Schneier on Security
A blog covering security and security technology.
« Friday Squid Blogging: Contaminated Squids |
| Liquid Ban Gone Weird »
June 23, 2008
New Technology to Detect Chemical, Biological, and Explosive Agents
"We have found we can potentially detect an incredibly small quantity of material, as small as one dust-speck-sized particle weighing one trillionth of a gram, on an individual's clothing or baggage," Farquar said. "This is important because if a person handles explosives they are likely to have some remaining residue."
Using a system they call Single-Particle Aerosol Mass Spectrometry, or SPAMS, the Livermore scientists already have developed and tested the technology for detecting chemical and biological agents.
The new research expands SPAMS' capabilities to include several types of explosives that have been used worldwide in improvised explosive devices and other terrorist attacks.
"SPAMS is a sensitive, specific, potential option for airport and baggage screening," Farquar said. "The ability of the SPAMS technology to determine the identity of a single particle could be a valuable asset when the target analyte is dangerous in small quantities or has no legal reason for being present in an environment."
Posted on June 23, 2008 at 6:07 AM
• 46 Comments
To receive these entries once a month by e-mail, sign up for the Crypto-Gram Newsletter.
So now the false positives will go way up. It is already the case that small amounts of gun powder on bags used sports shooters or hunters set off detectors at airports. (hint: don't let your range bag double as a luggage bag) Now dectectors will be set off by things like dried benzoyl peroxide residue on the skin of pimply faced teenagers. I feel much safer knowing that one trillionth of a gram of an explosive can't get through security.
What if I sat in the seat where the previous person had contact with explosives. I guess I will have hard time to explain how the particles got on me.
Sensitivity is fine, but it needs to be coupled with specificity. If the system detects, for example, nitrogen radicals, then it could confuse household ammonia as belonging in the class 'nitrogenous explosives'.
Detecting a picogram of cleaning fluid would be an exercise in futility.
Read the fine print on shampoo bottles. They commonly contain ammonia compounds, and head hair makes a fine wick to disperse aromatics.
Sorry, useless comment, but: can society just stop with the clever contrived acronyms already? "SPAMS." Give me a break.
I agree. How will I prove to officials that the single particle they lifted from me was due to secondary transfer?
If it is not coupled with other indicators then used on its own - it is a bad thing.
@@ Moderator / Bruce,
Is there a problem with the new comments page?
It's just that I have looked there and none of the above posts have made it onto the 100 new comments page.
So, now anyone can literally stop an airport just spraying little amounts of chemical residues on random people?
I guess I shouldn't shake the hand of a friend of mine who is a demolitions expert in the Army.
But if the test is that sensitive, I guess no-one should shake my hand after I've visited my friend (unless i've had a shower).
It would be interesting how many handshakes it takes before you can't detect any trace.
This is as useful as testing dollar bills for cocaine.
Its a mad world.
Dunno why you're uncritically posting this press release in your blog. SPAMS (perhaps named for all the false positives it will generate?) is exactly the sort of Homeland Security funded idiocy we should all condemn - you most of all. I appreciate that the researchers need to get funding from somewhere to continue their work, and that Homeland Security money is comparatively easy to get (and has a very low bar - in my experience - for demonstrating practical utility) - but the work as described has little/no serious airport security application.
Anecdotal, single data point FYI regarding current sensing: My work "briefcase," carryon and range bag are very often one and the same (a TT 3-day pack). I have, on more than one occasion, used the pack as a rifle rest, then boarded a plane with it as a carryon within 48 hours.
Never gotten a hit off the current nitrate swab tests. Seen them run, and nothing comes back, ever.
I guess the moral of this one is that soon it may be unwise to attempt to fly on a commercial airline (or do anything else that 'SPAMS' is being used to monitor) if you live downwind of a quarry, a military base, a gun-club or similar, sundry industrial premises or even, as I used to do, if you live in a house that was formerly let to workers at a munitions factory.
Railway technical and operations staff could have great fun explaining why they're liberally tagged with explosive residue from the fog-signals that are routinely carried and used on railways around the world. Ditto for numerous other quite innocuous occupations that bring one into contact with small amounts of explosives and related substances.
And, also, it will apparently, work for minute amounts of biological agents: that should get quite a few people concerned, from doctors and nurses through to garbage collectors and the like.
'SPAMS' sounds as if it may be a useful possible addition to the armoury of forensic examiners and other *after-the-fact* investigative professionals.
However, it appears to me to be neither safe nor ethical to use the technology as a tool for screening suspects from the population at random, which is what 'SPAMS' seems to be being pitched at in the LLNL press release that Bruce linked to.
This reminds me of the DNA profiling method called Low Copy Number (LCN), which allows a DNA profile to be produced from tiny traces of material at a crime scene -- as little as one or two cells, according to its advocates.
It's susceptible to contamination from a range of sources. For one thing, we all slough off a rain of dead skin cells as we go through life, so we leave traces of our DNA all over the place.
For another, forensic labs are simply not geared up to the extremely stringent precautions against cross-contamination that must be observed when using techniques as sensitive as LCN and SPAMS.
Alas, judges and juries are appallingly ignorant of these caveats, and prosecution lawyers are all too willing to suppress evidence that would call the reliability of these methods into question, so there is a huge potential for serious miscarriages of justice.
So, when will LLNL receive the C&D from Hormel?
@ Colossal Squid
Thanks for posting that link. From a very quick read it appears the system has been invented to monitor air for biological agents, but can also be used to detect other aerosols. At least it was tested that way. From this it is also easy to understand how it can be used to scan baggage and passengers.
What I don't understand is how it should lower false positive rate, indeed I agree with most other comments here that this system is probably not useful this way. The article you link to claims to make few errors in determining a substance. It can't however detect why a substance is present on the subject (passenger or bag). If that person was at a fireworks display a couple of days earlier, it would correctly detect explosives. Nevertheless, the person would be entirely harmless.
... "we can potentially detect an incredibly small quantity of material"
So then the real issue is how to find THAT particle !! And if THAT particle was detected (with little evidence of many others), what is the likelihood of it being deniable ??
"the target analyte is dangerous in small quantities or has no legal reason for being present in an environment."
So what's the smallest amount of explosive for which possession is a crime? Could a passenger with a single particle of TNT on their clothing, perhaps from someone else's medication, be prosecuted for carrying a (very small) bomb?
If you stick to your guns, there's no such thing as a false positive.
The problem is not the detector itself; it's what whoever paid for that thing does with the information.
I'd think a significant proportion of the population would always test positive; anyone who has recently been near fertilizer will test positive for nitrates and chlorates. I think many substances in hospitals contain them as well. Some heart patients even get nitroglycerin pills.
For like a month or so after new year, or the 4th of july in the US, most people would still have detectable traces of various (low-order) explosives on them.
The question is if a terrorist that has tried to minimize contamination, has a larger "explosives signature" than anyone who has been contaminated by accident. If not, this means the signal you are trying to measure is below the noise level, and there is no way this is going to work.
As has been said b4 there is a problem with a test that is too sensitive with respect to false positives.
However this is not new. DNA is like this. The CSI shows are very misleading as to what some DNA evidence means. Family members for example will have each others DNA all over them.
Funny, the thing that immediately leaped to mind was the case of the Birmingham 6, where one of the big indications of "guilt" was traces of nitrocellulose.
And that was 1970's technology. Detection sensitivity has gotten much, much better over the decades.
Lives ruined with false positives, while U wait!
A quick read of the paper suggests that specificity may not be nearly as much of a problem as sensitivity -- although the thing calls itself a mass spec, it doesn't do what we think of as chemical analysis at all. There's some fluorescence to see what surface compounds the particle contains, and then the MS part does a charge-to-mass ratio not on compounds but on the whole damn particle. So it will detect (in very small amounts) either threat particles made in just the way that exemplars were made in the lab for training or else a few things that look like them. It won't detect traces with a different size or shape distribution -- or threat particles someone didn't think of producing in the lab.
There's a telling line about possible false alarms toward the end of the paper, where a couple of particles that produced alarms are found not to be spores, but shapes similar to clumps that their anthrax spores very occasionally produced during lab training. So they simply removed those profiles from the database and the false alarm problem went away. Um.
That is extremely bad science; everyone knows that the datasets for formulating the hypothesis (these are evil particles) and testing the hypothesis should never be the same set.
But than again, that's just what the TSA does, isn't it? They show various pictures of bomb-like things someone made over and over again when training their screeners, and later on, they show the same pictures, injected into the x-ray machine, to verify that the screeners still recognize those pictures as bombs.
Who cares if they can recognize real-word threads, as long as the recognize the examples.
@c4, @Waldo, @Fabio, @Al, @Zippy, @SteveJ:
Sounds like some of you might be thinking that a false positive leads to something devastating such as a detainment or conviction. I don't think that is the case.
I used to fly on a regular basis and have been stopped 3 times after my bag was swabbed. They just opened my luggage and did a search by hand. I asked the screener and he said he gets false positives all day long. They didn't detain me or search me personally and it was rather routine and non-intrusive. Took maybe an extra 2-3 minutes to get through.
Who knows how many times they swabbed my bag after I checked it and then they searched it. Possibly very many times given that my bags apparently have something all over them that sets off the alarm and I often got notices that my bag had been searched after I checked it.
There could very well be many thousands of false positives per day in the states, but it is business as usual.
I bet the people who develop & sell SPAMS have a hard time getting emails to go through...
Rights and freedoms are often lost by stages. Stage one, they detect the smallest possible residue of explosives, but they don't detain or charge. Then later, step two, they start detaining and charging.
@Grumpy Physicist: "...one of the big indications of "guilt" was traces of nitrocellulose."
Yeah, maybe the defendant was simply playing an old Fender electric guitar! Those have nitrocellulose finishes.
Or he was playing ping-pong; the balls are made of NC.
When everyone is jailed, there won't be any more lines at the airport...
"Who cares if they can recognize real-word threads, as long as the recognize the examples."
Remember that TSA Seems to miss the examples with regularity, and they are still in business. So I'm not even sure THIS bar is low enough.
If memory serves, the nitro-cellulose traces that were one of the mainstays in the wrongful conviction of the Birmingham Six were, in their appeal, said to have come not from kneading TNT as the prosecution had suggested, but from a pack of 'shiny' playing cards that the defendants had been using immediately prior to their arrest.
In the summer of 2003, my wife was traveling with some high speed film that had to receive the nitrate wipe test. The talkative TSA agent (yes, an oddity in himself) said that he hated to inspect golf clubs (which, apparently you could carry-on in 2003). The clubs would retain traces of fertilizer which is heavily used on many golf courses. Every club would register as positive, so they would allow all golf bags.
Finally, I can be finally arrested for having a single molecule of something on my clothing. Brilliant, I'm sure this will never lead to people being detained wrongfully.
Just one more reason not to fly home from vacation. :-)
Whenever I get a positive test from the "whiff of air" machine at Las Vegas airport it just means they do a hand search of my bag. I believe the machine is reacting to the past presence in my bag or clothing of either flash paper or fireworks.
I'm not sure we should even call this a "false positive". It's a true-but-useless positive. Yes, I've been in contact with explosive agents; so what?
I miss the days when you could bring fireworks back in your checked luggage and not worry about it.
I have serious doubts about how useful this new equipment is.
There are pretty good explosives detectors already in existence, and the guys in airport security usually turn down the threshold on them or turn them all the way off, leaving them as expensive theatrical props, because they detect too many people walking through who just fertilized their lawns. Sure, you can tell that someone has ammonium nitrate on them, but was it because they were making a bomb or was it because they fed their petunias? Many other explosives also have quite legitimate uses around the home, farm, office, etc. -- the whole thing is useless. The more sensitive they make the detectors, the more likely it is that the things will just go off every thirty seconds with a false positive.
What you're looking for, if you're trying to avoid false positives, is something that will reliably detect LARGE quantities of explosives, and only large quantities -- something that will reliably find concentrations of explosives large enough that they could take down a plane, and which will ignore so little that it could just be powder residue from someone's hunting trip. You want to ignore traces and find large concentrations. This whole drive towards insane sensitivity totally misses the security realities of the situation. False positives are the enemy. You want reliable INsensitivity. That's a much harder problem.
I am thinking of patenting homeopathic explosives detection.
The detection of the explosive would be increased the more the explosive was diluted. So if you washed off the explosive, you would have diluted the explosive so that no molecules of the original explosive exists, so it would immediately flag the user as a terrorist. (Because if you tried to wash it off, you must have something to hide.)
I should have entered this in the movie plot product category. It is just as rational as some of the weird crap that has already been sold to law enforcement. (Like the cocaine detector using the power of dowsing.)
We have already brought back spectral evidence, why not bring back sympathetic magic as well?
Indeed, the "last 100 comments" page wasn't updating properly. It should be now.
While the security staff are busy checking the bags of all the travellers who innocently came into contact with explosives somehow, a terrorist would make sure that he was not the person who packed his bomb, but would have a separate team pack his load, clean it and have it deposited in a clean environment for his collection, ensuring that he and his cargo are not carrying any traces of offending substances. Once again, you have security staff being directed in the wrong direction!
Brilliant stuff. Now all someone has to do to shut down an airport is spread very small amounts of residue over the seats of the local airport train.
If someone wanted to shut down an airport all they would need to do is contaminate some $1 notes with explosives and spend them at shops outside the security area. Anyone who receives one of those notes as change (or one that was next to it in the till) would get flagged.
As for detecting large quantities, any large quantity of explosives would presumably be well wrapped and packaged (maybe by a different person as suggested by gmrza) so that only a trace escapes.
The best way to reduce the risk of airline hijacking in the US would be to dramatically increase the quality of the train service. If there were trains travelling at over 200Mph between the most populous parts of the US for a reasonable price the number of airline flights would decrease significantly. The smaller the number of flights the easier it will be to secure them.
"the target analyte is dangerous in small quantities or has no legal reason for being present in an environment."
Does that include hydrogen sulphide? Also the " has no legal reason for being present in an environment" does not imply that the person on whom (or on who's property) whatever was found does not in itself imply that they have done anything wrong. There may be more environmental contamination than is expected.
Detecting a picogram of cleaning fluid would be an exercise in futility.
Though there have been issues with cleaners not being security checked, even being allowed airside without even going through the same checks as passengers and aircrew are subject to.
While the security staff are busy checking the bags of all the travellers who innocently came into contact with explosives somehow, a terrorist would make sure that he was not the person who packed his bomb, but would have a separate team pack his load, clean it and have it deposited in a clean environment for his collection, ensuring that he and his cargo are not carrying any traces of offending substances.
The "decoys" don't even have to be contaminated with any actual explosive. "Nitrate" is simply part of a great many compounds. Whilst alkali metal nitrates can be used to make explosives they might be difficult to distinguish from other metal salts which cannot be used to make explosives.
I see physical DOS attacks on airport security are getting more and more likely.
SPAMS? This got to be a joke? This can't be real happening. Those Livermore boffins got to be smoking something wierd?
My unqualified $0.02
A few years ago, the Ben & Jerry's ice cream people started a campaign, in the name of children's safety, to get the federal government to lower the amount of dioxin (30 ppm) and other toxins allowed in food, preferably to zero.
They went on and on about how "any amount is too much" for 6 months or so -- until JunkScience.com sent some Ben & Jerry's ice cream to a testing lab and proved that it contained more dioxin than the existing standard allowed. (I'm not saying that the ice cream was unsafe -- I'm saying the standard already goes way beyond what safety requires.) That was the end of the campaign.
I expect that this new gadget will lead to a series of similar embarassments that (hopefully) lead to realistic "threat threshholds" being set for a variety of substances it detects.
Let's be sure not to let the needed public debate about those levels be silenced by TSA in the name of national security. I'm sure they'll try.
Schneier.com is a personal website. Opinions expressed are not necessarily those of BT.