Slashdot Mirror
Obstacles Near Emergency Exits Speed Evacuation
BuzzSkyline writes "Despite fire codes that require emergency exits be clear of obstacles, some types of obstacles actually speed evacuation. The counterintuitive conclusion resulted from a series of experiments performed at a TV studio in Japan. Researchers from the University of Tokyo asked 50 volunteers to exit the studio through a narrow door. Video tapes of the experiments show that people made it out quickest when a pole was placed about 30 degrees to one side of the exit. The lead researcher believes an obstacle reduces jamming and friction among people in crowds by decreasing conflicts as the crowd presses toward the exit. A paper describing the research is scheduled to appear in the journal Physical Review E in September, but a preprint is available on the Physics Arxiv."
Keep your eyes open and you'll see plenty of real world applications of this principle already in place.
The University of Tokyo seems to have a research group for everything these days...
It's the woman on the pole that's causing the premature evacuation
I wonder if those volunteers were realistic enough.. They should have set the place on fire to see some face stomping, and in the long run maybe save lots of lives..
People act very irrationally when they are afraid of being burnt alive for some reason.
It's shocking that anyone in this day and age still finds it surprising when scientific experiments produce counterintuitive results. So-called intuition and common sense are usually nothing more than widely held but unquestioned assumptions. That people involved in software as much as Slashdot readers and contributors should be surprised is even more absurd. We ought to know well that intuitive interfaces are really familiar interfaces; the only really intuitive interface, as some wit once remarked, is the nipple.
In any case, knowledge unverified by scientific experimentation is not knowledge at all. If there is anything surprising here, it is that we made it all the way to 2009 before someone thought to conduct experiments on a matter as important to public safety as emergency exits.
Proud member of the Weirdo-American community.
About 5 years ago when I was working with researches of behavior models and genetic algorithm (and others) based simulations, I was shown a demonstration of this very phenomenon. It wasn't a stick, or at a 30 degree angle, but a table placed in front of the door (like a reception desk at the entrance of an event).
Without the obstacle, people would rush to the door at once, creating a jam, slowing the actual exiting process. With the desk in the way, the people were forced to move to either end of the long table before they could get to the exit, in effect, creating 2 distinct exit points which worked to smooth the flow of people.
I can't remember if they said that it was based on a real life experiment or not, but it was actually very convincing.
but small exit ways can lead to death like what happened at the e2 nightclub.
http://en.wikipedia.org/wiki/2003_E2_nightclub_stampede
http://www.cnn.com/2003/US/Midwest/02/18/btsc.flock/
...but I remember reading articles about this in Scientific American or Discover years and years ago. I think research into this happened back in the late '80s, after incidents like the NYC "Happy Land" fire, where it was found placing a "grid" of poles near the exits would actually prevent people from stampeding, trampling, and blocking the exits. (Yes, I know it's Slashdot... so I'm not surprised it's actually old news...)
Personally I think it would be most useful to model humans :\
The biggest issue with a real emergency situation is panic. People being squished against fences, walls and other obstacles because there's too many people behind squeezing, making it more dangerous and less efficient. Same is really for people being trampled, it's very dangerous and almost impossible to help someone being trampled back on their feet in such a crowd for the risk of not getting up yourself. I'd be very careful placing obstacles which might lead to more well-behaved behavior in scientific tests (left, right, left, right, that's so much better) but would be very danerous in a real panicking crowd.
Live today, because you never know what tomorrow brings
From TFA:
Yanagisawa said that the next step is to program models of people intelligent enough to self-organize into a line.
Why bother with imaginary beings that don't actually exist?
These guys already figured this out several years ago. (Sorry, I couldn't find a non-subscription link.)
During the pilgrimages to Mecca, one of the things that people are supposed to do is go into a large stadium and cast rocks at three pillars. Zillions of people attend this event, and there have been numerous trampling deaths at the entrance to the stadium. These guys showed that having obstructions near the entrance improves traffic flow, and so they recommended to officials in Mecca to install such obstacles there, resulting in far fewer trampling deaths near the entrance. Other means of traffic calming were used to mitigate deaths elsewhere in the stadium.
and research the big shopping spree the day after thanksgiving when people trample each other trying to get in to the local Walmart when the doors just open
Politics is Treachery, Religion is Brainwashing
There's various places in fluid dynamics where 'obstacles' are put to improve flow aren't there? Those cone shaped things in jet engines for instance (and falcon's have similar cone shaped things in their nostrils.) Maybe this is like that.
In theory, theory and practice are the same; in practice they're different. (Yogi Berra & A. Einstein)
Think of it as impedance matching.
Welcome to the Turing Tarpit, where everything is possible but nothing interesting is easy.
Inconceivable!
sigs... don't talk to me about sigs....
How accurate can simulating an evacuation be? You loose the urgency of knowing when the building is collapse or when you are about to be turned into roast pork.
I am trying to figure out whether any of following is true : Too small a sample set ? How about ~50000 people getting out of a large stadium ? Was there something to do with the obstacle placed near the exit that led people to exit faster ? What about if there is a large piece of rock near the exit. ? What if there is a row of lights that guide people to form queues near the exit ? Would that be faster than this ? I think the study was conducted in the right spirit, but the conclusions cannot be easily drawn. To simply dismiss intuition is not the correct way to go about it. To say that an obstacle placed near the exit makes people exit faster is not the best way to put out the conclusion IMHO. How about an obstacle placed near the exit tends to reorder people which can lead to faster exit ?
During the experiment, the team also found that people exiting in a single-file line were by far the most efficient. Yanagisawa said that the next step is to program models of people intelligent enough to self-organize into a line.
It would also been interesting to see if a few spoilers can break the flow. (As in the onset of turbulence in a fluid?)
It would be ridiculous to see this research as an indictment of fire codes as the summary (and GP) imply. OK, by tuning the exhaust manifold you can decrease backpressure a bit. That's nice. Maybe even counterintuitive. But it's not what fire codes about, which is stopping people from piling junk in front of emergency exits that blocks people exiting, which is far more significant than this.
The building codes try to increase exit width to handle higher traffic flow. The reality, at least as suggested by the research, is similar to what landscape architects have known for generations: people walk faster on a narrow sidewalk than a broad one.
In an emergency, you hit the maximum carrying capacity of any pathway. The key to evacuating a densely occupied space is to convince people to spread out to multiple different exit points, which is confusing in an emergency situation.
I don't think anything is perfect, but when people approach a single door from a number of different angles optimum traffic flow doesn't happen.
but what about disabled people? I would think that the obstacle would cause big probles with wheelchairs and walkers.
They are not suggesting making the exit smaller, they are suggesting that an obstacle is placed further from the door to reduce the number of paths to the door and keep the number of people trying to push through the exit at any given time to a minimum. See Fig. 18 in the arXiv paper if you want to look at a diagram of this.
Interestingly enough, these results seem to have been known for a while (probably based on anecdotal evidence). I distinctly remember my fluid mechanics teacher telling our class almost exactly the same thing in 2006, explaining that a crowd headed for the exit behaved in similar ways to a fluid trying to pass through a small opening.
I came here for a good argument
This instantly reminded me of a roundabout, which also helps with congestion and crashes.
Which makes it obvious where to place such poles.
And where to put some girls on them. :P
Any sufficiently advanced intelligence is indistinguishable from stupidity.
Would a Russian or Italian be as effective?
Isn't this well known? I've seen TV documentaries with computer animations showing the difference between a crowd at an exit and a crowd at a partially blocked exit. Against intuition, the partially blocked exit allows more people to escape in a given time.
My first thought was that Bernoulli (one of them :-)) is smiling in his grave.
Ezekiel 23:20
The weird thing is that people who actually design stuff for crowd control have known this since at the least the 1980s. The goal is to get people ordered into efficient lines heading towards the goals and make sure people understand the process is fair and nothing is to be gained by jumping lines. For a real world example, see Heathrow's newer terminals versus its older ones, or any third world airport: if you make it easy to cheat by changing lines, and other people can see you do it, you get a mob in short order. So, keep lines narrow, and hard to switch from one to another, and people move faster. That means barriers - big ones. Just think Disneyworld, airports, good stadia.
so you're saying people could exit faster if they all spun around the door like a whirlpool?
Oh, and design the exit assembly areas so as to encourage dispersion from the final choke point at the exit: ideally, have the exit open to an amphitheater like shape so people will walk/run downhill/in various directions. Add attractors to get them away from the choke point fast: like, big sign advertising free beer 100 yards off to the side (seriously.)
RMS Titanic third class passengers had class-separation obstacles to contend with, locked barricades and gates to climb, to get to the boat deck, the survival rate among 3rd class males was 10%, and among 3rd class female passengers 25%.
Whereas the survival rate was 50% for male 1st class passengers and nearly 100% for female 1st class passengers.
Now I don't want to make any generalizations about obstacles, but be careful... unless thought out really well, the obstacles may do more harm than good in a real disaster.
A pole might stop people bumping into each other in a crowded room, but in a less-crowded room there's a risk, someone running or walking about accidentally crashes into the pole and injures themselves.
It should be easy to put a camera up and time people for a week, then install a pole and time them again. Have an intern count the number of people each time as well as speed.
excitingthingstodo.blogspot.com
OK, quick now, which will have more vehicles per hour getting through:
6 lanes of traffic squeezing down to 1 all at once, or 6 lanes of traffic taking away 1 lane every mile for 5 miles. Assume the traffic is light for 6 lanes but too much for 1 lane to handle without people waiting their turn, approximating a moderately-full venue when the fire alarms go off.
Why is the latter faster?
At each place where 1 lane is taken away, drivers in the lane going away and the adjacent lane take turns going ahead, with the other lanes going as fast as the traffic ahead will let them. With only 2 lanes contending for 1 each mile, things slow down but not nearly as bad as when 6 lanes suddenly shrink to 1, and all that taking-turns overhead rears its ugly head at a single choke point.
If you notice when you approach highway construction where more than one lane will disappear, they take away the lanes in stages, usually with some distance between each lane removal.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
I'm reminded of one airplane evacuation study where everyone exited the plane in a nice orderly fashion. Then they repeated the same study but paid the people based on the order they exited the plane. Let's just say the results were different. People climbing over seats pushing each other out of the way... Gov study in PDF
In any case, knowledge unverified by scientific experimentation is not knowledge at all.
Then I guess we don't really know that there infinite number of primes, or that that the one millionth digit of pi is 5, or that the American civil war happened, or that ... well you get the point.
Knowledge comes in many forms. Experiment is only one way to obtain it. It is a very powerful way, but it does have its limits and it most certainly isn't the only way.
The situation you're describing sounds like liquid heading downward. I'm pretty sure most doors aren't made in the floor. But if there was a pit, it might be fastest for people to jump into it from multiple directions.
Yes, the OP's correct this is well known. The programme I saw (again, in the UK) showed the effect of having a building's column designed near a fire exit. Everyone's first thought was "criminally stupid", until it was explained that this reduces the occupants' ability to crowd the exit, thus reducing the pressure from weight of bodies (live ones) thereby allowing more people out - rather than jammed in the exit.
politicians are like babies' nappies: they should both be changed regularly and for the same reasons
Nothing new. I recall reading a software review 10-15 years ago of a system that would review/simulate evacuations of large venues. One of the issues was crowd exits and pillars. Adding pillars to a wide door way created more edges, and more people can slip through on the edges of a doorway than those lined up in an orderly queue. The other issues I recall were that people often wanted to exit the way they entered rather than using the closest exit, so signs had to be clear. And people would like to gather their family or group together first, then exit.
"about 30 degrees to either the right or left side of an exit door" ??? Vertically, diagonally? That phrase makes zero sense...
-- Trinity in high heels carrying a whip: The donimatrix - there is no spoonerism
But sounds more like a computer science problem than a physics problem.
But place a big enough obstacle along the route and these people could refocus their navigation around one static object, rather than on the less predictable movement of others in the crowd.
This is pure speculation of course.
I've seen reports (3+ years ago) on simulations using a pretty simple model - people want to move away from the fire but still have some concept of personal space due to crushing being painful - which gave very similar results. Doesn't necessarily mean your hypothesis is wrong, but it certainly makes me wary of complex psychological theories.
I believe the wall is already a big enough static object.
-1 disagree is not a modifier for a reason. -1 troll, flaimbait, redundant, overrated are NOT acceptable substitutes.
You do NOT want people RUNNING DOWNHILL in any emergency situation. That leads to lots and lots of trampling. Someone running quickly hits someone running not so quickly and you have two bodies on the ground. People trip over the downed people and all of a sudden you have a very bad situation.
-1 disagree is not a modifier for a reason. -1 troll, flaimbait, redundant, overrated are NOT acceptable substitutes.
You're quite right, of course. The goal is big fanout + channelling to get the mob density down, plus gentle downhills to nudge the people in the right directions. A fire exit killer is often not the pure size of the exit, it is that that the first 100 people out form a crowd outside and slow the egress of the next 400 people. You can deal with the risks of people running downhill as long as the running reduces bottlenecks and trample risk - it's the difference between panic inside a burning nightclub and people fleeing the scene of a massacre.
Ninja Warrior is a fire drill.
Task Mangler
Very good analogy on the back-pressure.
The "blocking of exits" isn't what architects are curious about, it is the focus on funneling exits rather than just having a certain width offer a constant carrying capacity. Fire codes are tricky, in that they try to balance safety with general application practicality. For the most part (excluding A(ssembly) occupancies), distance separation of exits is focused more on ensuring that a problem in one area doesn't keep all occupants stranded. It works in many applications, but a mob exit isn't really what codes are designed around, for many reasons.
i remember seeing a docco about that a few years ago too. IIRC, the idea was borrowed from observations of exits in ant nests, or something like that.
What next? Bad drivers break up traffic jams? People lining up single file at a ticket booth? Zippers?
I wish the people who built my high school had known this. After a pep rally, I was shocked at how kids were squeezing their way through the doorway in a huge clump. Unfortunately, someone got their foot under my leg and I fell over, and I was stampeded for, oh, a good 40 seconds before anyone realized they were stomping on the back of a person. Then, the principal yelled at me for the incident. Herd mentality at its finest.
You are misinterpreting the research. The study says nothing about narrowing the exit, but instead talks about guiding the flow of people to the exit in ways that reduce congestion. It is similar to better aerodynamics reducing turbulent flow.
This was published in Nature in 2000. It was first discovered using Agent Based Modeling and then tested shortly after.
A) Agreed. What I've seen, even in a non-emergency, people will just wad up physically closest to the door, and you end up with people exiting "the right way" (straight ahead...) and people trying to come in from the sides at the exit simultaneously, lowering exit speed.
B) People are shockingly bad in an emergency. When I was at a hotel and the fire alarm went off, people were all leaving via the main exits. I get ready to open the fire door.. several people say "You can't open that, it's a FIRE DOOR." I pointed out "well, the fire alarm's going off." "Oh yeah that's right!!" That's right, people are so trained to not use the fire exit, they were not using it DURING A FIRE. (Actually, turned out someone was blowing cigarette smoke straight at a smoke detector..but anyway.)
Yes, insects figured it out. Here it was done with Argentine ants: http://www.abc.net.au/catalyst/stories/2008/04/03/2207179.htm
I think I know what's going on here.
I used to take a Chinatown streetcar and was always amazed at how the Chinese, at least in this town, had no sense of getting in line or lining up. The streetcar came, they just rushed the thing from all directions. Nobody waited in lines. I thought it was a cultural thing.
Could it simply be that by placing an obstacle near the doorway they forced people to actually line up, thereby proving that lining up actually helps?
You do realize that crowds in Japan act a little differently than elsewhere. Have you ever been shoved into an elevator, down a hallway, or onto a subway? If you're in Japan you have.