The scan takes fractions of a second and has shown to be 99 percent accurate during testing
Presumably it's 99% effective at matching random faces to the photos -- if a terrorist group has a pool of 1000 stolen passports, with a system that's only 99% effective at maching a random photo, I think they could do much better at finding a person (perhaps with some surgical alterations) that fits in that 1%.
The process eliminates the need for an airline employee to manually check every boarding pass and passport while boarding a plane.
So at least they are clear that this is a cost-saving method for airlines, it has nothing to do with speeding boarding or making boarding more secure.
This spring, Lufthansa announced that it boarded an A380 with over 350 passengers at LAX in less than 20 minutes—less than half of their normal time—using self-boarding gates linked to CBP’s facial-enabled traveler verification service,” McAleenan said. “No more fumbling with your boarding pass while you have two carry-ons, maybe a kid; no more trying to find your QR code or trying to refresh your screen.
Oh wait, I guess it is about speeding boarding... but is boarding pass scanning really what slows boarding? I always seem to get backed up in the line at the plane door, waiting for everyone to stow their carryons and sit down. Most busy flights have two agents scanning and most people who have flown more than once scan their pass/phone quickly.... I'm surprised it's any faster getting people to stand in the right spot and stand still for a photo.
Self-scanning in supermarkets was supposedly all about speeding checkout as a benefit to the consumer. This is about speeding boarding - that's the marketing friendly benefit to be shouted about.
In actuality supermarket self-scanning is often slower than a traditional checkout queue. Particularly when items require staff authorisation. I expect this to be no better than traditional boarding in most cases.
The real benefit to both this, and supermarket self-scanning, is another low wage employee can be canned. This isn't a very marketing friendly message so it's no surprise that this isn't the message that publicised loudest.
Per item? Sure. For small numbers of items, counting time spent waiting? Heck no, I regularly get through self-checkout much faster than people who get into a regular queue, in no small part because there's often a few times more self-checkout scanners than there are open checkouts.
When you have a full cart or two, the speedier scanning at a traditional checkout pays off. But usually both are available, not just self checkout, so you're free to choose when that isn't the case.
The big Tesco nearby has a single combined queue for all the 20 or so self-scan checkouts. They are rarely all in service. With a few items it is now always faster to go to an operator checkout unless you go in at 4am. The express lanes of "under 10 items" or "cash only" are a memory of a former, much faster time.
There is often an employee aggressively marshalling people to the self-scan checkouts at every available opportunity and trying to pull them away from the staffed lines.
In the local Pound shop they tried replacing all checkouts with self-scan, getting rid of all others. They had to reverse this a month or two later and put back half a dozen, now brand new, staff operated tills - ripping out a few weeks-old self-scanners to make space. They are again slower thanks to rarely having staff nearby to handle the steady stream of staff authorisations and errors, so there's a big delay every time someone wanders over from the back of the store to wave their bar code and immediately disappear again.
Sometimes it works though. At the Kings Cross Waitrose off peak there isn't a queue and the whole process is pretty quick. The good ones don't even bother weighing the stuff which speeds things.
Checkout queue optimization at a supermarket is pretty easy. It's all about time per scan. A good cashier (60-70 scans per minute) can be something like 50% more productive due to scan time savings. A slow cashier (10 scans per minute) is a drag.
Self-checkout is all about cutting labor hours, period. It is both slower and leads to much higher shrink (as much as 150% more) than a manned casher. Stores hope to make more money by keeping the store open longer. The hacks around the slowness of checkout without cashiers is for people optimizing for time is online ordering or shopping at midnight.
>> Self-checkout is all about cutting labor hours, period.
That is exactly the case. If you pay 10$ an hour for 3 shifts 8 hours each, this makes 3x8x10 = $240 a day. That is $7200 in 30 days per one checkout spot.
To setup equipment for self-checkout and maintain it order of magnitude cheaper compared to self checkout.
Another factor is eliminating managerial overhead while dealing with human beings, as one store manager told me referring to self-checkouts:
"They are never sick, they do not have an attitude, they do not quit and they can't be rude to customers".
Oh they can be rude to customers all right! And their attitude - push this button! Scan again! Put that item back on the scale! Alert! Alert! I despise them with every fibre of my being.
Yeah, I think the "rude to customers" thing (I've heard it too) is far too focused on the act of rudeness rather than the effects (unhappiness / lower retention).
Self-checkouts trade one kind of unhappiness for another, and many (most?) are horrifying experiences. There are a few that are reasonably quick and user-friendly, but certainly not all. Most seem to be scraping the bottom of the barrel in an effort to squeeze a few more $ off the purchase price.
That’s not the right comparison. What you want to compare is how long it takes at a staffed queue vs an unstaffed one once you get to the front of the line (for the same purchase).
Things might be faster now with the unmanned ones, but the contention is that that’s only because they opened more checkout terminals. Granted, they wouldnt have done that with staffed ones because of the increased expense, but let’s not get confused about the real cause.
This seems like a faulty comparison. Space for registers is naturally limited. A single traditional register is replaced by 3 self-checkouts at my local grocery stores. Usually they opt for 6x self checkout replacing two traditional registers, manned by a single employee who can help with problems. In these cases it's not a question of staffing, but of space and the increase in space efficiency.
As always it depends on what you want to measure. Comparing total checkout time including queuing though for 'speed (items/s)' is going to naturally favor the self-checkout because the choice between normal and self checkout is heavily weighted by the number of items a particular person has. Anecdotally I've never seen anyone go through self checkout with a full cart but see it all the time with the normal registers. Additionally there's rarely the same number of traditional registers actually staffed compared to the number of self checkouts.
I agree. I would say that a better measure than items/s would be customers/hr (closer to throughput rather than speed) averaged across an entire store with and without self-checkouts. As you said, the optimal setup will depend heavily on the mix of customer types ("express" vs. full cart) a particular store gets. There's probably a good ratio for most stores between the two and I think that's what they're all converging on now.
Agreed that space is a part of the comparison, but what I've seen is that although they put 3 checkouts in the y-axis space of a single staffed lane, they end up having to use up almost 3 times as much space in the x-axis to allow people to pass each other, which simply isn't necessary to allow in a staffed lane. So I don't believe the amount of space is that different.
No, there really is a huge difference. It's more like 1.5 times the y-axis space in your example, they could get 6 self service checkouts in the space of 2 staffed ones:
Space comes with a price too. If you ignore all costs, a traditional checkout that is read when you walk up is probably faster most of the time vs an available self-checkout.
>What you want to compare is how long it takes at a staffed queue vs an unstaffed one once you get to the front of the line (for the same purchase).
Maybe if you're selling a checkout system. Personally I care how fast I can go from "got things" to "left building". (or maybe more accurately: left home -> returned home, but that has a lot more variables)
The other benefit of self-checkout is you can fit perhaps 8 machines in the space that you used to use for 3 lanes, so even though an individual checkout might take longer, overall you can probably get higher throughput on the self-checkout.
It is possible for there to be multiple benefits. In fact this is the most desirable, since when there is an immediate economic benefit to both sides, it is incentivized to happen.
In the case of supermarket self-scanning, sure, the supermarket can reduce costs.
However, the benefit to the customer is in a shorter total time through checkout. Entropy causes customers to arrive at checkouts in spurts, rather than there being a steady flow. It is expensive for stores for checkout operators to remain idle (and would be passed on in higher prices). So customers have to queue instead of there being more checkout operators available during spurts.
On the other hand, the cost of maintaining extra self-checkout machines costs very little. So in practice, they tend to be available to handle spurts of activity, resulting in faster overall checkout times for customers too.
For the stores: yes, reduced costs. For the customers: reduced costs result in lower prices in healthy markets, and also time to traverse checkouts is faster.
The same can be applied to airlines. Fewer staff at the line at the gate results in reduced staffing costs: sure. But also: less standing around in lines and less juggling at the front of the line (especially for the infirm, or those with kids, etc) is also a clear benefit to passengers.
It is cynical to ignore the other half of the benefits and focus just on costs. And in any case, reduced costs for the company results in lower prices for consumers.
> On the other hand, the cost of maintaining extra self-checkout machines costs very little. So in practice, they tend to be available to handle spurts of activity
If only! At the (UK) Tesco metro near my workplace, the staff have this insane habit of keeping half or fewer of the available self-checkout machines in operation. The rest of them have a big symbol like a stop-sign and are unusable.
Only if a big queue builds up will the single attending staff member go round and laboriously make the other machines available, whilst simultaneously having to attend to the frequent weighing issues caused by the machines. By the time they're done with this the lunchtime rush is practically over.
I mentioned this to a work colleague who suggested the reason is that they want to reduce the amount of time counting cash out of the machines at the end of the day, so having fewer machines in operation is preferable.
That doesn't make sense - you would think the machines could track their cashbox balances perfectly, requiring dumping it in a bag, printing a status report/receipt/audit log, and attaching that to the bag. No counting required!
Aldi and Lidl - the discount supermarkets in the UK manage just fine with spurts without a self scan machine anywhere. They are constantly opening and closing checkouts as the people waiting varies. They seem to have a default to get staff off tills and back to the rest of the store when people waiting drops to some prescribed number.
Go in really early or late and there are no lines occupied until a customer waits - usually no more than a minute as they appear to have a strong focus on keeping an eye out.
I would say there's a shorter overall wait, and a far higher likelihood of getting out of the store in the shortest overall time than any of the alternatives at the competition.
So no, I haven't seen any of these much-touted speed benefits. Quite the opposite frequently.
The actual scanning might be slower, but the queues are much shorter for self-scanning at the stores where I by groceries.
Also, where I live we have portable scanners. You pick up a scanner when you enter the store, you scan all your groceries and put them directly in the bag (or under the stroller) and when you reach the exit you leave the scanner and pay at a terminal and off you go.
Man I hate these with a passion. I always forget one thing or so, and then every now and then you're selected for a random check, and you feel like a common thief and have to do the walk of shame to the actual register for re-scanning and paying.
Also, the self scanners require you to weigh your fruits & vegetables at the f&v section. Although it seems they're adding scales to the self checkout boots lately, so hopefully I can start using the in-store scanning the way I use Anglo-Saxon self-checkout booths soon - just load up in the shop, scan everything yourself at the end.
Portable scanners FTW! Once you get competent with them it's ridiculous how much time you save. Add to that bringing your own bags and you are scanned and out in little more time than it took to do the actual shopping.
Every now and then you get audited, but I'll happily pay that price in order to have the scanner.
Though the difference between a supermarket checkout and an airplane is once you're through the checkout, you're free to leave.
You don't walk out the supermarket door and have to wait for a hundred people to load their cars and make their way out of the parking lot. Which means that getting you through the checkout faster just means more time waiting to get out of the parking lot.
Here in the Netherlands (at Albert Heijn stores) one small change from the US system makes all the difference: they don't weigh the product after you scan it. No 'unexpected item in bagging area' etc. It goes extremely quickly.
If there's a cashier with more than one person waiting I will go to the self-scanners.
I think it’s much more convenient with self-checkout, because I’m not holding up the line. I can take my sweet time and pack my bags as I go. It’s really a lot of trust being placed with the customer, it’d be very easy to shoplift.
It's not easy (not any easier) to shoplift because there are random checks. People are much less inclined to put something extra in the bag if they know they can be checked. The ones that do shoplift are not using the bags with a lot of other groceries to do it, they likely hide items in their clothing. In such a mode, it doesn't make any difference how and whether they check out any actual paid items.
Identity checks are handled by TSA, not airline employees, so this scan has nothing to do with security. They check your boarding pass to make sure you have paid for the flight you are getting on.
Also "99% accuracy" can mean it has a 1% false-negative rate rather than false-positive, in which case they will just defer to scanning the boarding pass like normal.
Correct. And they're essentially acting as outward immigration, since the airline's passport records are used to mark the unwanted/foreigners (like me) as departed.
But the system is quite flaky, my arrival/departure record is missing a few departures - and I think that's where CBP want to improve things, possibly so they can more legitimately claim that someone overstayed when their records are incomplete. (And I guess also to spot passport swaps.)
The airline checks passports on the way out as they are required to bring you back if you are denied entry at the destination. They try hard not to pay that cost, so they want to make sure you have a passport, all required visas, haven't been barred from the destination, and so on.
It should come as no surprise the CBP will cooperate with other government departments, i.e. ICE. What would be interesting is finding out how they got funding for this rubbish.
Judging from when they had these at SecTac in June, it just made everything much, much slower.
> It should come as no surprise the CBP will cooperate with other government departments, i.e. ICE
Considering the "C" in both agency acronyms stands for the same thing, it would be more surprising if they didn't work very closely together. Personally, I'm surprised they're separate agencies in the first place.
On most flights within the Schengen area your boarding pass is no more manually checked. That's unless you fly business or are a high value status customer.
Instead you just scan your boarding pass / smart phone yourself.
I can't imagine that this is in any way slower than a face scan.
I thought airlines or airports were required to check your identity to endure that the checked in luggages fly along the passenger who checked them in, in case there is a bomb or something. Granted this rule is a bit absurd in a world where most terrorist attacks are suicide attacks.
Certainly when a borading pass, matched to a passenger who checked in luggage, is not used for boarding the luggage will be unloaded.
Identity check really depends on the airport and airline. Sometimes your identity is checked when you enter the secure area, sometimes at the gate and sometimes not all.
Even at airports that don't check identities airlines - notably budget airlines - may check id at the gate.
It may also happen that id is (spot) checked upon exiting the plane, but this is very rare.
That's only applicable for flights within the Schengen area. Passport checks upon leaving or entering the Schengen area are pretty strict. Essentially every piece of id is checked towards the SIS[1]. But within Schengen and depending on the airport there may be no checks at all.
I fly from Schiphol with great regularity and my boarding pass is definitely checked every time by a person (as is my passport) right before boarding the plane. Usually the same happens on the return leg, and the majority of my trips are within Schengen.
I'm actually surprised that Schiphol does that. My informal survey indicates that ex-Eastern European countries tend to check more as do the French.
In Zurich your id is virtually never checked on Schengen flights. Last time I flew through Vienna security wanted to see the boarding pass, but not the ID. Dusseldorf also didn't check when I flew through it last (alas, a few years ago).
I know at the Oslo airport it's fairly common to never show ID or your boarding pass to a person if you're flying within Schengen. --Check in at the automated kiosk, put on your own luggage tag, scan the tag at the bag drop, and you're done there. Passing through security means scanning your boarding pass rather than showing it to someone most of the time, and finally many of the gates have automated doors where you scan your boarding pass again to board the plane. The same applies to other Scandinavian airports as well.
From Schiphol on Schengen flights frequently (~50% of the time) no one checks anything in my experience outside of scanning the QR to get into security, and into the plance.
Definitely not a time-saving measure. The time limiting factor is the actual getting-on-the-plane-and-putting-your-carryon-away-and-getting-out-of-the-aisle part. Often I get my boarding pass scanned and go stand in line on the gangway.
In fact a mathematical algorithm was developed for loading faster[1] but I'm not sure anybody tried that yet.
> if a terrorist group has a pool of 1000 stolen passports, with a system that's only 99% effective at maching a random photo, I think they could do much better at finding a person (perhaps with some surgical alterations) that fits in that 1%
Is this worse than the current security measures in place while boarding?
Probably not since agents don't check ID at all, but if I'm going to give up my privacy by having my passport photo shared with airports without my knowledge or consent, I'd like to think that I'm getting something out of the deal.
> if I'm going to give up my privacy by having my passport photo shared with airports without my knowledge or consent
The same place where you literally already have to have your passport scanned, your face photographed, you're recorded in every area that isn't a restroom, where if you're not a US national your fingerprints taken, and they literally scan you beneath your clothes? You're worried about your privacy _there_?
A federal agency does that passport scanning and id check, I didn't expect them to share the same information with every airport agency in the country.
I'm fine with the TSA scanning my passport, and the airport filming me. What I'm not so fine with is the airport also having my photo and identity, so now they aren't just filming me every place I walk in the airport, but they are also correlating that to my identity.
They have my picture, but they don't have my name. But when the airports get the passport photo database to implement this photo check-in, then they'll have my name too.
>> The scan takes fractions of a second and has shown to be 99 percent accurate during testing
>> Presumably it's 99% effective at matching random faces to the photos
I think what they mean is that they carried out laboratory tests where a test subject stood facing the system's camera and the system tried to match their face to a database of pictures of peoples' faces, then decided whether the subject's face matched one in its database.
There is a lot that one can do to make this kind of testing give much better results than it should (e.g. use a small database, rather than the entire US passport photo database plus photos of visiting foreigners). This does not presuppose malice- a bit of sloppiness or choices dictated by budget restriction are plenty enough to "spike" the results.
On the other hand, even assuming 100% accurate evaluation one can expect a machine vision system's accuracy to fall way down when it's deployed in the real world, where conditions can no longer be as strictly controlled as in the lab.
So as a rule of thumb, we can probably assume that a 99% accuracy in the lab translates to ~75% accuracy in the wild.
> e.g. use a small database, rather than the entire US passport photo database plus photos of visiting foreigners
Doing so would be wildly inefficient. You need your passport to check in for your flight. When boarding the plane, you don't need to check someone's face against the entire passport database, only against the passports which have been checked into the flight, which is particularly important in terminals with shared domestic/international departures where entry is possible without a passport and the passport can be transferred to somebody after check-in.
But the main reason why it's anyway infeasible is because flights are rarely full of people of a single nationality; scanning peoples' faces against a wide-ranging passport photo database effectively means scanning their faces against some kind of global database of the faces of everybody from around the world. Assembling a database like that, apart from the ethical issues, is politically infeasible.
Remember - this system doesn't advocate for people to not need their passports at all when traveling. They still need them - just not after check-in. The question is whether the system is more or less accurate than a human being at matching passport photos to faces, and based on what we know about racial bias, the answer to that is probably that facial recognition is more accurate by now.
Well, the description of how the system works, in the article, makes it sound like it has a database of passport photos for US citizens, and a database of photos taken on entering the country for citizens of other countries:
>> The new veriScan system developed by the Metropolitan Washington Airports Authority—with guidance from U.S. Customs and Border Protection—scans the faces of travelers approaching the gate. The system then compares the photo to a gallery that includes images of that person—either their passport photo for U.S. citizens or the photo taken of foreign nationals when they entered the country.
So that's definitely not just the passports that have been checked into the flight- foreign nationals in particular may have entered the country at any time and place other than before boarding a flight at Dulles.
Which is why there are privacy concerns in the first place- because it can be used to track the movment of people across multiple trips.
you don't need to check someone's face against the entire
passport database, only against the passports which have
been checked into the flight
Even simpler than that, some UK airports have 'epassport gates' * [1] where you present your passport on a scanner and your face is photographed at the same time, so it only needs to compare your face against your passport.
Sadly the gates are still slow and unreliable, even at that seemingly simple task.
99% accurate doesn't necessarily mean 1% wrong identification. It can merely mean the 1% is a "don't know" indeterminant result that doesn't meet some specified confidence level, for which you fall back to the paper scanning method.
No terrorist organization is going to re-attempt 9/11, because why would you? Repetitions of the Mumbai attacks (notably in Paris) will be the norm, along with mass violence at a low intensity as Muslim radicals simply migrate in huge numbers into Europe and North America.
I agree that we just need to keep a secure level on flights, although I've come to a conclusion over the last few years. That conclusion is that either Terrorists are not smart enough or they just have an agenda that is being pushed by someone specifically on a specific target for whatever reason.
I'll expand a bit. For the past 5 years I've been travelling every 2 weeks using an airplane. Thats mainly within Europe. I've been to airports within Europe, specifically Santorini Airport, Mykonos Airport, Chania Airport where I've seen things going on that really doesn't make sense, nor it makes sense to my on why terrorist groups wouldnt prefer those kind of airports for their attack.
On Mykonos airport 2 years ago, Security check was handled by 2 young police officer in their very early 20's, no supervisor, no metal detectors no nothing. A wheelchair went ahead and they were trying to help the wheelchair to get front and they completely disregarded security. I wasn't searched or anything and I got boarded to a flight going to London. Someone could easily pass any kind of weapon in there. (The airport has since got renovated and prolly it goes with the European standards nowadays).
Santorini Airport 4 years ago to London, again flight going to London, there were so many people waiting and there was only 1 police officer in his late 50's trying to do security checks for hundreds of people. You understand how that system can fail. Again I got no checks on me or anything. Bare in mind both airports I am speaking of had no way to check cabin luggage for anything suspicious, it had to be done by the police officer looking into the bag, and that only was done if he thought that you could be carrying something. I saw him performing it once. (Again this airport has been renovated prolly keeps the European standards now)
Chania airport 1 year ago to Athens, they had around 5 security personel and 1 police officer. A friend of the police officer rocks up with a small hand bag, police officer silently asks him infront of me if he is carrying any firearms with him. He replies yes, gives him the handbag, passes through security and I saw his id he wasn't a police officer but a civilian, the police officer passes the bag behind the counters and proceeds giving it to him after the security check... (Nowadays I dont see police officers around the security but it did happen a year ago)
Thats only 3 airports I've been on and I've seen how easy it is to pass anything in an airplane and 2 of those airplanes were headed to London which is one of the big city capitals and targets. And I am pretty sure there are other places around Europe and other countries that the same thing is happening security wise. And if a civilian like me without resources can see those massive security breaches in a few of the airports he's been to, I am pretty certain that big terrorist groups with massive fundings and people everywhere cause they do have people everywhere, know about those security breaches. It still remains a question to me on why they never used those entrances.
P.S those 3 examples are from Greece, and well thats where I am from and thats where I use to travel a lot (back and forth to London) so thats why my examples are from here. Pretty sure other people have such examples to give.
Presumably it's 99% effective at matching random faces to the photos -- if a terrorist group has a pool of 1000 stolen passports, with a system that's only 99% effective at maching a random photo, I think they could do much better at finding a person (perhaps with some surgical alterations) that fits in that 1%.
The process eliminates the need for an airline employee to manually check every boarding pass and passport while boarding a plane.
So at least they are clear that this is a cost-saving method for airlines, it has nothing to do with speeding boarding or making boarding more secure.
This spring, Lufthansa announced that it boarded an A380 with over 350 passengers at LAX in less than 20 minutes—less than half of their normal time—using self-boarding gates linked to CBP’s facial-enabled traveler verification service,” McAleenan said. “No more fumbling with your boarding pass while you have two carry-ons, maybe a kid; no more trying to find your QR code or trying to refresh your screen.
Oh wait, I guess it is about speeding boarding... but is boarding pass scanning really what slows boarding? I always seem to get backed up in the line at the plane door, waiting for everyone to stow their carryons and sit down. Most busy flights have two agents scanning and most people who have flown more than once scan their pass/phone quickly.... I'm surprised it's any faster getting people to stand in the right spot and stand still for a photo.