New Explosive Detection Tech

cruci writes to tell us Yahoo! is reporting that a New Zealand company, Syft, has developed a new way to detect many different kinds of explosives (and their individual ingredients) in real time. Designed for what the company calls "photocopier simplicity", CEO Geoff Peck claims that the technology is ready to deploy immediately and is already deployed in some ports and hospitals. From the article: "The Voice100(TM) employs Selected Ion Flow Tube - Mass Spectrometry (SIFT-MS). While SIFT-MS has been in academic use for more than 20 years, Syft Technologies is the first company to offer a commercial instrument with the full discriminating analytical power of a laboratory-grade mass spectrometer."

Comments

Tubes

[p0tat03]

"The Voice100(TM) employs Selected Ion Flow Tube..."

It's a bunch of tubes I tell you!

Re:Tubes

[andrewman327]

Certainly not a truck bomb!

Anyway, mass spectrometry is an interesting technology that works very very well in the lab. The question is how practical can they make this machine? How much does it cost? TFA talks about how terrorism is mega expesive, so I get the feeling that they are just trying to lessen the sticker shock. And as the saying goes, no matter how idiotproof they make the device, TSA will just make a better idiot.

Fortunately (according to the manufacturer) this machine finds more than your run of the mill explosives, it can also find drugs:

The instrument has been calibrated to identify narcotics, chemical warfare agents such as the nerve gas Sarin, toxic industrial chemicals, and peroxide-based explosives including TATP and HMTD, both used in the July 2005 London bombings.

Machine super-sensitivity: not "a good thing"

[Riding Spinners]

If you have these exquisitly sensitive machines that can detect even a few molecules of material, aren't they by the same token super-vulnerable to being attacked by "chaffing" or overloading?

You have to look at the false positive and negative rates for detection. If you have a test that is 99.99% specific, it will still fail in practical use in an airport, as that means that 1 out of 10,000 people will come up positive. If you have a lot of people going through you will still have a big problem (London had over a million flights last year). This is the same issue as using automatic detection of terrorists – It's one thing to match/no match a known ID (e.g. biometric passport) to a person; it's another to match every passer by to every known terrorist.

Going back to chemical detection: this level of sensitivity will mean that every person runs the risk of coming up positive eventually. This amounts probably about 100,000 people in the U.S., and lots more elsewhere in the world.

Re:Machine super-sensitivity: not "a good thing"

[Riding Spinners]

AddressException said:

OK - don't use the super sensitive machine and let *ONE* terrorist through.

Nobody has found terrorists at any point in history with chemical analysis machines, and they've been in use for years (they can't detect a ceramic knife). The incident at Heathrow was taken care of by good old-fashioned detective work.

Maybe you've lost your faith in the art of investigation, but I sure haven't. I have, however, lost my faith in having a civilized conversation with you on Slashdot. (mods: feel free to mod this down as "flamebait")

Re:Machine super-sensitivity: not "a good thing"

[Jere H]

Your math is off twice.
70,000 / 365 is 200, not 2,000, which doesn't really matter because:

70,000,000 * .01% is 7,000 searches per year, not 70,000.
So it would still be about 20 per day. They already do more random searches per day than this.

What's the point

[z0I!)]

IANAC (I am not a chemist) but this guy seems to make a pretty solid arguement: http://www.interesting-people.org/archives/interes ting-people/200608/msg00087.html

Alternate method

[nizo]

Try to light all suspicious materials on fire. Nice and cheap (all you need is some guy you pay minimum wage with a handheld lighter). For larger items, have a can of hairspray handy to use as a cheap flamethrower.

They have it backwards

[pmancini]

We should be detecting bombers not bombs. Bombs form a nearly endless variety. Bombers are an easier class of object to detect, I believe. The fact that the bombers try to hide the bombs on their person or in their carry on luggage suggests they they themselves don't fear the system's scruitiny. In the old days they had to figure out ways of getting the bomb on the aircraft without them being anywhere near it. How times have changed.