quote:With one hand inserted into a sensor that monitors physical responses, the travelers used the other hand to answer questions on a touch screen about their plans. A machine measured biometric responses -- blood pressure, pulse and sweat levels -- that then were analyzed by software. The idea was to ferret out U.S. officials who were carrying out carefully constructed but make-believe terrorist missions.
GE Security CEO Louis Parker demonstrates some of the company's airport screening devices.The trial of the Israeli-developed system represents an effort by the U.S. Transportation Security Administration to determine whether technology can spot passengers who have "hostile intent." In effect, the screening system attempts to mechanize Israel's vaunted airport-security process by using algorithms, artificial-intelligence software and polygraph principles.
...
The company's goal is to prove it can catch at least 90% of potential saboteurs -- a 10% false-negative rate -- while inconveniencing just 4% of innocent travelers.
Were I writing the article, I'm not sure I'd put a "just" there. Let's say, hypothetically, that 10 terrorists and 1,000,000 legitimate passengers go through a certain airport in a month (by comparison, Toronto's Lester B. Pearson International Airport sees over 2,000,000 travellers per month). If the system meets its goal, it catches 9 terrorists and "inconveniences" 40,000 passengers. When a traveller is identified by the system as a potential threat, they would presumably be taken aside and interviewed by a real person. In my hypothetical example, that's 40,009 personal interviews in a month, or more than 1,300 daily. Assuming an 18-hour daily window of full-on airport operation, that's 74 personal interviews per hour. They had best be short interviews, even if there are 5 full-time trained and qualified people doing the interviews.
Let's make it marginally more realistic. There was an attempted hijacking at Toronto Pearson in 1998. To my knowledge, there has not been one since then, but I'll just take 1998. I'll lowball my estimate and assume that Pearson only served an average of 1,500,000 travellers per month that year, which makes for one terrorist and 18,000,000 legitimate travellers. Would needlessly interrogating 720,000 innocent people have been worth a 90% chance of preventing this hijacking?
As it was, the plane was stormed and the hijacker arrested. This is just a dummy example with made up numbers, but I think it's indicative of the problems inherent in trying to be the thought police when you can't actually read minds.
The other argument for this sort of thing is as a deterrent. I think that argument is stronger than the prevention argument, because I think airport security in general is much more useful as a deterrent than as a "last line of denfence." Presumably a system that was 90% successful at identifying terrorists would be a strong deterrent if it was applied across the board to all passengers.
Having said that, I'm pretty uncomfortable with the whole idea, and if I wasn't concerned about being arbitrarily deported to another country entirely, I'd consider answering the automated questions in Arabic to see if it resulted in me being flagged by the system.
Posted by TMedina (Member # 6649) on :
It sounds like a variant of the lie detector, working off of bio-feedback.
Which means this thing could trigger a false alarm for a hundred different reasons, the least of which being - oh wait, standing in line at the airport for hours on end.
How often do you get to produce a baseline from people willing, able and committed to blowing themselves up?
-Trevor
Edit: Any idea what's up with Sake?
Edit 2: Never mind, it just loaded - and logged me in, to boot.
Posted by Angiomorphism (Member # 8184) on :
I take it the other 6% of the "false negatives" that aren't inconvenienced are just nervous white people...
Posted by Dagonee (Member # 5818) on :
quote:Were I writing the article, I'm not sure I'd put a "just" there. Let's say, hypothetically, that 10 terrorists and 1,000,000 legitimate passengers go through a certain airport in a month (by comparison, Toronto's Lester B. Pearson International Airport sees over 2,000,000 travellers per month). If the system meets its goal, it catches 9 terrorists and "inconveniences" 40,000 passengers.
A perfect example of what's known as "the prosecutor's fallacy."
quote:I take it the other 6% of the "false negatives" that aren't inconvenienced are just nervous white people...
How are any false negatives inconvenienced?
The article isn't clear, but it seems to say the goal is a 10% false negative rate and a 4% false positive rate. What is this "other 6%"?
The Israel test had a 15% false negative and 8% false positive rate.
Posted by Nighthawk (Member # 4176) on :
And you're worried about traffic in Toronto? Try this in MIA, LAX, JFK...
It's a physical impossibility. Interviewers will outnumber the rest of the airport staff ten to one.
Toronto's *only* #22 internationally... JFK's #16, LAX #20, and you beat MIA at #31.
OK, so maybe that'll work in the US. Now look at Heathrow (61 million annually), Charles de Gaulle (49 million), etc...
And, is this a domestic or an international issue? Those numbers are even more dramatic; 85 million go through Atlanta every year.
Posted by Dan_raven (Member # 3383) on :
"And what do you plan to do with those who register positive." asked the reporter.
"Well," said the airport safety official. "There is only one thing we can do. Since they are, by definition, suicide bombers, we have to treat them like the suspected unexploded ordinance that they are."
"And that is..."
"Well, when we have a suspicious package, we take it to a clear, safe area near by, surround that area with detonating cord, and, boom, no more suspected terrorist."
"But...but...most of these people will be innocent people wrongly accused."
"Now, now" said the official. "We all must make some sacrifices for the safety of our homeland. Besides, the spectacle of exploding people and their luggage should keep the people stuck in line amused as they await thier turn."
Posted by Nighthawk (Member # 4176) on :
Yeah, because the one thing I need to remain calm while my biometrics are being scanned is to see the person before me explode on the tarmac.
Posted by TMedina (Member # 6649) on :
quote:When he said he had a short fuse, I had no idea...
-Trevor
Posted by Juxtapose (Member # 8837) on :
quote:Yeah, because the one thing I need to remain calm while my biometrics are being scanned is to see the person before me explode on the tarmac.
He said "amused." There were no promises made regarding calm.
Posted by twinky (Member # 693) on :
quote:Originally posted by Dagonee:
quote:Were I writing the article, I'm not sure I'd put a "just" there. Let's say, hypothetically, that 10 terrorists and 1,000,000 legitimate passengers go through a certain airport in a month (by comparison, Toronto's Lester B. Pearson International Airport sees over 2,000,000 travellers per month). If the system meets its goal, it catches 9 terrorists and "inconveniences" 40,000 passengers.
A perfect example of what's known as "the prosecutor's fallacy."
Are you sure? I didn't know that that particular misuse of conditional probabilities had a special name, but from the examples I've read I'm not quite sure I see how I'm committing it. It's been a long day, I'm pretty muddled, and I haven't tried to do stats in over two years, so it's entirely possible -- I just don't understand it.
A = person triggers a positive result from system B = person is a terrorist
If I'm not mistaken, their desired goal is twofold:
P(A|B) = 0.9 P(A|!B) = 0.04
Now, we don't know what P(A) or P(B) actually are, which is why I made them up by pulling numbers out of my ass. Once P(A) and P(B) are "known," can't I use P(A|B) and P(A|!B) to sort out how many people come up as legitimate positives and how many come up as false positives?
quote:And you're worried about traffic in Toronto?
I'm just using Toronto as an example, since on order of magnitude it's closeish to my dummy example.
Posted by Dagonee (Member # 5818) on :
quote:I didn't know that that particular misuse of conditional probabilities had a special name, but from the examples I've read I'm not quite sure I see how I'm committing it.
You're not committing it. You're describing it. Perfectly.
The people who think this is usable are committing it.
It was a compliment, dude. Posted by twinky (Member # 693) on :
Ahahahahahahahaha! Oh, man. As I said, it's been a long day, and I'm pretty muddled.
(Just take it as a compliment that I believed you and went and researched it, assuming I was wrong. )
Posted by Destineer (Member # 821) on :
It's also called the base-rate fallacy.
Posted by airmanfour (Member # 6111) on :
Polygraphs suck. And are unbelievably unreliable. All that's going to happen if these things are deployed is that a chapter entitled How To Fool a "Hostile Intent" Detector will be inserted into the terrorist handbook between directions for making an IED and avoiding being tagged by INS.
Posted by SoaPiNuReYe (Member # 9144) on :
I don't see the point in implementing such a device...
Posted by Dagonee (Member # 5818) on :
quote:Just take it as a compliment that I believed you and went and researched it, assuming I was wrong.
Yes, but my writing was capable of misinterpretation, so I am properly chastised. Had I written a complete sentence, no confusion would have occurred. Hoist upon my lazy petard! Posted by AvidReader (Member # 6007) on :
If I have to show up two hours early again and they won't let me have a bottle of water on the plane, you can bet I'll be one of those testing as a terrorist.
The England hooplah didn't happen until after I was already in Indianapolis on vacation, so I had to deal with the crap coming home. I didn't feel safe. For one thing, the gate checker didn't recheck my purse or coats and barely glanced at the bags. Then our checked luggage turned back up at the gate when it wasn't supposed to. God forbid we actually did have something we weren't supposed to.
It felt like jumping through hoops so they could say, "Look at us. We're doing something." I was certainly feeling hostile when I went through all that for sloppy security.
Posted by rivka (Member # 4859) on :
quote:Originally posted by AvidReader: If I have to show up two hours early again and they won't let me have a bottle of water on the plane, you can bet I'll be one of those testing as a terrorist.
So true.
Posted by human_2.0 (Member # 6006) on :
Some of the best reading I've done about airport security and what a joke it is:
"We can't keep weapons out of prisons. How can we ever hope to keep them out of airports?" - Bruce Schneier
"We can't keep weapons out of prisons; we can't possibly keep them off airplanes." - Bruce Schneier
quote:Originally posted by AvidReader: If I have to show up two hours early again and they won't let me have a bottle of water on the plane, you can bet I'll be one of those testing as a terrorist.
quote:It felt like jumping through hoops so they could say, "Look at us. We're doing something." I was certainly feeling hostile when I went through all that for sloppy security.
I was in San Francisco last Friday and when the little voice said we were at code Orange I actually yelled "PANIC!". Well, not too loud, I knew they would probably arrest me.
Posted by dkw (Member # 3264) on :
quote:The idea was to ferret out U.S. officials who were carrying out carefully constructed but make-believe terrorist missions.
Would fake "terrorists" in a training exercise have bio-feedback results even remotely similar to actual terrorists? Isn't the whole idea behind measuring physical responses that they are something that you (supposedly) can't fake?
Posted by rivka (Member # 4859) on :
I wondered that as well, dkw.
Posted by Angiomorphism (Member # 8184) on :
If you're feeling particularly sneaky, don't fly!
Posted by human_2.0 (Member # 6006) on :
Or just sneak on board Posted by BlackBlade (Member # 8376) on :
quote:Originally posted by human_2.0: Or just sneak on board
I read that too a day or two ago, I'd love to see how the boy managed it. But then again the lay out of each airport is different.
This reminds me of when my dad interviewd for the CIA he was so nervous he could not even state his name without the polygraph saying otherwise.
Maybe my own father isnt who he says he is
I was once in an applied physics class with a teacher I very much enjoyed learning from. He asked me about a homework assignment that I had obviously failed to turn in. I insisted that I had turned it in just to see if I could get him to believe me. He was adament that I had not, I said, "I will not admit to failing to turn in the assignment until I have taken a polygraph test" my teacher with a glimmer in his eye said, "alright..."
He went to the storage closet and pulled out a machine that measured electrical resistance and attached a finger sensor. He explained that if the electrical resistance stayed the same, I was telling the truth, if it dropped I was lying.
He asked me to confirm my name, him as my teacher, and that I had handed in the assignment. I am not particularly gifted a lying, but I passed questions 1-2 but utterly failed question 3.
I feel guilty the moment I see a police car with his siren on anywhere near me, even when I have done nothing! I could see myself as being one of those false positives.
Posted by human_2.0 (Member # 6006) on :
The eyes always give it away. The test they describe should have been done with accomplished liars from prison though.
Posted by Amanecer (Member # 4068) on :
BlackBlade- that is a rocking story!
Posted by human_2.0 (Member # 6006) on :
quote:Human beings are easily angered when they are kept waiting, e.g., in airline terminals, hospital emergency rooms, and heavy traffic. As adrenaline and noradrenaline levels rise, flyers, patients, and commuters may be more prone to aggression and violence than they are when permitted to move freely about. (N.B.: In England, more nurses are attacked in emergency departments than in psychiatric wards.)
I think putting hostile intent detectors in airports is going to have very bad results. They should actually try one as a test in a real airport just to see what results it gets.
Posted by pH (Member # 1350) on :
quote:Originally posted by rivka:
quote:Originally posted by AvidReader: If I have to show up two hours early again and they won't let me have a bottle of water on the plane, you can bet I'll be one of those testing as a terrorist.
So true.
Oh, yes.
I almost got INTO it with the airline woman when we went to check in our bags. Because I couldn't take Chapstick or EYEDROPS, despite the fact that I have eye allergy/dryness issues, especially on planes. She was like, "Just use them now." Lady, I am here two freaking hours early to jump through your hoops. I will not have access to my eyedrops for another five hours.
Then she and her coworker yelled at me for the entire time I was repacking. To which I replied, ")@#($&@, I am DOING WHAT YOU TOLD ME TO DO! Do you not SEE me doing it?!"
![[Smile]](smile.gif)
![[Wink]](wink.gif)
)
![[Dont Know]](graemlins/dontknow.gif)