AIT Gumby Software - 50% failure rate - results in many quickie patdowns
#1
Original Poster
Join Date: Nov 2005
Location: on the path to perdition
Programs: Delta, United
Posts: 4,784
AIT Gumby Software - 50% failure rate - results in many quickie patdowns
Was going to post this in the other forum but I think it will turn into a debate so I will save the mods the trouble of moving it.
Executive summary: If going through an AIT that utilizes the new Gumby software I believe the false positive rate can be as high as 50% based on my very limited observations. As such, PAX should be prepared to receive a quickie patdown even if going through the AIT. Because so many PAX receive a quickie patdown some slip through without one.
Obervations:
While at DCA I decided to observe the AIT with Gumby software. Of 50 PAX who I observed going through, approximately 50% failed to get the all green. All but two of those failing got a quick patdown of the non conforming area. The two that did not get a quickie patdown slipped through the system because there were so many passengers failing that the two TSOs at times could not keep a proper accounting of who failed and who passed. At times there were three PAX past the AIT but not yet clear or a waiting a quickie patdown.
Because there were so many failures there was a male and female TSO present to do the patdowns. They swapped places on an as needed basis.
Secondary observations: TSOs did not change gloves. Women with dense frilly, afro, pony tails, etc. where subject to hair pat downs.
I did not count but some of the patdowns were a result of PAX failing to remove everything.
Conclusions: The efficacy software is questionable. Because of the poor efficacy a minimum of three TSOs are required. One entry TSO, a male and female TSO on exit. At DCA there was a forth TSO standing around doing nothing (who after 5 minutes eventually did my opt out patdown). As such, there is no reduction in work force with the new software.
Those PAX doing a opt out could wait 1 to 5 minutes for a response. In my case it was a solid 5 minutes - to the point I asked the tray TSO to secure my possessions as so many other people were going through that I could not secure them. To her credit she immediately responded, gather them and stood by them and even told another TSO she was busy. I thanked her for it once I got my pat down.
Discussion ...
Executive summary: If going through an AIT that utilizes the new Gumby software I believe the false positive rate can be as high as 50% based on my very limited observations. As such, PAX should be prepared to receive a quickie patdown even if going through the AIT. Because so many PAX receive a quickie patdown some slip through without one.
Obervations:
While at DCA I decided to observe the AIT with Gumby software. Of 50 PAX who I observed going through, approximately 50% failed to get the all green. All but two of those failing got a quick patdown of the non conforming area. The two that did not get a quickie patdown slipped through the system because there were so many passengers failing that the two TSOs at times could not keep a proper accounting of who failed and who passed. At times there were three PAX past the AIT but not yet clear or a waiting a quickie patdown.
Because there were so many failures there was a male and female TSO present to do the patdowns. They swapped places on an as needed basis.
Secondary observations: TSOs did not change gloves. Women with dense frilly, afro, pony tails, etc. where subject to hair pat downs.
I did not count but some of the patdowns were a result of PAX failing to remove everything.
Conclusions: The efficacy software is questionable. Because of the poor efficacy a minimum of three TSOs are required. One entry TSO, a male and female TSO on exit. At DCA there was a forth TSO standing around doing nothing (who after 5 minutes eventually did my opt out patdown). As such, there is no reduction in work force with the new software.
Those PAX doing a opt out could wait 1 to 5 minutes for a response. In my case it was a solid 5 minutes - to the point I asked the tray TSO to secure my possessions as so many other people were going through that I could not secure them. To her credit she immediately responded, gather them and stood by them and even told another TSO she was busy. I thanked her for it once I got my pat down.
Discussion ...
#2
Join Date: Feb 2008
Location: Nashville, TN
Programs: WN Nothing and spending the half million points from too many flights, Hilton Diamond
Posts: 8,043
In another thread, there was discussion on "training" the software with real field generated data. This is why it may be necessary or at least recommended.
But then, someone would have to look at the images in the training process.
But then, someone would have to look at the images in the training process.
#3
Join Date: Feb 2010
Posts: 733
My sample is much, much smaller than yours.
However, I was able to observe 3 people go through whole body imaging with ATR. The failure rate of the software was 100%:
Of course we are told these things are explosives since that is what these machines are designed to detect. Not one of these people was swabbed.
Oh, and no one changed their gloves.
However, I was able to observe 3 people go through whole body imaging with ATR. The failure rate of the software was 100%:
- I watched a man with pleats get his legs rubbed, which included his "resistance."
- I watched a woman with curly hair get both her hair and the area of her bra clasp patted down.
- I watched a man wearing glasses get his face patted. No joke.
Of course we are told these things are explosives since that is what these machines are designed to detect. Not one of these people was swabbed.
Oh, and no one changed their gloves.
#4
Join Date: Dec 2007
Posts: 629
There should be a sign in front of the ATR NoS that says "please remove hair, glasses, sweat, and clothing with folds or pleats". I have to admit to being curious about whether I would alarm the ATR NoS without my glasses, with a fully shaven head, a genuine orange prison jumpsuit or, possibly spandex cycling clothes, industrial strength antiperspirant, and some anti-anxiety pills to further reduce sweating. I wonder what the false positive rate would be then.
#5
Join Date: Feb 2011
Location: Houston, Texas
Posts: 959
There should be a sign in front of the ATR NoS that says "please remove hair, glasses, sweat, and clothing with folds or pleats". I have to admit to being curious about whether I would alarm the ATR NoS without my glasses, with a fully shaven head, a genuine orange prison jumpsuit or, possibly spandex cycling clothes, industrial strength antiperspirant, and some anti-anxiety pills to further reduce sweating. I wonder what the false positive rate would be then.
#6
A FlyerTalk Posting Legend
Join Date: Sep 2006
Location: where the chile is hot
Programs: AA,RR,NW,Delta ,UA,CO
Posts: 41,668
There should be a sign in front of the ATR NoS that says "please remove hair, glasses, sweat, and clothing with folds or pleats". I have to admit to being curious about whether I would alarm the ATR NoS without my glasses, with a fully shaven head, a genuine orange prison jumpsuit or, possibly spandex cycling clothes, industrial strength antiperspirant, and some anti-anxiety pills to further reduce sweating. I wonder what the false positive rate would be then.
Careful. If you appear too relaxed, you'll probably get SPOT'd by a BDO because it isn't normal to be cool and relaxed at an airport checkpoint.
#7
Join Date: Aug 2006
Location: DCA / WAS
Programs: DL 2+ million/PM, YX, Marriott Plt, *wood gold, HHonors, CO Plt, UA, AA EXP, WN, AGR
Posts: 9,388
At DFW over a 30 minute observation period at one checkpoint, 80+ percent of males were subject to a pat-down after the AIT scanner (even when it flashed green). Almost no women were pulled aside.
#8
A FlyerTalk Posting Legend
Join Date: Sep 2006
Location: where the chile is hot
Programs: AA,RR,NW,Delta ,UA,CO
Posts: 41,668
There should be a sign in front of the ATR NoS that says "please remove hair, glasses, sweat, and clothing with folds or pleats". I have to admit to being curious about whether I would alarm the ATR NoS without my glasses, with a fully shaven head, a genuine orange prison jumpsuit or, possibly spandex cycling clothes, industrial strength antiperspirant, and some anti-anxiety pills to further reduce sweating. I wonder what the false positive rate would be then.
I watched the screens at AMS (they are quite visible to the pax being screened, the security person and anyone else sitting in the gate area). I watched one man who was nearly bald, no glasses or facial hair, didn't appear to be sweating. I don't know why, but the screen image showed 'anomalies' all over his face.
FWIW, another young guy went through and one large 'anomaly' displayed in the center of his image. He pulled up his t-shirt and showed a monster belt buckle to the screener and was good to go.
#9
A FlyerTalk Posting Legend
Join Date: Sep 2006
Location: where the chile is hot
Programs: AA,RR,NW,Delta ,UA,CO
Posts: 41,668
#10
Join Date: Apr 2010
Location: Southwest Florida
Programs: AA lifetime Gold , DL Gold, Hilton Diamond
Posts: 572
My wife and I just recently went through the MMW at TPA, which has the new gumby software, the screener at the entry to the MMW clearly stated to us that make sure we have absolutely nothing in our pockets. My wife went first and the AIT picked up her small belt buckle that is permanently attached to her slacks, the female screener just slightly patted her down on the area shown on the screen and verified it was the belt buckle and she was good to go.
I cleared without any problem, but it was very slow at the MMW, no one behind me so I was able to observe my image and talk to the screener about the new system. My image did show some small areas with a box around it, but the screener didn’t even bother to check those areas, even after I pointed it out on the screen, he said it was nothing to be concerned about.
He said they like the new system, it was faster than waiting for the out of sight screener to either clear the passenger or describe the area that needed to be checked, and that the passengers could see for themselves where the scanner has picked up any anomalies. So for me this has eliminated the privacy issue with the NOS and I find it acceptable.
As far as the BSX AIT’s, even after they are modified with the gumby software, I will never go through one because of the radiation issue, I will take the grope instead, but so far I have been lucky, at least half the airports I usually fly in and out of have the MMW or no ATI’s have been installed and so far I have been able to either avoid the BSX or they were not using them when I went through security.
Mr. Elliott
I cleared without any problem, but it was very slow at the MMW, no one behind me so I was able to observe my image and talk to the screener about the new system. My image did show some small areas with a box around it, but the screener didn’t even bother to check those areas, even after I pointed it out on the screen, he said it was nothing to be concerned about.
He said they like the new system, it was faster than waiting for the out of sight screener to either clear the passenger or describe the area that needed to be checked, and that the passengers could see for themselves where the scanner has picked up any anomalies. So for me this has eliminated the privacy issue with the NOS and I find it acceptable.
As far as the BSX AIT’s, even after they are modified with the gumby software, I will never go through one because of the radiation issue, I will take the grope instead, but so far I have been lucky, at least half the airports I usually fly in and out of have the MMW or no ATI’s have been installed and so far I have been able to either avoid the BSX or they were not using them when I went through security.
Mr. Elliott
#11
FlyerTalk Evangelist
Join Date: Oct 2006
Location: Marriott or Hilton hot tub with a big drink <glub> Beverage: To-Go Bag™ DYKWIA: SSSS /rolleyes ☈ Date Night: Costco
Programs: Sea Shell Lounge Platinum, TSA Pre✓ Refusnik Diamond, PWP Gold, FT subset of the subset
Posts: 12,509
Wirelessly posted (Motorola DynaTAC: BlackBerry9630/5.0.0.1030 Profile/MIDP-2.1 Configuration/CLDC-1.1 VendorID/104)
I wonder if that's because the TSA has trouble staffing female screeners to do the gropes. One could imagine the pushback if they were done by male screeners.
Originally Posted by Global_Hi_Flyer
At DFW over a 30 minute observation period at one checkpoint, 80+ percent of males were subject to a pat-down after the AIT scanner (even when it flashed green). Almost no women were pulled aside.
#13
Original Poster
Join Date: Nov 2005
Location: on the path to perdition
Programs: Delta, United
Posts: 4,784
SATTSO - on the exit side of the AIT there were two TSOs, one female and one male. At the exit of AIT one TSO would reset the screen and instruct the next PAX into the AIT. The other would watch the screen and would prevent the PAX from leaving the roped off exit area until the all clear or Gumby image showed up. If a Gumby image the TSO would pat down the PAX if they were the same sex. If not the TSOs would swap places and so a male TSO would pat down a male PAX. Because the PAX's sex was random and there were so many false positives the TSOs were swapping positions very frequently. Had there been fewer false positives perhaps only one TSO would have been needed.