Hamburg, Germany, 26. March 2006. With only 8 of the 16 exits opened, the task for this evacuation certification was to evacuate 853 passengers and 20 crew in 90 seconds. This all happened in darkness, the footage is from infrared cameras.
(via Monkeyfilter)
Very realistic test. No fire, no smoke, no injuries, only able-bodied adults, and everyone on board prepared to participate in the exercise and therefore remaining calm and orderly.
They don’t even have to use dangerous smoke to make it more realistic, just some mild irritant. They also need to have random people in aisle and middle seats who won’t move, simulating injury or infirmity. Geriatrics are always seated on the aisle.
Completely flat and level too. I once saw, on a documentary, some testing that a company did privately – to simulate the rush to get out, they offered a large payment based on escape time which dropped off exponentially. As far as I can tell without having gone through it, it was a pretty good result in terms of realism. People were climbing over chairs, pushing each other, holding other people back etc. In terms of people getting out, not so good. IIRC, they reckoned 3 times as many people die as would do if they could stay calm.
I agree with the first poster. To make this a realistic test airlines needs to use “real” people and not load the test with trained airline employees.
They need to throw in a few elderly people, children, teenagers, folks that barely fit in the seats and those ***holes who you know are never going to leave a plane without trying to retreive their laptops from the overhear bin. And what about those folks in wheelchairs? Ever been to a flight to Palm Beach? Last time I went there were 16 people who needed to be wheeled up the jetway.
These types of tests are pretty much on the same level as the MPG numbers that the U.S. government releases for new cars. You KNOW that your car is never going to get that milage in a real-worl setting because they were arrived at in a controlled environment (45 mph on a track with no AC running).
These kinds of tests are a JOKE. They allow airlines to think that they are prepared for emergencies. When a disaster happens there will be panic — and lots of it.
The airlines are doing everyone a disservice with this travesty.
The real piece they are missing is luggage. You know in a real flight situation half the people are going to be ignoring the instructions and trying to get their bags out of the overhead compartment.
They should pay them X based on how quickly they get out and add a multiplier if they have their bag(s) with them.
Not a joke at all. A baseline. This isn’t an advertisement, it’s an engineering test for certification.
If I were an engineer trying to figure out how many exits I need to make the A380 roughly as egress friendly as a 757, for instance, I need to have a baseline for comparison. If the test for the 757 is under particular conditions, I’d need to replicate those conditions on the A380.
To put this another way, as an engineer, how would I design a *reproduceable* egress test plan to take into account injured people, panicky idiots, slow folks, and smoke? I would think the results of such a test would be all over the map, and while that scenario may be realistic, it’s not much use to a test engineer.
You’re all treating this like it’s some kind of scam. Yeesh!
It is a scam. Ever been in an “incident”? I have. Gear-up landing. Not the most fun you can have in an evening.
Yes, this was designed as a demonstrative test, but the results are meaningless. It in no way simulates anything like that which the occupants experience in any situation requiring emergency egress.
Dave wrote, “If I was an engineer,” meaning he’s not one. If he was he’d know that when it comes to aircraft, engineering takes a back seat to marketing. Most emergency exits in aircraft are narrower than the minimum legal width of a door in a private house, a standard determined by fire code. This is done to keep the exit rows as shallow as possible in order to fit more seats in. In order to keep weight down (and reduce engineering problems), most emergency exits don’t simply open outward but rather require someone pull them inside the aircraft, another violation of basic ground fire safety code.
The majority of aviation-related deaths were caused by smoke and fire. Fire’s a special beast which induces an incredible panic. It’s what causes hundreds of people to try and jam through one emergency exit while ignoring the other free surrounding exits. This test only showed that 900 calm people can quickly exit a plane.
The only similarity to a real-life emergency was that the aircraft was dark. There were no sick people. There were no infirm. There were no people trying to grab their belongings. There was no confusion. There was no smoke. There was nothing which might cause discomfort. There were no shoes under the seats. There were no bags blocking the floors.
Dave, introducing some reproducible reality factors is rather easy. You can run the tests with volunteers who don’t realise that the test concerns exiting the plane. You tell them it’s a seating comfort study, then without warning kill the lights, pump in some benign but slightly irritating smoke, then watch what happens. No psychological study is valid when the participants know what’s being tested, and panic reaction is much more a psychological than physical test.
Are you a politician? He wrote “If I was an engineer trying to figure out…”, which could equally well mean he is an engineer, but that he’s not one who’s trying to figure out that particular problem. I don’t know if he is or not, but I’m not going to jump to a conclusion based on flawed logic arising from a quote out of context.
Dave: Baseline testing makes sense to me, I hadn’t even thought of that. I’ll go and get my pointy ‘D’ hat.
Fair enough. I may have parsed his sentence incorrectly. However, he qualifies his expectations.
That “all over the map” is the realistic expectation. Human panic behaviour is irrational and unpredictable. It’s not reproducible. Reproducibility is important to the overall system, not a single subset.
As I’ve explained, this test does nothing to demonstrate the likely egress in the event of an emergency. The only way this could be considered a “baseline test” is if you’re interested in how fast people can get out in a non-emergency situation, in which case switching off the lights is nothing but a deviation which also negates the results.
If you want to get “statistically reliable” results, you’d need around 27,000 volunteers, a rather unlikely proposition. You’re therefore limited in how you can test. Since it takes close to a thousand people to run one test and you therefore probably have only one shot, which of the two following scenarios do you think would give you any useful information:
1) “Hi. Please take a seat. Once everyone’s ready, we’ll turn off the lights and see how quickly you can get out in an orderly fashion. OK, ready… and … now!”
2) “Hi. Please take a seat. OK, we’re here today to study your comfort levels in this new Airbus A380. Just relax, read a book, do wha… OHMIGOD!!! FIRE! FIRE! GET OUT! GET OUT NOW!!!”
In both scenarios there are almost 900 people in the plane. In the second scenario, people are much more natural and behave much more like plane passengers. There’s your reproducibility. It mimics real life. The screaming of instructions reproduces the initial shock and discovery which is found in real life. The reactions of the unprepared reproduce real life situations. The test as shown in the video is useless from an engineering as well as a safety standpoint.
I don’t agree, Dave learned me good.
The second, as you say yourself, would seem likely to give you unpredictable results unless you use an impractical amount of people.
The first though, would give you a baseline from which you could conceivably use other data to extrapolate meaning, even if the test by itself were meaningless. If the plane you’re testing, plane A, has an identical baseline result to plane B, and there is real-life crash data for plane B, it would be a pretty good starting point assume that plane A’s real-life result would be similar to plane B’s. Not ideal obviously, as there may be things in plane B’s layout that cause no hinderance to a calm crowd, but that becaome apparent when people start fighting their way out.
To be honest I don’t see why they can’t do both, as you suggest they can’t. The panic test would help highlight any deficiencies that don’t show up in the comparison between baselines and refine the extrapolation.
Actually, I’m educated as an Aerospace Engineer (BS, Iowa State University, 1986) and currently working as a Systems Engineer for a Japanese electronics company.
Just to set the record straight.