julianYorke July 7, 2005 9:48 AM

I wonder what the false positive rate is? seems like it would be bad. Will they use this in airports now? Once this technology is known it seems that there will be hundreds of ways to get around it, but maybe if used in conjunction with other security it could be helpful.

Francois Kashy July 7, 2005 12:04 PM

@Philip and Julian, unfortunately, there isn’t enough information to begin to answer those questions. Would anyone like to summarize the nature of this millimeter-wave system for us non-physics types?

Probitas July 7, 2005 12:48 PM

Sure, it can detect guns, knives or bombs, whether they are made of metal, plastic or composite. But can it slice and dice as well? Is it both a dessert topping and a floor wax?

I am guessing the world will never really know.

Ari Heikkinen July 7, 2005 1:06 PM

It’s interesting to note when someone introduces a security measure that’s effective people are often worried it’s too intrusive. Then when someone introduces a security measure that’s less intrusive people are worried it’s not effecient. There’s no such thing as security without tradeoff.

Davi Ottenheimer July 7, 2005 1:27 PM

Thanks for the tip. The New York Times has a good write-up, republished in “Newszine”:

A quick search found an article from 1997 extolling the virtues and relevance to airport screening:

“Applications that make use of penetration of solids include concealed weapon detection for airport security-Smith says that tests in this area have detected plastic as well as metal weapons through clothing and even through 0.5 in. of sheetrock.”

It looks like the US Justice Department has been financially backing the development of the technology to get enhanced sensors.

The NYT article says US$7million was invested over eight years, but another article claims Lockheed Martin used a $200 million grant from the National Institute of Justice:

This article also explains that a millimeter wave sensor detects “blackbody radiation signatures” that are already present (as opposed to actively reflecting radiation off a subject). These are then processed and compared to a database of known weapons.

In effect, this means any surveillance camera can be significantly more useful as it can provide visual alerts directly related to radiation signatures.

Amazingingly, even though Brijot will be selling each unit for $60K/unit they already claim $100 million in orders from around the world.

The questions that come to mind are what is the accuracy of the signature algorithm, how complete is the database of signatures, and how will they be secured/maintained?

Phillip Hofmeister July 7, 2005 8:50 PM

If someone broke a gun up in to it’s componants and spread them accros say 2 or 3 (or even 1) person. Would it still recognize the gun?

Clive Robinson July 11, 2005 6:53 AM


The answer is a little complex, but broadly yes any black body system can be trained to recognise component parts if it’s resolving charecteristics are suitable.

To grossly over simplify, asume it measures the density of an object, even individual gun components will have a similar density, and the main parts (breach block barrell) will need to remain in one piece.

Even if the resolving power is low, the mean density in a given area will increase if high density objects are there.

A sensible policy would be to use this and similar technology to direct people into different search lines. If you appear to have high density objects go into line B where, another slightly more intrusive detector is used. If that registers then move to line C. You use progressivly more intrusive / labour intensive methods untill the object is clearly identified.

Obviously people in line A get moved through fairly quickly and with less inconveniance. You would however need to randomly select between 10-30% of people over and above those detected to reduce the likley hood of targeting issues (both by terorists trying to avoid, and people beleving they are being singled out/victamized).

Leave a comment


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

Sidebar photo of Bruce Schneier by Joe MacInnis.