I've begun using the MMW systems equipped with the ATR/gumby software since there's really no health risks from low energy RF and no more virtual strip search being conducted by a voyeur in a viewing room. (Forget x-ray/backscatter tho, ATR or not). I'm sure you've all heard the barkers yelling about "all belts off, nothing in your pockets, not even your ticket, or chapstick, or lint", yada yada yada. Well, I was so focused on ensuring that all my pockets had absolutely nothing I forgot to remove my leather belt with a large metal buckle. It was under my pullover sweater so wasn't visible to the guy that ordered me into the machine.

The result after getting scanned? Absolutely nothing. Green screen, good to go. It seems the machine could care less that I went through with a belt. I know that particular belt couldn't pass the WTMD (I'd tried it before) but MMW was fine with it. I'm starting to think that all this yelling that a piece of paper will cause the machine to alarm is BS. I'll make sure to wear the belt through MMW in the future unless a screener notices and browbeats me into removing it. If it alarms in the future I'll update this thread.

Since you have broached the subject, I will tell of my experiences.

My home airport, as well as some to which I travel, have also installed the ATR system. I always wear pleated pants. All green. I have tried a wadded up tissue, green. I left a bit of change in my pocket, green. I wore earplugs on a lanyard around my neck and under my shirt, green.

I have decided to try various tests each time I go through. I agree with you that the false alarms seem to be over hyped. I think they were occurring with operators. The computers do no react to them.

The only yellow I have gotten was when my belongings were behind me on the belt and I moved my head to keep an eye on them. My whole face was alarmed, but the operator waved me through because he saw me move.

I'm torn if I should go through the ATR with MMW if I ever encounter it. Unfortunately the airports I frequent all have cancer boxes.

How do you tell if the MMW has the ATR retrofit? Does it explicitly say so?

I'd bet it works by detecting sharp edge discontinuities on the outline of a person, but I'm not sure. If so, anything "flat" like a knife or belt buckle would be ok, but not anything that sticks out, like, say, a gun or a pocketbook. Anybody have any technical information on how it works (it likely doesn't work very well...)

I know my home airport is equipped so no problem there. When I fly into my destination, I make a habit of checking out security from the sterile side on the way to bag claim. First to see if they have scanners, second to see if they are MMW or BSX, third, if MMW do they use ATR and fourth are there lanes that do not lead to a scanner at all. If it is impossible to see the CP from the sterile side, I ask at the TDC upon arrival to depart. If they do not know, yes that happens, when you get to the scanner there will be a small flat panel on the outgoing side that will be getting the attention of the operator. They will not be using radios. It moves much, much faster, I would guess 3:1 faster that the old viewer in the booth system.

I'm sure by now you know how to distinguish the MMW from the Rapiscans.
TSA says that ALL MMW's have ATD.
When they upgraded the units they put an "ATD" label on the machine like this picture shows:

What are other people's thoughts on going through this? I'm tempted to actually go through it -- essentially zero cancer risk (nonionizing) and less invasive than the ridiculous patdowns. If there is truly no "imaging stored" it is not much more invasive than WTMD.

The issues are:

1. I don't necessarily trust that they don't save the 'raw' image for 'later review' or 'quality control' etc., etc.

2. I like opting out so others see the protest, as they may not know the difference between the machines.

3. I'm not convinced that it works (see belt buckles above.) At least the WTMD would catch a metal detonator (presumably), but the AIT would, in all honesty, almost certainly let the underwear bomb on board.

Your concerns match mine. These are my hopes which may be displaced when it comes to TSA:

1. If there is no anomaly, there is no reason to store so I hope they do not. I would suppose they wold store the anomaly at least until it is resolved.

2. I do not like opting out but I did it for the reason you have stated. I am convinced no one cares, at least not enough.

3. I do not care if it works, at least personally. I just want to get to my plane unmolested. This seems to fit the bill. I do care if it works when it involves someone with nefarious intent. I know, it's a double standard.

There is still the underlying idea that the scanner is producing an image of your naked body under your clothing. True, with ATR, the screener is no longer seeing that image, and in theory, it is only being saved in memory temporarily for the software to process it, but nevertheless, an under-the-clothing image is being produced. As an extreme example, would it still be okay if you had to go into a private cell and strip down, and have several cameras take pictures of you to be processed, as long as those images were not saved and not viewed by a screener?

With a frisk, no one is seeing under your clothes, and there is no image to be potentially saved.

Now whether having a machine take an under-the-clothing image is more or less invasive than a government employee running his or her hands up and down your clothing is subject to debate. Both are disgusting options.

I agree.

Yes, of course.

The NoS (both x-ray and MMW) doesn't take a photo in the sense of a film photograph or a film medical x-ray - that is, all at once. It works by scanning the mmw beam back and forth, working up-and-down (raster scan) and recording the reflected signal on each point. Think of viewing a statue in a perfectly dark room with a laser pointer; you'd sweep it back and forth to illuminate one part at a time. At no instant would you see the whole statue, but in your mind you could build up a picture. Now imagine storing each part of that picture - the value of the pixel and where it was in relation to the other pixels - in a computer. The computer could then put the pieces together to create a complete image.

The raw data in a NoS is not intrinsically a photograph, but a string of digital data representing the sequence of pixels. That data can be processed to form a detailed image, or it can be processed to look for sharp discontinuities. The following is conjecture, but I imagine it identifies the discontinuities that represent the edges of the body - this is why everyone needs to adopt the same pose - and then looks for edges in other, unexpected places. Metal, ceramic, plastic (etc) items reflect MMW differently than your body does, so if there's a section on your chest (ie, between the edges it already found) that is discontinuous, it will alert on that.

The variables are how big the pixels are (tradeoff of resolution versus scan time), how much of a discontinuity is detected and how many pixels need to be discontinuous to be an alarm (tradeoff of false positives vs false negatives.)

All of that's theory; getting it to work in practice is harder than it sounds, so I'm not surprised there are serious problems. In the late 90's/early 2000's there were a lot of research groups working on MMW scanners but most decided it wasn't reliable in practice. There was a lot of interest in non-airport, non-human-scanning applications like looking for flaws in composite materials or debris in packaged food.

In theory, it should have found the belt buckle; either it was too small or not reflective enough or something. See previous paragraph.
It's hypothetical for me (for now), since I don't go to the US/UK and Australia has only run a few trials. Like you, I find the lack of health risk and the reduced privacy concerns make this less offensive.

My remaining personal concern would be the inability to watch my belongings while being scanned. I don't like the "I surrender" pose, and it's slower than the WTMD. But it's probably no worse than the "frisking" pose and time for the patdown* I get 50% of the time due to my metal hip.

*In Australia, Europe and Asia it's still a "pat" down. US, not so much.

My societal concerns are the exorbitant cost of the scanners and the problems they cause for people with innocent medical devices (ostomy bags, insulin pumps etc). If these machines were effective at finding explosives, I still don't believe that the risk justifies the cost and invasiveness of these devices. And as they're not effective, the cost and invasiveness are completely outrageous. So at this point, I would still opt out on principle, in the (slim) hope that it would inform other passengers and put a dent in the gov't's claim that "everyone likes the scanners".

But if, like many of you, I was flying through US airports several times a week, where the alternative is not really a "pat" down and where other passengers and the government don't care, I too might decide that the MMW/ATR was the easier course of action.
This doesn't worry me for the reason given above. The scanner isn't producing "an image" in the photographic sense. And the data that's required to form the images we're seen is a lot more than the data required to describe edges. I guess it's possible that they're storing the high-res data that could be processed into a high-res image, but the raw data is not, in the first instance "an image". There's no "image" stored temporarily while the software processes it; there's data that could have been processed to be an image and is instead being processed to find edges. Perhaps the distinction doesn't matter to most people but for me, it's quite different than producing, viewing, or storing "an image".

Thanks for the clarification.

I suppose then I should amend my observations to instead state that even if an image, as we would think of one, is not being stored in RAM and then processed, the raw data that can be potentially used to build up an image is still being generated. That is, the machine is still searching under your clothes.

But then again, a the WTMD cannot generate an under-the-clothes image, but it can still find items that are concealed under the clothing.

I must admit that ATR does make it harder to opt out on practical grounds, assuming that there truly are no potential health hazards of repeated scans, and that the data points necessary for creating an image are not being stored on disk for calibration against the "gumby" images. Opting out then becomes more of a way to make a statement, to protest what you feel is the unnecessary expenditure of hundreds of millions of dollars, but its alternative -- the frisk -- is just as much of a violation.

Great, my intuitions were right.

RadioGirl, I assumed that the MMW uses a transmission measurement and an inverse radon transform, but I must admit that was biased by the movement that I've seen of the actual device around the person, (looks similar to a CAT scan ) but of course this would not prohibit a backscatter type measurement... are you absolutely sure that it is a raster scan and measurement of reflection/scatter, and not simply shining the beam through to the other side and measuring attenuation, then reconstruction of an "absorbance" index which could produce a nice outline of the person?

Yeah, but either way the raw data for postprocessing is there. I'm scared that they will retain the data for hindsight if stuff gets through. In actuality, I'd LIKE them to do this for QC, but I simply don't trust them given the previous inability to divide by n when calculating radiation dose. Too much of a bad history.

OT, but sorry I didn't give you a ring in Sydney! Next time.

I'm absolutely sure that it's reflection/scatter, not a through transmission. The attenuation by the human body at these frequencies is so great that a transmission system would only generate a silhouette where the beam missed the body completely, and blackout where the beam hit the body. The detailed MMW images we've seen show far more detail than that.

If they had to transmit enough power to get a tomographic image, it would literally cook people in a very obvious (and painful, probably deadly) manner. (Old industry joke: "We sponsor research on RF health effects because it makes us feel all warm inside." )
I'm not absolutely sure what they do with data/storage/reconstruction; my previous comments were my conjecture and opinion. To both you and Saulblum, I don't blame anyone for not trusting TSA!

Yeah, that's what I thought they were doing because the images were indeed so poor -- the silhouette could just give an outline of the body with any sharp edge outlined. So it is scatter? Interesting. Doubt they could really detect "artfully concealed" contraband.

It's a reasonable position to use the MMW with ATR. However, whether "opting out" really matters or not - by going through the machine, you are "casting your vote" in favor of: (in no particular order)

1) A reckless bureaucracy that has repeatedly proven it will do whatever it wants and openly lie about it, without fear of consequence.

2) Whatever hare-brained scheme they come up with next.

3) The unrestrained waste of millions of tax dollars, adding onto a monstrous debt, for less than zero benefit.

4) The slow decline of the ability of the average American to understand probability, economics, and various other numerical disciplines.

5) The routine abuse of the thousands of people who can't "hold the pose", especially the frail and elderly.

6) The much easier ability to smuggle a metallic weapon through the checkpoint, in the very rare instances someone might want to cause harm.

There are so many reasons to reject this technology. The "pat down" is terrible, certainly. But it's really hard to respect the choice to surrender to the AIT when I think about what it represents.