Australia Restores Some Sanity to Airport Screening

Welcome news:

Carry-on baggage rules will be relaxed under a shake-up of aviation security announced by the Federal Government today.

The changes will see passengers again allowed to carry some sharp implements, such as nail files and clippers, umbrellas, crochet and knitting needles on board aircraft from July next year.

Metal cutlery will return to return to cabin meals and airport restaurants following Government recognition that security arrangements must be targeted at 'real risks'.

I'm sure these rules won't apply to flights to the U.S., where security arrangements must still be targeted at movie-plot threats.

Posted on December 17, 2009 at 12:54 PM • 32 Comments

Comments

JoshuaDecember 17, 2009 1:50 PM

Actually, nail clippers and metal cutlery returned to US flights some time ago. Sounds like Oz is a little late in catching up on those accounts.

(Though I don't know about umbrellas or knitting needles.)

Jeffrey CharlesDecember 17, 2009 2:01 PM

On a return trip from London to San Francisco recently my wife asked if her knitting needles (long, metal ones) would be permitted. The security folks said "Sure, no problem. But you can't take those tiny scissors" (she uses to cut her yarn). This was expected and she was resigned to losing the cheap scissors.

More of a surprise was the experience of the woman in front of us. She pulled out a box of wooden matches and a lighter and asked the question. Security guy says "hang on, let me ask my supervisor" and turns to a fellow behind him. "No, can't take the lighter. The matches are fine."

At least we weren't required to remove our shoes (though I was on the next flight home a couple weeks later).

jennDecember 17, 2009 2:03 PM

When I was in Australia a few years ago, the screener removed my bamboo size 0 knitting needles from a pair of socks I was working on. Apparently, what amounted to dull, 5 inch long toothpicks were enough of a threat to destroy 2 weeks of patient work.

JeffDecember 17, 2009 2:27 PM

And yet if you shop in the right places you can still buy rather large ominous looking combat knives made out of non-metallic materials, and walk right through the metal detectors.

So while you are turning over your knitting scissors, nail files and umbrellas, the guy next to you might have a 6 inch plastic combat knife taped to his back. (yea yea.. movie plot)

I feel safer already.

feloniusDecember 17, 2009 3:01 PM

From what I've heard from folks I know that travel a lot more than I do, in the US, the security levels are largely dependent on which airport you're flying from. So, Ohio (randomly chosen, I don't actually know) might allow sewing needles, but Nebraska won't. Again, I don't know specifically, but I know someone who travels a fair amount as a coach for a sport, and he said that the levels at different airports can be wildly different.

Ctrl-Alt-DelDecember 17, 2009 3:13 PM

I have encountered the knitting needle paradox several times in the last two years. In one case the woman next to me - middle-aged and kindly-looking - had a pair of wooden needles about 30 cm (1 foot) long. She said she'd never been challenged over them, even when her bag had been searched.

However, those flights were between Australia and New Zealand and in one case case between Australia and Singapore, so we were a long way from places of peak security.

Walt DanielsDecember 17, 2009 3:18 PM

My wife is also a knitter. She has never had any problem with needles, but frequently get questions about some of the other knitting gear she carries. She even had them on flight in Nov 2001 where the only question was fear she might be knitting an Afghan. She regularly has scissors taken, even the blunt ones that are supposed to be allowed. Her circular needles would probably do a good job of removing someones head (as would a necktie) if put around someones neck and pulled tight.

PlonksDecember 17, 2009 3:38 PM

What about water?

They don't allow it yet... and i guess that showing them that you can drink it should demonstrate that it's safe, shouldn't it?

Last week i had to leave a 33cl bottle back because of stupid rules that, i think, only benefit airport shops.

joyDecember 17, 2009 3:42 PM

I'm a crocheter/knitter and I found a pair of still sharp scissors with brightly colored plastic going part-way down the blades... TSA hasn't challenged me since. Apparently, if it looks like it belongs to an elementary schooler, they're ok with it. They've never give me a problem with knitting needles and I've been flying with them since 2002.

JeremiahDecember 17, 2009 3:50 PM

As a knitter, the jokes about using circular needles as a garrote fall solidly under the movie plot threat genre. The amount of force needed to be sustained against a resisting opponent would simply pull the cable out of the fixture on the needle.

I switch to bamboo needles from my more expensive metal needles to allow the tools to A)Look less threatening and B)Reduce my monetary loss if they are taken.

To Jenn:
Before flying the hassle of putting a life line into your live stitches will save those 2 weeks of work when the TSA (or other personnel) confiscate the needles.

ChrisDecember 17, 2009 6:36 PM

@Plonks

Actually, Australia doesn't have any generalised liquid restrictions on domestic flights (just the obvious ones, like corrosive or flammable liquids). So you can take as much water as your carry-on weight allowance can cope with.

We only need to follow the 100ml rule when flying internationally, presumably to keep the US happy.

GeorgeDecember 17, 2009 7:02 PM

The TSA's website on prohibited and permitted items says that "Scissors - metal with pointed tips and blades shorter than four inches" are permitted. It says nothing about knitting needles.

But it doesn't matter. What is allowed depends entirely on the training, understanding, and "interpretation" of the particular officer who screens you. And that can and will vary significantly between airports, checkpoints, and time of day. And in addition to never knowing what rules apply at the moment we're screened, we're supposed to have unquestioning faith that this wildly variable system provides effective protection against terrorists.

What am I missing here?

Clive RobinsonDecember 18, 2009 12:19 AM

@ George,

I'll try and give you an answer to your,

"What am I missing here?"

Belive it or not a question of "probability in action" (and no I'm not being funny).

Think about what you said above your question a little more,

"And in addition to never knowing what rules apply at the moment we're screened, we're supposed to have unquestioning faith that this wildly variable system provides effective protection against terrorists."

You make it sound like it is "Random" in nature.

Which in many respects it inadvertently is.

Bruce and others (including myself) have argued in the past that "random screening" is going to be more effective than "determanistic searching". There are many reasons for this and they are both intuatively and mathmaticaly sound.

However I've also argued (as have others) it has to be done in a very specific way.

This is so a number of reasons,

1, So that people know why it is being done the way it is.

2, Also so it cannot be "gamed" by a sentiant subject group (the criminals).

3, Also that there needs to be certain controls and other procedures in place to prevent abuse.

The problem the TSA have is they are inadvertantly proving the very definate need for the first and third points, whilst very probably not actually stoping the second (which is the whole point of "random searching").

What the TSA's rather silly and arbitary rules and lack of supporting controls has done is leave them open to all sorts of valid argument against their system. Both in it's actual "designed in" effects and in it's "inadvertant effects".

The first thing I should say is that they are not by design implementing a "random search" system. It is the failings of their system that is inadvertantly making it a (very ineffective) "random search".

The second thing I have to say is that after considerable thought I have changed my mind on "random searching", not because it does not work (it does), but because it cannot be implemented correctly with humans (and this may be the view of others now since we don't hear about it much anymore).

Peter A.December 18, 2009 7:11 AM

Clive Robinson:

"random searching [...] cannot be implemented correctly with humans"

What do you actually mean?

1. Random search as a defence against whatever human action (e.g. smuggling) is unimplementable/ineffective or just:
2. That human beeing selecting other human beeings for search does not do it truly randomly.

I can agree with 2. but not with 1.

And if 2. is what you are saying, there are really trivial ways for TSA (or any other organization) to do it properly.

Ricky BobbyDecember 18, 2009 7:57 AM

@ Jeff

Why do creative attack vectors become "movie plots" that should be dismissed?

I understand their is risk issue here, where all risks cannot be mitigated. But is someone taping a non-metallic knife to their back really a movie plot? After all, 9/11 was caused by boxcutters!

I really hate the fact our liberties are being taken because of these far our risks, so I just accept the risk. I understand the likelihood of getting killed on the way to work in my car is far far greater than getting hijacked in a plane. So, I am not really concerned, I wish everyone could understand risk management.

the other AlanDecember 18, 2009 8:15 AM

@Plonks

While they won't let you take a bottle of water through security, they WILL let you take an empty water bottle through. Next time, drink your water, then put the empty bottle in your carry-on. Then, once through security, simply fill it at a fountain. Works every time and doesn't force you to spend $4 on $0.02 worth of water.

Clive RobinsonDecember 18, 2009 8:44 AM

@ Peter A,

""random searching [...] cannot be implemented correctly with humans"

What do you actually mean?"

Humans are sentient and as either the testors or the testees they can manipulat the system either overt, covertly or neglagently.

All truly random systems will at times suffer "queing problems" in that either to many or to few will be selected. In a non sentient system this is not generaly an issue.

Humans on the other hand will observe the system and realise that there are ways to "game" the system as test subjects.

Other humans will see oportunities in the system to excercise prejudice or power again either overtly or covertly.

Other humans have designed rules where the penalties for infringing are so minimal that the system can be gamed repeatedly untill success is assured.

I could go on but in essence if the testors efect the experiment by observation the oposit must also logicaly be true. That is people who will be tested will effect the process by acting on observation.

All though the observation issue could be fixed. It is pointless to do so unless the penalties of transgretion are realistic (taking away a lighter or bottle of water is not a realistic penaltie).

PattiDecember 18, 2009 9:00 AM

I'm a cross-stitcher, and I've been flying (on business, at least 1x/month) with my tiny embroidery scissors for years. They're about 2.5 inches long and have pointy tips. The TSA has had no issues with them whatsoever, but the were taken by security in Canada on a return flight from Ottawa last month. I looked at the guy and said "but the TSA has never had a problem with them." His answer was "well, we're not the TSA." I told him to just go ahead and take them, but it still took them almost 10 minutes to finish combing through all of my things and decide if they were going to let me through at all or not. A woman next to me was allowed through with a full-size (6") pair of rounded tip scissors with no fuss.

uk visaDecember 18, 2009 9:01 AM

Now that Australian authorities have come to their senses let's see how long it takes US and UK authorities to come to theirs... I wouldn't hold your breath though; governments don't give up control as readily as they take it...

jaredDecember 18, 2009 11:00 AM

Australia is already rather sane. I'm able to meet someone at Sydney airport - at the gate - and without taking my shoes off going through security. Volume of liquids are limited per international standards but don't have to get taken out of one's carry-on.

Smaller planes at smaller airports don't have screening. Upon landing at a bigger airport one enters the terminal land-side, picks up luggage and then re-checks luggage and goes through security if transferring to another flight.

An issue in the air is what is a small plane? One airline flies props, another airline wants to fly regional jets of the same size. Using the latter would attract the costs of security because of being jets. So the second airline says it wouldn't be economic to start competitive service.

Peter A.December 18, 2009 11:21 AM

Clive Robinson:

"Humans are sentient and as either the testors or the testees they can manipulat the system either overt, covertly or neglagently."

That is true and obviuos for any screening system. I was rather interested on your opinion on random selection for screening.

"All truly random systems will at times suffer "queing problems" in that either to many or to few will be selected. In a non sentient system this is not generaly an issue."

Well, yes, a local properties of a subsequence of random sequence may not be very random. A "selection density" may be locally higher than the planned average, causing longer queues at the checkpoint and it may be an issue but I don't think it is a big one. Often there are several lanes and if a queue to one is getting longer people will select other one.

"Humans on the other hand will observe the system and realise that there are ways to "game" the system as test subjects."

Sticking to the randomness issue: you cannot game an RNG. Observing the rate of selection for (more thorough) screening won't help the attacker. A probablility of beeing selected would be constant for a long period (until administration decides to change the rate). The truly random selection would have a deterring value of having some chance of being caught that cannot be gamed by behaving or looking innocent.

"Other humans will see oportunities in the system to excercise prejudice or power again either overtly or covertly."

That's the property of a non-random screening system. While profiling is good when implemented correctly (i.e. increases the chance of detecting a violation) it is a very narrow gap between a rational profiling and prejudice...

"Other humans have designed rules where the penalties for infringing are so minimal that the system can be gamed repeatedly until success is assured."

Not entirely true. While penalties for minor infringements are nearly a non-issue for a perpetrator (that's the reason why such minor violations like scissors should be dropped), penalties for major infringements (bomb, gun) are quite severe.

Bryan FeirDecember 18, 2009 11:46 AM

@Ricky Bobby:

"Why do creative attack vectors become "movie plots" that should be dismissed?"

Answer: Because focusing on the 'movie plots' means expending effort dealing with specific, detailed, but highly unlikely cases. This is effort that could almost assuredly be better spent hardening the system in general and doing defense in depth rather than riveting steel plate over the hole that someone imagined was there. Movie plot thinking often leads to Maginot line security, where you may have blocked off one attack well, but left the edges open so the enemy can just walk around.

"After all, 9/11 was caused by boxcutters!"

In part. You can make a better case, though, that 9/11 was really caused by the then-SOP for dealing with hijackers, which boiled down to 'stay calm and let them have what they want, and everybody should get out alive'. Said SOP worked just fine in cases where the hijackers actually cared about getting out alive themselves.

Security systems in general rarely have single points of failure. They have weak spots that can be attacked, sure, but they also tend to have enablers and support structures that allow those weak points to become catastrophic.

RogerDecember 18, 2009 3:26 PM

@Jeffrey Charles:
> More of a surprise was the experience of the woman in front of us. ... "No, can't take the lighter. The matches are fine."

This shouldn't be a surprise. It's a fire safety measure, nothing to do with terrorism, and is motivated by a number of fatal aircraft fires caused by butane filled lighters. For more than a decade ICAO (international civil aviation organisation) has been trying to get all member nations to implement these requirements, and the US was rather late to the show. In Australia, carrying lighters on an aircraft (other than refillable lighters that are empty) has been illegal for many years -- since the mid-1990s IIRC.

Incidentally,
"No, can't take the lighter. The matches are fine."
is not quite true. Butane filled lighters are forbidden altogether (in the cabin or cargo hold.) But under ICAO regulations matches are not permitted in carry-on luggage; they may only be carried on the person.

The reason is that there are examples of matches in carry-on luggage starting a fire under the prolonged vibration of flight, and this spreading to an uncontrollable fire before it was detected. Matches in your pants pocket suffer much less vibration, and even if they should ignite you will know about it while the fire is still small!

RogerDecember 18, 2009 4:59 PM

My wife is also a madly keen knitter -- in fact, she gets agitated if she has to sit around waiting anywhere (e.g. a flight departure lunge!) without her knitting. So she was already well aware of the ban being lifted when I told her -- but apparently it is going to take 7 months to rewrite these regulations, so no knitting for our Christmas flight to see the family!

The Australian knitting needle ban has long been a particular subject of derision for its obvious absurdity (at least to anyone who is actually familiar with knitting needles), and has taught a generation of peaceful craft workers the phrase "security theatre", as well as basic methods for smuggling non-metallic objects [1].

It never made a blind bit of sense. I don't know if it's true, but the belief is widespread that it was motivated by an attack that occurred on a Qantas flight in Australia in May 2003:
http://en.wikipedia.org/wiki/Qantas_Flight_1737
A man attempted to hijack a flight with two sharpened wooden stakes, and stabbed two flight attendants, one of them repeatedly, before being subdued. (The attacker, one David Robinson, was not an Islamic extremist, but an unmedicated schizophrenic.)

However the fact of the matter was that this hijack attempt was utterly unsuccessful. The attacker was subdued in about 20 seconds, the injured persons were not severely injured, and they soon recovered and returned to work. So having proved you can't hijack a plane with large sharp stakes, knitting needles, which are even smaller and blunter, were banned?!?

Anyway, it has long been noted that knitting needles are forbidden, but chopsticks and pencils are permitted (so long as you don't try to use them as knitting needles, which most officials don't mind, but a few pompous asses apparently think makes them look absurd.) The relevant regulations were apparently written by someone with no actual knowledge of knitting needles, because they encompass not just conventional needles but also "circular needles". These are actually two tiny little blunt points (usually made of metallised plastic although some high quality ones are made of rosewood), joined by a lightweight plastic tube. Try as I might I can conceive of no way whatsoever to use these things as a weapon. (Even if garrotting was a practical way to hijack a plane, which it is not, circular needles are certainly far too fragile to use as a garrotte; and in any case, yarn, fishing line, silk ties and Ethernet cables are all permitted !!) Our regulations also made no distinction about the material of the needle: knitting needles made of flexible, lightweight resin were just as banned as steel ones.

So good riddance to this utterly absurd idiocy.

However, sometimes the fiddly details matter ...
sharp-pointed scissors, even ones with 3 inch blades, really are quite dangerous weapons. There is a technique for "weaponisng" these scissors which is fairly well known among criminals, and turns them into a fast, lethal weapon that takes much less training to use effectively than knife-fighting. (That is, a well-trained knife-fighter is a lot more dangerous than a guy with scissors, but if you take two neophytes, give one scissors and the other a knife, and give them both 5 seconds of instruction, the knife guy is a dead man.) So if you are going to allow on embroidery scissors, you might as well let people take combat knives. The same is not true of round-nosed scissors, which are not particularly dangerous.


____
1. See, for example:
http://www.yarnharlot.ca/blog/archives/2007/03/...

GeorgeDecember 18, 2009 7:10 PM

@Clive Robinson: What the TSA's rather silly and arbitary rules and lack of supporting controls has done is leave them open to all sorts of valid argument against their system. Both in it's actual "designed in" effects and in it's "inadvertant effects".

The interesting thing is that the TSA has been consistently successful at masterfully deflecting each and every "valid argument against their system." Their most frequent technique is to simply ignore the arguments. They can get away with it all the time because the lowest level flunkies (who implement the "randomness") have absolute authority to silence any "valid argument." They're trained to use the powerful magic words "Do you want to fly today?" to instantly put anyone who challenges their (arbitrary) authority in their proper place. It works every time, and eventually the bullying will yield a compliant sheeplike public.

And when they can't ignore a "valid argument," the TSA's other successful tactic is to assert the existence of "robust intelligence," the details of which are necessarily secret. If critics only knew the highly classified information, they'd immediately understand both the need for the TSA to act exactly as it does and the reason for its effectiveness. They would then apologize profusely about the unjustified and inappropriate doubt and criticism they've needlessly spouted.

Unfortunately, since all this information is necessarily classified for National Security reasons, we'll just have to accept on faith that there's a valid reason that we can never know for everything the TSA does; and also that all these measures are highly effective against the threats we can never know about. It may *look* random and arbitrary, but if you only knew the horrifying secret nightmare threats for which the TSA's leadership works round the clock in secret devising effective secret protection strategies, you'd be on your knees begging forgiveness for ever having doubted them.

How can anyone argue with that?

Clive RobinsonDecember 19, 2009 9:03 AM

@ Peter A,

"A "selection density" may be locally higher than the planned average, causing longer queues at the checkpoint and it may be an issue but I don't think it is a big one. Often there are several lanes and if a queue to one is getting longer people will select other one."

Yes and this is where the problem starts.

If you assume a true random number generator it will at some point be a significant distance above or below the average.

Now the humans have to decide what to do about this to get it back to the average when high, as the delay has knock on effects for not just the passengers but the airlines (if you miss your gate but your bag is on the aircraft etc).

These rules are distinctly predictable in operation in most cases and hence the reason even a truely random generator will be gameable under certain conditions. Which is why I said,

"Humans on the other hand will observe the system and realise that there are ways to "game" the system as test subjects."

Whilst what you say,

"Sticking to the randomness issue: you cannot game an RNG.

Is true. It is what is done with the output where the problem is which is why,

"Observing the rate of selection for (more thorough) screening won't help the attacker."

May not be true, because,

"A probablility of beeing selected would be constant for a long period (until administration decides to change the rate).

It depends on what they do. If they open new gates without uping the number of screeners than the odds of being caught move more favourably to the attacker.

Worse as the people on the gates are not likley to understand what random is let alone apply it correctly this is when the system gives rise to exploitable human failings. Which is why I said,

"Other humans will see oportunities in the system to excercise prejudice or power again either overtly or covertly."

For instance the RNG rate is above capacity and more than one gate is in use it would be very easy for a brown shirt to make a choice of the next person according to there prejudices or desires, and these may be sub conciouse.

Which is why an otherwise random process ceases to some extent to be random or as you say,

"That's the property of a non-random screening system. While profiling is good when implemented correctly (i.e. increases the chance of detecting a violation) it is a very narrow gap between a rational profiling and prejudice..."

Which brings us around to the question of how do you game the system.

Using the Newtonian scientific method of 'observe, make a hypothosis, test, observe...'

The first two steps there is little you can do about, any passenger or worker in an airport in that area can do this.

It is the third stage you can effect by making "testing" expensive in one way or another which was my point,

"Other humans have designed rules where the penalties for infringing are so minimal that the system can be gamed repeatedly until success is assured."

To which you replied,

"Not entirely true. While penalties for minor infringements are nearly a non-issue for a perpetrator (that's the reason why such minor violations like scissors should be dropped), penalties for major infringements (bomb, gun) are quite severe."

The problem is "recognising" what is critical to an attack plan. Prior to 9/11 they tested to see if box cutters could get through even if found and the answer was yes...

So as I said random testing works it's the humans in the system that make it gameable to an attacker etc etc. So the old saw about "the devil is in the details" applies.

Kermit the BogDecember 20, 2009 4:51 PM

Australia may have started to see some sense with respect to air travel restrictions, but it is set to introduce some draconian net censorship laws.

http://news.smh.com.au/breaking-news-national/...

The government is trying to say that this is about "protecting the children", but in fact the scope is already creeping to include other areas. Even during the trial, innocent websites (such as that of a dentist) was blocked by the filter.

NobodyDecember 21, 2009 5:39 PM

>After all, 9/11 was caused by boxcutters!

No, 9/11 was caused by 40years of experience that hijackings in the US don't lead to anything other than a certain amount of tedious waiting.

Together with a concern that the airline would be sued if anyone stopped the hijackers and a passenger was injured - which meant that SOP was to do nothing.
If a middle eastern gentleman leapt to his feet on an airline today with a box cutter the response from the passengers and cabin crew might be rather different.

But, nevermind it's important to protect against the last threat. which is why we are buying $Bn anti submarine aircraft for Afghanistan.

Ricky BobbyDecember 23, 2009 12:55 PM

This whole thought of "Security Theater" is dumb. Airport Security is doing better at protecting us than nothing. Security is about mitigating potential risks. Its obvious some risks have more potential than others, but this does not mean they can not take place. So, these low risks should probably be acceptad risk. Every secured system has a weakness if someone has the time and money.

According to many people here this airport security is "security theater". Because they do not perform good security, because they mitigate some risks but not others, which may done haphazardly. At the same time, I bet many people here have control over firewalls, IPS, AV, etc... but probably no DLP watching all emails for PII data. So how is this different? This is a major major risk of data loss, but I bet AV is in place everywhere. So does this constitute Security Theater as well? If so, every Security attempt in the World is Security theater, because there is always a way in.

Mike leavyJanuary 15, 2010 2:07 PM

In Canada, before the recent underpants bombing, Allen keys were considered a threat. There were 3 in a kit attached to a tripod that were detected at screening and had to go in checked baggage. I looked them up on the Canadian Transport Safety web site and they are not specifically covered, but Tool Boxes (and their contents such as saws, giant wrenches) are in a general way. I complained and got a response within a couple of days: "Tools are under review". Note the tripod itself would make a substantial, albeit expensive, club.

Leave a comment

Allowed HTML: <a href="URL"> • <em> <cite> <i> • <strong> <b> • <sub> <sup> • <ul> <ol> <li> • <blockquote> <pre>

Photo of Bruce Schneier by Per Ervland.

Schneier on Security is a personal website. Opinions expressed are not necessarily those of Co3 Systems, Inc..