities without being
> armed, presumably as a propaganda
> exercise. The second was apparently
> meant to be detonate but was located
> and rendered safe before it functioned.

May I inquire what is the source of the latter bit of this information?

Anti-FederalistFebruary 7, 2008 9:20 PM

This idea stinks. I'm already a bit bothered that GPS chips are already in phones - supposedly, only activated on user request, or when E911 is involved.

Now, they want ANOTHER excuse to get access to GPS data?! Screw that.

@Bruce:
IMO, this is Security Theater(TM) incarnate.

@Kashmarek:
"Mount the detectors on top of cell phone towers."

I like that idea - it helps remove part of the privacy problem for end users.

@Timmy303:
Don't even get me started on the Writer's Strike. Their demands violate the basic principles of cyberspace - and they're being snobs, to boot.

@wsinda:
"- Even if you detect the bomb, you may be too late to prevent the attack. (How long does it take to drive a van -- or an ambulance -- to the center of a city and push the red button?)"

Worse yet - the detection may be of the radioactive plume, not the undetonated material.

@Roger:
"...through quite thick shielding..."

This was always my question about radiological material detection - in the unlikely event it be brought in (I personally believe the whole "terrorism" FUD to be an excuse to destroy civil liberties).

Because if someone DOES try this, you can bet your soul that they're going to shield it from here to Kingdom Come...

RogerFebruary 13, 2008 5:56 AM

@JMS:
"May I inquire what is the source of the latter bit of this information?"

Numerous. If you Google Argun+radioactive you will get several thousand hits, most of which are relevant. For example:
http://www.ahrq.gov/research/pedprep/pedchap6.htm

If you mean, what is the _primary_ source, then that is the Chechen security service. You may or may not regard them as a reliable source, but it should be noted that the previous device had extensive media coverage, and the Chechen rebel leader who called the press had his headquarters very close to where the second device was found.

Leave a comment

Allowed HTML: <a href="URL"> • <em> <cite> <i> • <strong> <b> • <sub> <sup> • <ul> <ol> <li> • <blockquote> <pre>

Photo of Bruce Schneier by Per Ervland.

Schneier on Security is a personal website. Opinions expressed are not necessarily those of IBM Resilient.