Things in checked luggage that look suspicious
Nov 30, 2009, 3:46 pm
#46
Join Date: Sep 2009
Posts: 3,702
Try these images from dual-energy X-ray systems that try to identify materials http://www.vidisco.com/DualXray.asp. It shows weakness in regular x-ray as compared to that vendor's dual-energy system. Especially interesting is the similarly-orange book+TNT image and the test pattern image with sugar and salt.
Another vendor's examples show dual-energy material identification system weaknesses as compared to Z(eff) systems.
I know TSA's policies aren't really aimed at smart terrorists, but the boogey-man would buy, rent, or borrow equipment similar to the target system (Was Britney's x-ray a Rapiscan 620DV as in Bob's post?) and make something that looks the same on screen as a book.
Take a look at Rapiscan's 620DV sales brochure and think about what level of sneakiness would get past you after looking at just 20 passengers worth of bags. A 5-lb bag of sugar might get noticed, but a sugar-filled book would be just another book. Snow globes probably look enough like a Bullwinkle-cartoon hush-a-bomb to cause a panic.
Another vendor's examples show dual-energy material identification system weaknesses as compared to Z(eff) systems.
I know TSA's policies aren't really aimed at smart terrorists, but the boogey-man would buy, rent, or borrow equipment similar to the target system (Was Britney's x-ray a Rapiscan 620DV as in Bob's post?) and make something that looks the same on screen as a book.
Take a look at Rapiscan's 620DV sales brochure and think about what level of sneakiness would get past you after looking at just 20 passengers worth of bags. A 5-lb bag of sugar might get noticed, but a sugar-filled book would be just another book. Snow globes probably look enough like a Bullwinkle-cartoon hush-a-bomb to cause a panic.
And yes, we already know that terrorist have ALL of our neat devices to learn how to get things through both te checkpoint ant baggage. We have found inside caves in various countries our x-rays, wtmd, hhmd, etd machines - you name it.
Nov 30, 2009, 4:37 pm
#47
Join Date: Aug 2009
Location: MSP
Programs: DL Silver, UA Silver, AMX Plat, Carlson Gold, SPG Gold
Posts: 188
I carry on the regulator and dive computer, check the mask, fins, booties, bcd and wetsuit. I can always rent that stuff, but dont want to rent a regulator or computer. I even keep the regulator and computer in a little bag inside my rollaboard in case I have to gate check it. I can pull out the regulator and at least carry that on.
Nov 30, 2009, 7:30 pm
#48
Join Date: Apr 2009
Posts: 252
Kind of curious, what about the book & TNT do you find interesting?
And yes, we already know that terrorist have ALL of our neat devices to learn how to get things through both te checkpoint ant baggage. We have found inside caves in various countries our x-rays, wtmd, hhmd, etd machines - you name it.
And yes, we already know that terrorist have ALL of our neat devices to learn how to get things through both te checkpoint ant baggage. We have found inside caves in various countries our x-rays, wtmd, hhmd, etd machines - you name it.
Since they have the machines, then they could create mix recipes to match any density they choose. A little bit of microballoon filler, and the slightly darker orange of the TNT would be indistinguishable from the paper.
Of course, I'm no explosives expert, but with these machines, TSA has converted its high-volume detection problem into a visual inspection process, where I do know a little.
Dec 1, 2009, 2:12 pm
#49
Join Date: Sep 2009
Posts: 3,702
What I find interesting is that the uncut book and the TNT look similarly orange in the image. If the book was cut smaller, or was filled with TNT, then the image wouldn't look so much like a framed bomb, but a plain book with a phone on it.
Since they have the machines, then they could create mix recipes to match any density they choose. A little bit of microballoon filler, and the slightly darker orange of the TNT would be indistinguishable from the paper.
Of course, I'm no explosives expert, but with these machines, TSA has converted its high-volume detection problem into a visual inspection process, where I do know a little.
Since they have the machines, then they could create mix recipes to match any density they choose. A little bit of microballoon filler, and the slightly darker orange of the TNT would be indistinguishable from the paper.
Of course, I'm no explosives expert, but with these machines, TSA has converted its high-volume detection problem into a visual inspection process, where I do know a little.
It is also important to note that these machines are not "take it as it is". TSA has had our machines "tweaked". As I am sure you already know, these machine produce false colors. Out machines have been adjusted to change the colors that you have seem on those sites you link to. What colors? Can't tell you; I'm sure you will understand that's SSI (not that I am saying it is a perfect solution, of course it is not).
Edit: on another note, the x-ray itself inside the entire machine are placed at different angles for different machines. This is also done by design. Now there are so many places it can be place because of physical restraints, but it is done, and the angles of each machine is also SSI. And if you think about it, this also helps to explain why sometimes your bag is check and another time it is not, even when you put your bag in at the same angle with everything in the same place. Or course, person interpertation does play a very large (largest?) role.
Last edited by SATTSO; Dec 1, 2009 at 2:21 pm
Dec 1, 2009, 3:44 pm
#51
Join Date: Apr 2009
Posts: 252
Yes, sometimes we have to visually inspect certain items to verify what it is. I believe this thread is about checked luggage, yet these are machines used in chekpoint only. And it is interesting if we add what YOU said, and tie that into another thread involving additional screening.
It is also important to note that these machines are not "take it as it is". TSA has had our machines "tweaked". As I am sure you already know, these machine produce false colors. Out machines have been adjusted to change the colors that you have seem on those sites you link to. What colors? Can't tell you; I'm sure you will understand that's SSI (not that I am saying it is a perfect solution, of course it is not).
It is also important to note that these machines are not "take it as it is". TSA has had our machines "tweaked". As I am sure you already know, these machine produce false colors. Out machines have been adjusted to change the colors that you have seem on those sites you link to. What colors? Can't tell you; I'm sure you will understand that's SSI (not that I am saying it is a perfect solution, of course it is not).
Still, filtering a couple x-ray bands into a false color RGB image is a far cry from hyperspectral imaging, which is why the CRT mediated TSA eyeball inspection system alarms on things like brownie mix, books, cheese, and soap, and flour. If you've only got RGB (and maybe blinking) to pack density, opacity, organic, inorganic, or a few other estimators of materials, paper could very well look like TNT. If you jack up your TNT-paper discrimination, you might lose sensitivity to other items of interest.
I really don't care about the specific colors, just that TSA appears to be making magical claims about how it can pick 1-in-5-billion needles out of haystacks while getting confused about bars of soap. If 3.5oz of toothpaste is actually dangerous, a soap-bar or paperback-worth of TNT would be more like the hay in the haystack.
Edit: on another note, the x-ray itself inside the entire machine are placed at different angles for different machines. This is also done by design. Now there are so many places it can be place because of physical restraints, but it is done, and the angles of each machine is also SSI. And if you think about it, this also helps to explain why sometimes your bag is check and another time it is not, even when you put your bag in at the same angle with everything in the same place. Or course, person interpertation does play a very large (largest?) role.
Dec 1, 2009, 5:55 pm
#52
Join Date: Sep 2009
Posts: 3,702
Yep. I missed that this was checked luggage. There you can use bigger, fancier machines with more spectral bands.
Still, filtering a couple x-ray bands into a false color RGB image is a far cry from hyperspectral imaging, which is why the CRT mediated TSA eyeball inspection system alarms on things like brownie mix, books, cheese, and soap, and flour. If you've only got RGB (and maybe blinking) to pack density, opacity, organic, inorganic, or a few other estimators of materials, paper could very well look like TNT. If you jack up your TNT-paper discrimination, you might lose sensitivity to other items of interest.
I really don't care about the specific colors, just that TSA appears to be making magical claims about how it can pick 1-in-5-billion needles out of haystacks while getting confused about bars of soap. If 3.5oz of toothpaste is actually dangerous, a soap-bar or paperback-worth of TNT would be more like the hay in the haystack.
Yep, that personal interpretation in a low rate detection problem is a large problem. Well defined visual inspection catches only about 80% of what you are looking for, and with the varied materials people travel combined with and the wide range of possibly dangerous items, it quickly becomes an intractable inspection problem as the false alarm rate swamps the needles in the haystacks: Can you practically swab each bar of soap? If the 3,000th baby powder bottle looks just like another baby-powder bottle, do you go further?
Still, filtering a couple x-ray bands into a false color RGB image is a far cry from hyperspectral imaging, which is why the CRT mediated TSA eyeball inspection system alarms on things like brownie mix, books, cheese, and soap, and flour. If you've only got RGB (and maybe blinking) to pack density, opacity, organic, inorganic, or a few other estimators of materials, paper could very well look like TNT. If you jack up your TNT-paper discrimination, you might lose sensitivity to other items of interest.
I really don't care about the specific colors, just that TSA appears to be making magical claims about how it can pick 1-in-5-billion needles out of haystacks while getting confused about bars of soap. If 3.5oz of toothpaste is actually dangerous, a soap-bar or paperback-worth of TNT would be more like the hay in the haystack.
Yep, that personal interpretation in a low rate detection problem is a large problem. Well defined visual inspection catches only about 80% of what you are looking for, and with the varied materials people travel combined with and the wide range of possibly dangerous items, it quickly becomes an intractable inspection problem as the false alarm rate swamps the needles in the haystacks: Can you practically swab each bar of soap? If the 3,000th baby powder bottle looks just like another baby-powder bottle, do you go further?
And you seem to have a misunderstanding of what happens: no x-ray or any machine we have can detect an IED. All of our fancy machines can do is help to identify specific characteristics that a common object might share with IEDs and their component. That's it. Where is the false-positive coming from?
Dec 1, 2009, 8:34 pm
#54
Join Date: Apr 2009
Posts: 252
It was some half-remembered thing from TSO Ron about the danger of a bunch of soap bars. If the machine could tell you that soap was soap or cheese was cheese or a common object wasn't an IED component then you wouldn't have to do the open the bag to do a different inspection.
I do understand that. Much of your system relies on unreliable visual inspection because the available objective tests have unworkable error characteristics. (Think puffer.)
False positive is when x-ray calls for a bag check and then don't find an IED.
I think TSA is pretty sloppy with 'false positive' in that they count finding suspicious but non-WEI items as "successes".
My use of "false-positive" comes from the first round of inspection which causes a second round of inspection. The first part of your process seems to be that you run the bag though the visual x-ray inspection, if it passes the visual x-ray, then you don't open it up and you send the traveller on their way. If it fails the visual x-ray, then you open it up and look at it with a Mark I eyeball, and if it still looks suspicious you escalate it to a swabbing or a BAO or whatever. Each escalation decision for a non-WEI is a false positive.
Each stage of decision making in the process has its own sensitivity, specificity, false positive, and false negative error rates. If the potential terrorist tricks the first "layer of security", the bag check and swabbing layers don't come into play. On the other hand, if the first layer is overly sensitive, then the Mark I Eyeball inspection at bag check ends up being the main test. Take a look at false positive paradox and plug 2,000,000 people per day and one seventh of the week-at-a-glance numbers in for the x-ray process.
If there was a box could detect an IED components reliably (high sensitivity, high specificity), like the puffer promised, then the number of bag checks would be much reduced, and when you did do a bag check, you'd likely find something wicked. Think of a Geiger counter looking for radioactivity -- there aren't many things that could cause false alarms, in contrast to a visual inspection for dense objects in people's luggage.
And you seem to have a misunderstanding of what happens: no x-ray or any machine we have can detect an IED. All of our fancy machines can do is help to identify specific characteristics that a common object might share with IEDs and their component. That's it. Where is the false-positive coming from?
False positive is when x-ray calls for a bag check and then don't find an IED.
I think TSA is pretty sloppy with 'false positive' in that they count finding suspicious but non-WEI items as "successes".
My use of "false-positive" comes from the first round of inspection which causes a second round of inspection. The first part of your process seems to be that you run the bag though the visual x-ray inspection, if it passes the visual x-ray, then you don't open it up and you send the traveller on their way. If it fails the visual x-ray, then you open it up and look at it with a Mark I eyeball, and if it still looks suspicious you escalate it to a swabbing or a BAO or whatever. Each escalation decision for a non-WEI is a false positive.
Each stage of decision making in the process has its own sensitivity, specificity, false positive, and false negative error rates. If the potential terrorist tricks the first "layer of security", the bag check and swabbing layers don't come into play. On the other hand, if the first layer is overly sensitive, then the Mark I Eyeball inspection at bag check ends up being the main test. Take a look at false positive paradox and plug 2,000,000 people per day and one seventh of the week-at-a-glance numbers in for the x-ray process.
If there was a box could detect an IED components reliably (high sensitivity, high specificity), like the puffer promised, then the number of bag checks would be much reduced, and when you did do a bag check, you'd likely find something wicked. Think of a Geiger counter looking for radioactivity -- there aren't many things that could cause false alarms, in contrast to a visual inspection for dense objects in people's luggage.
Dec 2, 2009, 6:49 am
#55
Join Date: Sep 2009
Posts: 3,702
It was some half-remembered thing from TSO Ron about the danger of a bunch of soap bars. If the machine could tell you that soap was soap or cheese was cheese or a common object wasn't an IED component then you wouldn't have to do the open the bag to do a different inspection.
I do understand that. Much of your system relies on unreliable visual inspection because the available objective tests have unworkable error characteristics. (Think puffer.)
False positive is when x-ray calls for a bag check and then don't find an IED.
I think TSA is pretty sloppy with 'false positive' in that they count finding suspicious but non-WEI items as "successes".
My use of "false-positive" comes from the first round of inspection which causes a second round of inspection. The first part of your process seems to be that you run the bag though the visual x-ray inspection, if it passes the visual x-ray, then you don't open it up and you send the traveller on their way. If it fails the visual x-ray, then you open it up and look at it with a Mark I eyeball, and if it still looks suspicious you escalate it to a swabbing or a BAO or whatever. Each escalation decision for a non-WEI is a false positive.
Each stage of decision making in the process has its own sensitivity, specificity, false positive, and false negative error rates. If the potential terrorist tricks the first "layer of security", the bag check and swabbing layers don't come into play. On the other hand, if the first layer is overly sensitive, then the Mark I Eyeball inspection at bag check ends up being the main test. Take a look at false positive paradox and plug 2,000,000 people per day and one seventh of the week-at-a-glance numbers in for the x-ray process.
If there was a box could detect an IED components reliably (high sensitivity, high specificity), like the puffer promised, then the number of bag checks would be much reduced, and when you did do a bag check, you'd likely find something wicked. Think of a Geiger counter looking for radioactivity -- there aren't many things that could cause false alarms, in contrast to a visual inspection for dense objects in people's luggage.
I do understand that. Much of your system relies on unreliable visual inspection because the available objective tests have unworkable error characteristics. (Think puffer.)
False positive is when x-ray calls for a bag check and then don't find an IED.
I think TSA is pretty sloppy with 'false positive' in that they count finding suspicious but non-WEI items as "successes".
My use of "false-positive" comes from the first round of inspection which causes a second round of inspection. The first part of your process seems to be that you run the bag though the visual x-ray inspection, if it passes the visual x-ray, then you don't open it up and you send the traveller on their way. If it fails the visual x-ray, then you open it up and look at it with a Mark I eyeball, and if it still looks suspicious you escalate it to a swabbing or a BAO or whatever. Each escalation decision for a non-WEI is a false positive.
Each stage of decision making in the process has its own sensitivity, specificity, false positive, and false negative error rates. If the potential terrorist tricks the first "layer of security", the bag check and swabbing layers don't come into play. On the other hand, if the first layer is overly sensitive, then the Mark I Eyeball inspection at bag check ends up being the main test. Take a look at false positive paradox and plug 2,000,000 people per day and one seventh of the week-at-a-glance numbers in for the x-ray process.
If there was a box could detect an IED components reliably (high sensitivity, high specificity), like the puffer promised, then the number of bag checks would be much reduced, and when you did do a bag check, you'd likely find something wicked. Think of a Geiger counter looking for radioactivity -- there aren't many things that could cause false alarms, in contrast to a visual inspection for dense objects in people's luggage.
Personally, if I honestly thought there was an IED in a bag in the x-Ray, I would tell the passengers, "see my ...? Follow it as I run out of the building!" I joke, of course, but it's to highlight a point. Terminals are not evacuated with each an devery bag check considering how many times a terminal is evacuated compared to how many times bag checks are done should tell you what a relative rate of false alarms is (though some evacuations are justified as people mistakenly carry gernades with them sometimes thinking it's ok).
Bag checks fall into several categories: clearly a prohibited is there; verify what an item is. A bag check is NEVER done when someone believes an IED is there.
s I understand it there will never be a "box" or device that will be able to detect IEDs or their components. Here is why. The liquid bomb the nojinka terrorist used had a timing device. It was a cheap digital watch with holes drilled into it to allow for wires to be inserted. If that is inside a bag, how would a machine catch something like that, or an electrical switch used in in thousands of everyday electrical devices?
All that our best technologies can so is this: determine that an object has the same density of a KNOWN explosive, and that's it. Sadly, many items have simiar densities of know explosives, and many common items have chemical components of known explosives in them (though they themselves are harmless).
Dec 2, 2009, 7:22 am
#56
Join Date: Jun 2006
Location: TPA
Posts: 282
I've only had my checked baggage rummaged through by TSA in PHX once and I think that happened b/c I had a Xbox 360 in it.
So, if I have a TSA approved lock on my checked bag is TSA likely to pull it b/c there is a lock and uh o, it looks suspicious? I just don't want my luggage to miss the plane next week b/c it gets pulled.
So, if I have a TSA approved lock on my checked bag is TSA likely to pull it b/c there is a lock and uh o, it looks suspicious? I just don't want my luggage to miss the plane next week b/c it gets pulled.
Dec 2, 2009, 9:52 am
#57
Join Date: Apr 2009
Posts: 252
Your concept of false positive is way off base. Not to be rude though. Bag checks are called to verify exactly what an obect is. If a TSO really thought there was an iED in a bag such a belief would work it's way up the command to a BAO (now that we have them) and to the local bomb squad. The vag would REMAIN in the x-ray tunnel, assuming we are talking about the operator of the x-ray being the one who belives there was an IED in there.
Personally, if I honestly thought there was an IED in a bag in the x-Ray, I would tell the passengers, "see my ...? Follow it as I run out of the building!" I joke, of course, but it's to highlight a point. Terminals are not evacuated with each an devery bag check considering how many times a terminal is evacuated compared to how many times bag checks are done should tell you what a relative rate of false alarms is (though some evacuations are justified as people mistakenly carry gernades with them sometimes thinking it's ok).
Bag checks fall into several categories: clearly a prohibited is there; verify what an item is. A bag check is NEVER done when someone believes an IED is there.
Bag checks fall into several categories: clearly a prohibited is there; verify what an item is. A bag check is NEVER done when someone believes an IED is there.
One way that TSA differs from normal testing processes is that actual weapon-carrying terrorists are so incredibly rare. In order to make the security theatre seem worthwhile you have to define non-WEI items as WEI.
s I understand it there will never be a "box" or device that will be able to detect IEDs or their components. Here is why. The liquid bomb the nojinka terrorist used had a timing device. It was a cheap digital watch with holes drilled into it to allow for wires to be inserted. If that is inside a bag, how would a machine catch something like that, or an electrical switch used in in thousands of everyday electrical devices?
By the very same reasoning TSA will never be able to detect [all] IEDs or their components. Every day TSA lets millions of watches, switches, batteries, and small bottles fly uninspected. So what's the point of it? Security theatre? CYA?
All that our best technologies can so is this: determine that an object has the same density of a KNOWN explosive, and that's it. Sadly, many items have simiar densities of know explosives, and many common items have chemical components of known explosives in them (though they themselves are harmless).
IOW, it isn't TSA skill that is keeping us safe, it is the fact that the world isn't anywhere near as dangerous as TSA would have us believe.
Dec 3, 2009, 5:31 am
#59
Join Date: Jul 2008
Posts: 65
My 15 gel candles packed in checked luggage to give as Christmas gifts for friends and employees of a hotel in French Polynesia caused quite a ruckus with TSA several years ago (back when you could watch your checked luggage being searched). They wanted to know WHY I "needed" 15 candles! Use your imagination, guys... some of us bring GIFTS when we travel!
Dec 3, 2009, 9:42 am
#60
Join Date: Jul 2002
Location: Victoria, BC
Programs: UA 1k, AA Exec Plt 2MM, HH Diamond, *wood Gold, disgruntled Amex Ex-Centurion
Posts: 584
FedEx, my what a splendid idea. A 10 kg package from the Canadian west coast to Germany is just a bargain basement $285 - if I want it there by Monday (today is Thursday). Also, it will likely spend additional time in customs. Probably I will have to pay customs, brokerage and VAT on top of that - there is no personal allowance for things that don't accompany you into the country.
Now let's say I take another 10 kg package back from Germany to Canada: a cool € 214 or, in these days of record low dollar exchange, just about $310. Due to the time difference, this one would even make it until tomorrow. And again, I would have to pay brokerage - even though I would save the duties this time (Canada allows to declare "goods to follow" when entering the country).
In other words: one 10kg package both ways: $599, excluding any insurance, brokerage fees and duties. I had airfare this year that was less roundtrip (YVR or SEA to Germany).
So please spare me the FedEx litany. And show me the 10 kg (22 lbs) pet elephant.