FlyerTalk Forums - View Single Post - False Positives on Explosives
View Single Post
Old Sep 23, 2013 | 9:13 am
  #85  
exerda
Moderator, Omni, Omni/PR, Omni/Games, FlyerTalk Posting Legend
20 Countries Visited
1M
40 Nights
20 Years on Site
 
Join Date: Oct 2004
Location: Between DCA and IAD
Programs: UA 1K MM; Hilton Diamond
Posts: 72,600
Originally Posted by TSORon
These are not “false positives” but actual detection of the chemicals they were designed to detect.
They're false positives because the TSA treats them as an indication of explosives, and yet explosives are not present. We can argue about the way to resolve the resulting false positive "hit," but it is nonetheless a false positive the way the TSA reacts to it.


Originally Posted by TSORon
ETD machines cannot perform a complete chemical breakdown of the samples provided, they are just not designed to do that. They can only tell us when they detect what they are programmed to detect.
True, but they can give an overall signature rather than alarming on just components. The GC/MS machines I've used in a lab setting spit out a chart with peaks for different compounds and fragments, and the distribution of those peaks will correspond fairly well to the overall compounds present. Even back in the early 90s when I worked with that equipment, the computer attached to the GC/MS was able to analyze the overall signature and report on compounds matching the fragments detected in their proportions.

This would at least reduce the false positive rate without significantly compromising the ability to detect explosives.



Originally Posted by TSORon
If an alarm happens they we need to find out why, it’s as simple as that. That means a search, since lie detectors are somewhat unreliable.
And if one doesn't find anything in that search, that should be the end of it. Unfortunately, as I and others who have gone through it can attest, rather than admitting their sampling process or machine may be at fault, the TSA continues to treat the pax as a would-be terrorist, making threats, causing the pax to miss their flights, and of course bringing in LEOs for more intimidation.


Originally Posted by TSORon
Yes, its not really all that uncommon. Explosives are used not just by terrorists and the military but by quite a few different commercial concerns and police agencies. Hence the reason we dont make a big deal out of it, its about like finding a 1955 penny in a bucket full of pennies.
Several years ago when I "broke the ETD machine" (words by a TSO to me when it alarmed on my swab, but then also on just about everything else they tested), it wasn't a big deal, because the TSO let it drop there.

I think my much more recent experience (this year) of being detained in a back room with threats of "bad things" (calling in the FBI and missing my flight) with the constant refrain of, "You've got to understand it's just the times we live in," was certainly a big deal. And this of course after a complete dumping of all my possessions and of course finding absolutely nothing... yet the TSOs were beholden to their stupid (broken) ETD as if it had to be right and everything else their investigation showed them was wrong.


Originally Posted by TSORon
If we were to make a big deal out of incidents such as you describe what purpose would it serve? We verify that they are not carrying explosives and have no designs on destroying an aircraft or harming the folks on board, then its not really an incident. Therefore no need to make a big deal about it.
Like I said above, if only it worked that simply... maybe it even does at some checkpoints.


Originally Posted by TSORon
I’m not a chemist, but from what I have read they are many many different recipes for explosives, and none have a single compound.
But they all do have chemical signatures; when analyzed by ETD, there will be some ratio of each of the component parts which matches a known explosive. Yes, for improvised nitrate-based explosives, it's trickier (since you can successfully, to some degree or another, nitrate many common organic bases into an explosive compound)--but then, the ETD should be alarming on the combination of nitrates and those organic bases, not just on the bases themselves.

Also, the level of alarm should be different. Match a signature for a known explosive compound (not just the nitrates or an organic base), and one alarm goes off and requires a more detailed check. Match just the nitrate or an organic base (say, glycerol), and a lesser alarm goes off, requiring just an x-ray and possible hand search of items--not the full back room rigmarole.


Originally Posted by InkUnderNails
There may be some, and in the initial roll out of the technology there were likely many. However, what we are calling "false positives" are more than likely from one of three sources:
  1. Cross-contamination from poor handling practices for the TSO's
  2. Improper preparation or cleaning of the machine following testing of a known sample for calibration and the subsequent cross contamination
  3. Random alarms (if they will do it for the WTMD and Scanners, they will do it for the ETD.)

I have observed the first, but I only suspect the other two and I could be wrong.
I've observed the second on several occasions. One landed me in the back room grope-fest, and the TSOs couldn't understand why the ETD kept alarming. Well, duh, because you're not properly cleaning the machine! Try a different machine, with different TSOs performing the check. Once they finally did that, they let me go--but it took an hour to get them to that point.
exerda is offline