EVERYBODY OUT: As pedestrians throng near an exit, they slow one another's progress. A new study suggests that judiciously designed obstacles might reduce such conflicts.
Image: ©iStockphoto/Lars Christensen
-
The Best Science Writing Online 2012
Showcasing more than fifty of the most provocative, original, and significant online essays from 2011, The Best Science Writing Online 2012 will change the way...
Read More »
If you find yourself in a crowded building during an emergency, perhaps the last thing you want to see is an obstructed exit. But a new study by a group of Japanese researchers shows that wide-open exits are not always the most efficient at speeding pedestrians through. A judiciously placed obstacle, such as a column, can actually reduce bottlenecking and evacuation times.
Daichi Yanagisawa, a graduate student in the School of Engineering at the University of Tokyo, and his colleagues examined various ways of reducing conflicts—friction, essentially—between individuals as they try to squeeze through an exit.
In research set to appear in the journal Physical Review E, Yanagisawa and his co-authors tested various theories using a model incorporating both the friction of conflicting pedestrians and the slowing effects of obstacles that they must circumnavigate. The researchers also ran evacuation drills with 50 human subjects working toward a narrow exit.
The research team found that the problem with wide-open spaces in front of exits is that evacuees can approach from all sides, allowing the maximum number of pedestrians to enter into conflict at the exit. Reducing exit access with an obstacle can pare down the severity of those conflicts. "When a proper obstacle is set up at an appropriate position in front of the exit," Yanagisawa says, "it blocks a pedestrian moving to the exit and decreases the number of pedestrians moving to the exit at the same time."
Not just any old obstacle will do, of course. Yanagisawa notes that its size, shape and orientation has to be tailored to the space and to the width of the exit itself. In general, however, off-center obstacles are more effective than those placed directly in front of an exit. "When an obstacle is set up at the center of the exit," Yanagisawa says, "it makes pedestrians detour and slows down the evacuation." An obstacle placed to one side, on the other hand, "decreases the probability and the impact of conflicts without making pedestrians detour a lot," he explains.
Andreas Schadschneider, a professor at the University of Cologne's Institute of Theoretical Physics in Germany who has studied pedestrian and traffic flows, says that the obstacle effect has popped up occasionally in the literature for several years. But the new work is, to his knowledge, the first time it has been empirically tested. "So, it no longer remains a vague theoretical prediction, but should be considered a serious effect that might be used for substantial increase in the safety of sports arenas and other large public buildings," Schadschneider says.
He adds that even empirical tests with human subjects cannot fully replicate the conditions of an evacuation, however. For one thing, real-world evacuees might be discouraged by the appearance of an obstacle and retreat toward a different exit, which could negate the obstacle's benefits. Real evacuation simulations are extremely difficult to carry out and can be extremely dangerous, Schadschneider says, pointing to a 2006 test of Airbus's A-380 aircraft in which one subject suffered a broken leg and dozens of others sustained lesser injuries.
The researchers also note that their model does not yet compensate for the intelligence of real-life pedestrians, and Yanagisawa cautions that the new results do not imply that building managers should start blockading fire exits. The findings, he says, merely suggest that researchers should start to "consider installing obstacles as one option which holds potential to shorten a total evacuation time."
See what we're tweeting about
17 Comments
Add CommentWill these findings be shared with managers of large arenas that may benefit from this? Earlier this year U visited the Mets mew statdium, Citifield. Most of the fans leaving the level and side of the stadium I was on were all exiting via one major stairwell. The entrance to the stairwell was incerdibly crowded and I could see how it could easily become dangerous in an emergency sutuation. I was also recently at Orchard Beach in the Bronx. I was waiting with a relatively large yet orderly crowd that was waiting to board the public buses. All was well and good until the police officer posted at the entrance to the bus depot was called to an emergency. All hell broke loose and the beach-goers at the end of the line tried to jump the line. So, in addition to physical guides, the presence of someone with authority is probably also critical.
Reply | Report Abuse | Link to thisI hope this will be applicable soon.
Reply | Report Abuse | Link to thisPossibly something along the lines of lanes would be helpful, separated by railings. This prevents people from clumping together like bricks and even if people run, they have a low chance of being stampeded by a crowd.
Reply | Report Abuse | Link to this"The researchers also note that their model does not yet compensate for the intelligence of real-life pedestrians"
Reply | Report Abuse | Link to thisso true, people make safer and more efficient ways of dong things, and the universe make better idiots.
this could also do more harm than good if used incorrectly, more research definitely needs to be done before it is put into widespread use
Reply | Report Abuse | Link to thisFire inspectors will never go for it. Ok, never is a long time. It make take some time if this is practical for it to be adopted into fire codes.
Reply | Report Abuse | Link to thisSince emergency exits are meant to be used in life or death situations they need to be developed to accommodate a crowd of people using their only their reptilian brains. Simulated emergencies don't provide a plausible life or death scenario, (which seem to short circuit higher cortical functioning) I'd suggest that a high fidelity simulation with "me first" minded actors would produce a much better estimate of human behavior in an emergency... In fact, since a selfish actor is only trying to move from a higher density space to one of lower density, average human behavior could likely be approximated by a molecular model. Obstacles that promote laminar flow would likely speed traffic through the exits...
Reply | Report Abuse | Link to thisDiagrams of the proper placement would be appreciated.
Reply | Report Abuse | Link to thisExit doors have a maximum throughput. Can a better throughput be achieved by organizing the pedestrians in some way before they reach the exit? (Worst case, imagine two fat people trying to jam through the door at the same time.) This might slow the throughput of the area before the exit door, but that is not the gating factor of the throughput of the system (getting people out of the burning theatre using the available exits.) Laminar flow, like ufologist said.
Reply | Report Abuse | Link to thisThe off-center observation is similar to the fact that poles that prevent automobiles from entering bike trails ought to be placed in odd numbers, otherwise you're funneling bicyclists going in opposite directions into the same path.
This "research" isn't just garbage... it's dangerous and deceptive garbage.
Reply | Report Abuse | Link to thisWhat a brilliant idea to "study" emergency exits without modeling emergencies! Would it make any difference if you told the subjects of Daichi Yanagisawa's ridiculous experiment that the last ten people out the door would be burned alive? Who knows? Who cares?
This silly "research" will be published in Physical Review E, and nothing else really matters. "Look, Ma, they published my paper. I may get a job!"
But wouldn't it have been a good idea to model a few catastrophic possibilities? What happens if somebody falls down in the exit? What happens if ten people fall down?
These simple questions encapsulate the only important issues involved in designing exits from arenas and other large public buildings. What's the worst-case scenario? In the worst case, how many people die?
But of course these situations cannot be realistically modeled under existing guidelines for research with human subjects in the United States, and probably not even in Tokyo, so in order to understand the situations which this ludicrous "research" pretends to address, the only reasonable guide is analysis of crowd behavior at emergency exits during actual emergencies.
This junk from the University of Tokyo should be discarded from Physical Review E, and if the gear-heads who run that review are absolutely determined to study human interactions in critical situations, instead of restricting their activity to subjects where they won't endanger the public...
Forget it! That review can't be fixed, and it's more useful to consider how useful research in this field might be conducted, and there are already plenty of useful paradigms available, although they don't really fit the usual guidelines for grant-worthy research in the physical sciences.
How do we know anything about crowd behavior at exit-points in emergency situations? By analyzing the records of actual emergencies, and those records could be significantly improved by installing (well-insulated and well-protected) video cameras around the exits of actual arenas and other large public buildings, and analyzing what actually happens.
In particular, data about exactly the same subject as the subject of this ludicrous gobbledegook from Physical Review E would be easy enough to obtain, since obstructed exits in public buildings are all too common, and if any sort of obstruction improves the efficiency of exits in emergencies, observation of real-world emergencies is the only reliable way to find it.
HEHE
Reply | Report Abuse | Link to thisgood
Reply | Report Abuse | Link to thisI feel that this study really lacks a complete consideration of a true emergency. While an orderly evacuation will no doubt go faster then a disorderly evacuation, the obstacle that is placed is meant to detour people away from the exit and thus reduce it's traffic flow.
Reply | Report Abuse | Link to thisIn other words by reducing traffic to that exit you inherently impose order on the exit which will speed evacuation times. The real flaw to this thinking is that you do not choose your exit in an evacuation, it chooses you.
In the event of an emergency there stands to be a good chance that some of the evacuation paths may be completely blocked by what's causing the evacuation leaving you with very limited paths of escape. This being the case you have no choice in the matter but to choose an exit that is available to you no matter how crowded it might be.
Under these conditions adding an obstacle will only further reduce order because you will still have the same amount of people trying to go through the same exit except now there are planned obstacles in your way inhibiting your path even more as you try and get out.
I'm afraid that I don't believe this strategy can speed up life-threatening evacuations, just the opposite it will impede evacuation times. The only way to make the evacuations faster are to add more places to evacuate from and/or increase the size of the evacuation doors.
Interesting results. Since other studies have shown that large groups of people tend to flow like water, my guess is that the optimum configuration is probably the same as that for water: a trumpet-shaped funnel.
Reply | Report Abuse | Link to thisthey should install vats of KY Jelly to dump on panicked crowds to help them slip by each other easier
Reply | Report Abuse | Link to thisPerhaps a series of experiments using animals can be designed to test the speed of evacuation under different exit-approach configurations. Some way of adapting the "aspect ratio" of the chosen animals (such as by some sort of prothesis) to resemble that of humans should be used --and that could be used to immobilize some of the animals (such as by activating an electromagnet under the floor) to simulate fallen or injured people.
Reply | Report Abuse | Link to thisI saw that evacuation drill of an A380 aircraft and it was not a test of a real life evacuation.
Reply | Report Abuse | Link to thisAll the participants were applicants who wanted to take part in it. They were all young and fit , had a vest with their number on it.
They all knew what to do and it was not an emergency , they well expecting it and were just waiting for the start signal.
In a real emergency , there would be stunned , injured people , mothers with babies , toddlers , old or infirm and all
very frightened . With smoke and probably flames , with the aircraft very likely at anything but straight and level.
Getting fit , primed young keen people to evacuate in 90 seconds pleased the manufacturer , it means nothing in the real world.
Bon Voyage and safe flying.