Bomb-Sniffing Mice

I was interviewed for this story on a mouse-powered explosives detector. Animal senses are better than any detection machine current technology can build, which makes it a good idea. But the challenges of using animals in this sort of situation are considerable. The neat thing about the technology profiled in the article, which the article didn't make as clear as I would have liked, is how far it goes in making the mice just another interchangeable part in the system. They're encased in cartridges, which can be swapped in and out of the system. They don't need regular handling:

Unlike dogs, which are often trained for explosives and drugs detection, mice don't require constant interaction with their trainers or treats to keep them motivated. As a result, they can live in comfortable cages with unlimited access to food and water. Each mouse would work two 4-hour shifts a day, and would have a working life of 18 months.

If we are ever going to see animals in a mass-produced system, it's going to look something like this.

Posted on February 9, 2011 at 11:39 AM • 42 Comments

Comments

AlexFebruary 9, 2011 12:03 PM

"false-alarm rate was less than 0.1 per cent"

LAX has 59,542,151 passengers per year. Each one of those will be screened either at LAX (departures) or another airport (arrivals). That means 59,000 false alarms per year. Assuming evenly distributed and allowing half of them to occur elsewhere, that's still 80 a day.

anonFebruary 9, 2011 12:09 PM

I don't really want to know what the TSA's plans are for these mice when someone smuggles explosives in a body cavity...

uk visaFebruary 9, 2011 12:17 PM

I'm not sure Peta is going to like the idea of mice stuffed in cartridges... but the prospect of watching the TSA being savaged by Peta does highlight how much fun can be had watching the bright humiliate the stupid!

Ken HaglerFebruary 9, 2011 12:19 PM

I think the lack of interaction would be considered a negative by the Gestapo. The general public has been thoroughly duped into believing in the infallibility of drug/bomb/whatever-sniffing dogs which really "alert" on command, which makes them an ideal cover for harassing people on a whim.

With a mouse locked away in a box somewhere, there's no opportunity for some Checkpoint Charlie to think, "Oh, this guy has long hair, he must have drugs," and surreptitiously signal it to "find" something as he would with a dog.

RockDoggyFebruary 9, 2011 1:09 PM

A couple of edits, and you have my IT job.

"They're encased in cubicles, which can be swapped in and out of the company."

Brandioch ConnerFebruary 9, 2011 1:10 PM

@Alex
"That means 59,000 false alarms per year. Assuming evenly distributed and allowing half of them to occur elsewhere, that's still 80 a day."

That shouldn't be too difficult to mitigate.

Just run all of the initial positives through a 2nd and 3rd screening (mousing?). Then investigate the ones that show two or three positive "hits".

kashmarekFebruary 9, 2011 1:13 PM

Until one of those rodents brings along or picks up a transmittable disease. Think plague.

NobodySpecialFebruary 9, 2011 1:25 PM

The article does a bit of marketing sleight of hand - it says the mice are more sensitive to explosive than the imaging scanners, which aren't supposed to detect explosive.

Sensitivity of explosive detectors isn't the problem - it's possible to make a detector that can pick up a single molecule of an explosive. The problem is using it in an airport where millions of people pass through - including construction workers, miners, gun owners ....

Ty SarnaFebruary 9, 2011 1:59 PM

What kind of interface do these cartridges have? Are they, perhaps, actual USB mice?

Davi OttenheimerFebruary 9, 2011 2:05 PM

"Unlike dogs...", that started to show a need for improved job benefits like time off, bathroom breaks, health-care...

Reminds me of the pre-American depression business model: why solve management problems when you can just invest instead in the promises of a new mousetrap?

Clive RobinsonFebruary 9, 2011 2:51 PM

"They're encased in cartridges, which can be swapped in and out of the system."

Hmm I reckon you could get a mouse in a 12guage so "lock and load" time for "squeak".

However there is one thing that is stretching it a bit,

"Each mouse would work two 4-hour shifts a day, and would have a working life of 18 months."

I've kept and breed quite a few "rodents" in my time including mice and as a general observation,

1, Mice are generaly not very active during the day.
2, Their actual life is on average 18months.

Which begs the question of how old they are when trained etc. Or more correctly it raises the question of if the person writing the document has actually ever been in close contact with mice and trained them etc...

February 9, 2011 3:11 PM

> Each mouse would work two 4-hour shifts a day

With an hour for lunch in between the shifts.

lazloFebruary 9, 2011 3:42 PM

@Brandioch Conner: That's assuming that the FP's are random, which I'd posit is a poor assumption. No one seems to make distinctions like this often, but there's a world of difference between a system that's got a 1% failure rate against 100% of the population vs a system that has a 100% failure rate against 1% of the population.

Nick PFebruary 9, 2011 4:11 PM

New Scientist is awesome. Now back to the topic. It sounds like a nice idea but i got a better one: drug sniffing mice for police departments. Much cheaper to manage than police dogs and pigs.

ShaneFebruary 9, 2011 4:46 PM

Lord.

If we're going this far, why not just arm every capable adult boarding a flight with edged weaponry? I'd like to see some underwear sparklers and box-cutter wielding zealots go up against a couple hundred armed American civilians with a self-interest in survival.

Well? Doesn't seem any less idiotic than putting disease ridden rodents that have historically wiped out a third of the world's population in every airport in the country, or bombarding every passenger with radiation and/or supervised molestation.

But oh how those big government contracts do indeed inspire the lab rats (no pun intended).

In other news, thank the stars, at least a temporary win for recuperating our constitutional rights - Patriot Act extension == DENIED.

MattFebruary 9, 2011 4:52 PM

"Is that a mouse in your pocket, or are you just happy to see me?"

"It's not a mouse, it is my pocket bomb detector."

Dirk PraetFebruary 9, 2011 5:24 PM

I'm actually quite curious if they have done any research into usage of common repellents such as citrus or red pepper smells. If for example the mice react positive to citronella, many people using an insect reppelent would create FP's. During summer, that could be a lot. But if they're trained not to, wouldn't there be a risk of them getting confused if someone is wearing both citronella and a bomb ?

Imperfect CitizenFebruary 9, 2011 6:02 PM

Cool idea. Unless you are Klondike Kat. "Savoir Faire is everywhere!"

Re: Patriot Act extension denial maybe this nightmare will end for my family? I hope so for my son's future.

Davi OttenheimerFebruary 9, 2011 7:43 PM

@Dirk Praet

good question. i have been wondering if cat owners would cause more false positives or less.

...was just discussing the other day how wolf piss smell has been used to covertly undermine illegal hunting and protect game.

Clive RobinsonFebruary 9, 2011 10:22 PM

@ Davi,

"...discussing the other day how wolf piss smell has been used to covertly undermine illegal hunting..."

Not just "piss" but spraint as well. Most animals that are hunted by preditors react to the various chemicals emitted / excreted by the preditors.

Oddly this is true for some animals that have never encountered the preditor (such as "lab mice" and "cats"). Thus it appears to be a built in response similar to the human reaction to "sour" or "bitter" taste (which broadly indicate poison).

This appears so effective that some zoo's actually sell on the waste some of their big cat exhibits produce. For instance lion excrement is used very successfully against deer and domestic cats. So much so that an artificial version of the "lion poo" smell is sold in a repelent called "silent roar" for gardeners in the UK.

Nick NFebruary 10, 2011 12:32 AM

"false-alarm rate was less than 0.1 per cent"

"LAX has 59,542,151 passengers per year. Each one of those will be screened either at LAX (departures) or another airport (arrivals). That means 59,000 false alarms per year. Assuming evenly distributed and allowing half of them to occur elsewhere, that's still 80 a day."

@Alex, the 0.1% figure is obviously the lowest/closest accuracy figure they determine with, I presume, a little over 1000 people. It seems like they only had one false positive out of ~1000, (1/1000)*100=0.1%

They would need a larger sample to determine a more accurate false positive rate.

Peter A.February 10, 2011 4:36 AM

Mice in cartridges... My goodness! Is this a LARP in "The Flintstones" setting?

Dinosors! do it with dinosors! that'll be massively COOL! Just imagine a velociraptor handled by a TSO dressed up in Fred Flintstone-style tattered rags armed with a flint-studded heavy club walking the isles in the waiting hall... precious. And it will scare the sh*t out of the bloody tirrorists!

Ok, my imagination just bolted. Personally, I would do it with hamsters. Hamsters are cool. No cartridges needed, just some duct tape...

Adam TrickettFebruary 10, 2011 9:52 AM

It works better with insects:
* they are small
* they are cheap
* fast generation time
* quick to train
* very little "animal rights" to worry about

By better half did this work for DARPA over a decade ago. The University of Georgia even took out a patent, US Pat # 6,919,202.

JoeFebruary 10, 2011 10:44 AM

For those worried about disease.

Mice bred and kept in a controlled setting are not carriers of disease. Think of your local pet store. None of their rodents are carrying plague.

No one is suggesting the use of wild rodents in this scheme.

BacopaFebruary 10, 2011 1:44 PM

I have a lot of experience handling rodents for the pet trade and I have to agree that expecting 18 months out of a mouse is a bit much. Also, mice stink. They are the dirtiest of all the rodents. We can joke about gerbils, but they are much cleaner and could easily work for 18 months with plenty of time for training and a long retirement thereafter.

As for cat owners getting false positives, there may be some ways around that. Rodents infected with toxoplasmosis are less likely to perform alert behaviors in response to cat urine odor. From what I understand toxoplasmosis has almost no negative effect on rodents, it's only a problem in humans, pigs, and other large mammals who are dead-end hosts. Alternatively, SSRI antidepressants also reduce rodents' alert response to cat urine odor. Maybe they could dose the mice with Prozac.

BTW, chances are quite good you are infected with toxoplasmosis even if you never owned a cat. If you are also Rh-negative, watch out.

ShaneFebruary 10, 2011 3:06 PM

@Bacopa - reinforces my point quite nicely - why are these new studies in 'security' becoming less and less concerned with exposing massive amounts of the public to disease and / or the possibility thereof? (Radiation, touching, swabbing, mice, et al).

The government justifies the scanners by stating that the risk of backscatter-induced cancer is nearly negligible, yet justifies their use of them to combat airplane related terrorism, which likely poses even less of a risk statistically than getting cancer from a scanner.

See any problems here? Now we want to add rodents to the mix?

Sounds to me like someone has sh!t bass-ackwards.

NobodySpecialFebruary 10, 2011 3:38 PM

@Bacopa - so your proposal is that the TSA should employee mice with brains damaged by a parasite or on drugs.
And who are going to do the job for 18months before retiring.

- Although I suppose it does sort of fit .....

JimMarch 27, 2011 12:41 AM

PeTA going after the TSA is not the bright humiliating the stupid; it's stupid vs. stupid all the way.

Whoever wins, humanity gains as a whole.

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 Co3 Systems, Inc..