Traffic jams without bottlenecks—experimental evidence for the physical mechanism of the formation of a jam
March 24, 2012 8:01 PM Subscribe
The mathematical theory behind shockwave traffic jams was developed more than 20 years ago using models that show jams appearing from nowhere on roads carrying their maximum capacity of free-flowing traffic - typically triggered by a single driver slowing down. After that first vehicle brakes, the driver behind must also slow, and a shockwave jam of bunching cars appears, traveling backwards through the traffic. The theory has frequently been modeled in computer simulations, and seems to fit with observations of real traffic, but had never been recreated experimentally until recently (PDF of SCIENCE). The authors also released video of their experiments which has since been posted to YouTube.
Abstract.
A traffic jam on a highway is a very familiar phenomenon. From the
physical viewpoint, the system of vehicular flow is a non-equilibrium system
of interacting particles (vehicles). The collective effect of the many-particle
system induces the instability of a free flow state caused by the enhancement
of fluctuations, and the transition to a jamming state occurs spontaneously if
the average vehicle density exceeds a certain critical value. Thus, a bottleneck
is only a trigger and not the essential origin of a traffic jam. In this paper, we
present the first experimental evidence that the emergence of a traffic jam is a
collective phenomenon like ‘dynamical’ phase transitions and pattern formation
in a non-equilibrium system. We have performed an experiment on a circuit to
show the emergence of a jam with no bottleneck. In the initial condition, all the
vehicles are moving, homogeneously distributed on the circular road, with the
same velocity. The average density of the vehicles is prepared for the onset of
the instability. Even a tiny fluctuation grows larger and then the homogeneous
movement cannot be maintained. Finally, a jam cluster appears and propagates
backward like a solitary wave with the same speed as that of a jam cluster on a
highway.
Traffic waves previously and very previously
Abstract.
A traffic jam on a highway is a very familiar phenomenon. From the
physical viewpoint, the system of vehicular flow is a non-equilibrium system
of interacting particles (vehicles). The collective effect of the many-particle
system induces the instability of a free flow state caused by the enhancement
of fluctuations, and the transition to a jamming state occurs spontaneously if
the average vehicle density exceeds a certain critical value. Thus, a bottleneck
is only a trigger and not the essential origin of a traffic jam. In this paper, we
present the first experimental evidence that the emergence of a traffic jam is a
collective phenomenon like ‘dynamical’ phase transitions and pattern formation
in a non-equilibrium system. We have performed an experiment on a circuit to
show the emergence of a jam with no bottleneck. In the initial condition, all the
vehicles are moving, homogeneously distributed on the circular road, with the
same velocity. The average density of the vehicles is prepared for the onset of
the instability. Even a tiny fluctuation grows larger and then the homogeneous
movement cannot be maintained. Finally, a jam cluster appears and propagates
backward like a solitary wave with the same speed as that of a jam cluster on a
highway.
Traffic waves previously and very previously
Snark: So the guy doing 40 in the left lane isn't just clueless, he actually believes he is some kind of traffic god?
posted by gjc at 8:12 PM on March 24, 2012 [1 favorite]
posted by gjc at 8:12 PM on March 24, 2012 [1 favorite]
> guy doing 40 in the left lane
You mean, "guy in an exit-only lane, going at the same speed as the driver ahead?" Or "guy who's actually following FHA recommendations of NOT BEING A TAILGATER?"
:)
But seriously, I think he posted the wrong Youtube link. That's my old amateur video, not the ring-road setup discussed in that paper:
http://www.youtube.com/watch?v=Suugn-p5C1M
posted by billb at 8:51 PM on March 24, 2012 [4 favorites]
You mean, "guy in an exit-only lane, going at the same speed as the driver ahead?" Or "guy who's actually following FHA recommendations of NOT BEING A TAILGATER?"
:)
But seriously, I think he posted the wrong Youtube link. That's my old amateur video, not the ring-road setup discussed in that paper:
http://www.youtube.com/watch?v=Suugn-p5C1M
posted by billb at 8:51 PM on March 24, 2012 [4 favorites]
The guy doing 40 in the left lane may well be some kind of traffic god, if he's doing that because he's observed that that is the average speed of cars in his lane. Before dissing him, do a quick check to see whether there's a large and oscillating gap between his car and the one in front of him.
posted by flabdablet at 8:53 PM on March 24, 2012 [4 favorites]
posted by flabdablet at 8:53 PM on March 24, 2012 [4 favorites]
Before dissing him, do a quick check to see whether there's a large and oscillating gap between his car and the one in front of him.
No, he's varying his speed to match the gap, specifically to dampen the oscillation. All he's proved is that if you drive slower than everyone else, traffic will always go ahead of you.
posted by charlie don't surf at 8:59 PM on March 24, 2012
No, he's varying his speed to match the gap, specifically to dampen the oscillation. All he's proved is that if you drive slower than everyone else, traffic will always go ahead of you.
posted by charlie don't surf at 8:59 PM on March 24, 2012
You only have to drive on the freeways in Southern California about a week or so to observe this. The most maddening experience around here is dragging through a half-hour or so of bumper to bumper traffic to discover that the cause of it was a crash in the opposite lane which caused a few people to slow down marginally just to rubberneck--that slowing sets off the shockwave that ends up with traffic at a complete halt.
posted by yoink at 9:01 PM on March 24, 2012 [5 favorites]
posted by yoink at 9:01 PM on March 24, 2012 [5 favorites]
Billb's earlier video, which I think is interesting anyway, even if it's not the right one.
posted by kenko at 9:10 PM on March 24, 2012
posted by kenko at 9:10 PM on March 24, 2012
For clarity, here is metafilter's own billb's awesome video
posted by Blasdelb at 9:11 PM on March 24, 2012
posted by Blasdelb at 9:11 PM on March 24, 2012
the cause of it was a crash in the opposite lane which caused a few people to slow down marginally just to rubberneck
The point of the linked research is that your hypothetical rubberneckers were merely a trigger, not the cause; the cause of jams is traffic density in and of itself.
No, he's varying his speed to match the gap, specifically to dampen the oscillation
That's not what I'm doing when I'm in slow-driver-in-the-fast-lane traffic god mode. What I'm doing is attempting to keep my speed as close to constant as possible. I don't want to dampen the gap oscillation; I want to stop it propagating backward behind my car.
posted by flabdablet at 9:13 PM on March 24, 2012 [5 favorites]
The point of the linked research is that your hypothetical rubberneckers were merely a trigger, not the cause; the cause of jams is traffic density in and of itself.
No, he's varying his speed to match the gap, specifically to dampen the oscillation
That's not what I'm doing when I'm in slow-driver-in-the-fast-lane traffic god mode. What I'm doing is attempting to keep my speed as close to constant as possible. I don't want to dampen the gap oscillation; I want to stop it propagating backward behind my car.
posted by flabdablet at 9:13 PM on March 24, 2012 [5 favorites]
> No, he's varying his speed to match the gap, specifically to dampen the oscillation
No, in my above video I'm trying to "bust" a daily merge-lanes jam which forms on I-5 North near the Seneca exit just S of Seattle. This particular jam was very weird: a single driver could sometimes cause it to dissolve. None of the other rush hour backups act this way, and I've commuted on numerous different routes.
I say "was weird" because it doesn't form anymore, at least during the few times I've been down there. Too many drivers now know the trick. Something similar happened on my other commutes: changing lanes isn't impossible anymore, and I see four or five gap-leaving drivers every day. Back ten years ago I'd only notice a few of us per week (not counting long-haul truckers.)
Also this stretch of highway acts totally different today, because the state govt installed "traffic smoothing" overhead electronic signs with per-lane variable speed limits. This pushed the usual jam back by about a mile, so now it forms near the Columbian exit rather than up at I-90 and the city center. I guess the jam is also smaller, since my commute definitely takes less time than before the signs went online. Article: http://seattletransitblog.com/2009/04/11/what-is-active-traffic-management/
posted by billb at 9:19 PM on March 24, 2012 [5 favorites]
No, in my above video I'm trying to "bust" a daily merge-lanes jam which forms on I-5 North near the Seneca exit just S of Seattle. This particular jam was very weird: a single driver could sometimes cause it to dissolve. None of the other rush hour backups act this way, and I've commuted on numerous different routes.
I say "was weird" because it doesn't form anymore, at least during the few times I've been down there. Too many drivers now know the trick. Something similar happened on my other commutes: changing lanes isn't impossible anymore, and I see four or five gap-leaving drivers every day. Back ten years ago I'd only notice a few of us per week (not counting long-haul truckers.)
Also this stretch of highway acts totally different today, because the state govt installed "traffic smoothing" overhead electronic signs with per-lane variable speed limits. This pushed the usual jam back by about a mile, so now it forms near the Columbian exit rather than up at I-90 and the city center. I guess the jam is also smaller, since my commute definitely takes less time than before the signs went online. Article: http://seattletransitblog.com/2009/04/11/what-is-active-traffic-management/
posted by billb at 9:19 PM on March 24, 2012 [5 favorites]
It has a lot to do with changing lanes to exit the freeway, essentially crossing over three or more lanes in short order. If there were flying off ramps from the left, or fast lane, splitting up in the middle of the overpass (a four-way starting in the middle of the freeway to create double one-ways to either side, they would reduce a large source of disorder. That would eliminate a freeway crossing street though, which in theory could be done by entering the freeway and crossing three or lanes into the fast lane and getting off there. It sounds the same, but the positive tradeoff being that one is expected to go faster to change lanes, not slower. [post script pony request: a metafilter drawing pad for illustrating]
posted by Brian B. at 9:20 PM on March 24, 2012
posted by Brian B. at 9:20 PM on March 24, 2012
Of course, the only way this deliberate slowing actually achieves anything for anybody but me is if everybody behind me is also a traffic god and is also committed to driving in such a way as to take account of conditions more than one car ahead.
This really isn't rocket science. If everybody always obeyed the two-second rule, highway traffic would always go at the maximum speed allowed by traffic density. But people don't do that, because during peak times that maximum speed is well short of the speed that people think they have a God-given right to drive at. And once a critical mass of drivers is tailgating to the extent that the only safe reaction to brake lights ahead is to jam on their own, backward-propagating waves of stopped cars are completely inevitable.
posted by flabdablet at 9:23 PM on March 24, 2012 [5 favorites]
This really isn't rocket science. If everybody always obeyed the two-second rule, highway traffic would always go at the maximum speed allowed by traffic density. But people don't do that, because during peak times that maximum speed is well short of the speed that people think they have a God-given right to drive at. And once a critical mass of drivers is tailgating to the extent that the only safe reaction to brake lights ahead is to jam on their own, backward-propagating waves of stopped cars are completely inevitable.
posted by flabdablet at 9:23 PM on March 24, 2012 [5 favorites]
Awesome traffic simulator which lets you adjust the speed and density to exactly replicate the results above.
You can also demonstrate the benefit of ramp metering - once a jam is established by the onramp traffic volume, the freeway volume has to drop by about half to clear it. You can get far more overall throughput by restricting the onramp volume so that the jam doesn't form in the first place (even at the cost of traffic queuing on the ramp).
posted by dave99 at 9:27 PM on March 24, 2012 [4 favorites]
You can also demonstrate the benefit of ramp metering - once a jam is established by the onramp traffic volume, the freeway volume has to drop by about half to clear it. You can get far more overall throughput by restricting the onramp volume so that the jam doesn't form in the first place (even at the cost of traffic queuing on the ramp).
posted by dave99 at 9:27 PM on March 24, 2012 [4 favorites]
It has a lot to do with changing lanes to exit the freeway
The research above shows that above a certain traffic volume, jams will form simply from random minor fluctuations in speed. The volume of traffic is the root cause.
The take away for me is that we can't fix traffic jams by focusing on lane changing and 2-second rules and blaming rubberneckers. There is the only the hard solution, which is reducing the overall traffic volume through offering alternative modes of transport.
posted by dave99 at 9:33 PM on March 24, 2012 [3 favorites]
The research above shows that above a certain traffic volume, jams will form simply from random minor fluctuations in speed. The volume of traffic is the root cause.
The take away for me is that we can't fix traffic jams by focusing on lane changing and 2-second rules and blaming rubberneckers. There is the only the hard solution, which is reducing the overall traffic volume through offering alternative modes of transport.
posted by dave99 at 9:33 PM on March 24, 2012 [3 favorites]
Just another reason that we should all be letting robots drive cars for us.
posted by KeSetAffinityThread at 9:44 PM on March 24, 2012 [3 favorites]
posted by KeSetAffinityThread at 9:44 PM on March 24, 2012 [3 favorites]
The research above shows that above a certain traffic volume, jams will form simply from random minor fluctuations in speed. The volume of traffic is the root cause.
I mean, yes and no. Volume sets the stage, but the triggers are needed to set off the actual jam. The triggers can't be totally eliminated, but they can be mitigated.
posted by Garm at 9:50 PM on March 24, 2012
I mean, yes and no. Volume sets the stage, but the triggers are needed to set off the actual jam. The triggers can't be totally eliminated, but they can be mitigated.
posted by Garm at 9:50 PM on March 24, 2012
Taking Traffic Control Lessons — From Ants:
posted by cenoxo at 9:57 PM on March 24, 2012 [7 favorites]
When opposing streams of leafcutter ants share a narrow path, they instinctively alternate flows in the most efficient way possible. Studying how ants manage this could provide the basis for a system of driverless cars running on ant traffic algorithms.Go to the ant, you sluggard; consider her ways, and be wise.
"They never get stuck in traffic," said Audrey Dussutour, a University of Sydney entomologist. "We should use their rules. I’ve been working with ants for eight years, and have never seen a traffic jam — and I’ve tried."
People have long been fascinated with the ability of ants to organize colonial activities in patterns as sophisticated as any urban engineer’s megalopolis blueprint. In recent years, scientists have turned ant traffic flows into algorithms applicable to data transmission and vehicular traffic.
posted by cenoxo at 9:57 PM on March 24, 2012 [7 favorites]
Here are more cool links:
Federal Hwy Admin: MERGE PRINCIPLES
http://ops.fhwa.dot.gov/publications/fhwahop09037/principles.htm
1. Go slow to go fast
2. Keep sufficient gaps
3. "Zippering"
So I guess the long-haul truckers had it right all along. Are those mudflaps with Yosemite Sam and "BACK OFF" now recommended by traffic theorists? :)
Nebraska finds that, during congestion at interstate highway construction sites, signs to promote merging at the last minute work best. So if you see someone waiting until the last minute before changing lanes ...that's what we're all supposed to do.
"Time machine" lets commuters in Duisberg DE view future jams and avoid them. But the researchers discovered that, once people were able to see the future, they modified their behavior, which altered the future. Doh.
posted by billb at 10:26 PM on March 24, 2012 [1 favorite]
Federal Hwy Admin: MERGE PRINCIPLES
http://ops.fhwa.dot.gov/publications/fhwahop09037/principles.htm
1. Go slow to go fast
2. Keep sufficient gaps
3. "Zippering"
So I guess the long-haul truckers had it right all along. Are those mudflaps with Yosemite Sam and "BACK OFF" now recommended by traffic theorists? :)
Nebraska finds that, during congestion at interstate highway construction sites, signs to promote merging at the last minute work best. So if you see someone waiting until the last minute before changing lanes ...that's what we're all supposed to do.
"Time machine" lets commuters in Duisberg DE view future jams and avoid them. But the researchers discovered that, once people were able to see the future, they modified their behavior, which altered the future. Doh.
posted by billb at 10:26 PM on March 24, 2012 [1 favorite]
I don't want to dampen the gap oscillation; I want to stop it propagating backward behind my car.
I kind of like damping as a term here actually.. Sure, it is more an active control system than just a passive damper, but it does capture the spirit of the thing. Unlike charlie don't surf though, I'm sure it does actually create the beneficial result you describe.
posted by Chuckles at 10:27 PM on March 24, 2012
I kind of like damping as a term here actually.. Sure, it is more an active control system than just a passive damper, but it does capture the spirit of the thing. Unlike charlie don't surf though, I'm sure it does actually create the beneficial result you describe.
posted by Chuckles at 10:27 PM on March 24, 2012
waiting until the last minute before changing lanes
I have trouble being convinced of this though... The avoiding standoffs and not closing ranks thing is well proven. This though.. If you have a long stretch with no ramps and you have an isolated lane restriction..... Okay, forget it, how about we have a 2 lane example of that simulation to prove this please :P
posted by Chuckles at 10:37 PM on March 24, 2012
I have trouble being convinced of this though... The avoiding standoffs and not closing ranks thing is well proven. This though.. If you have a long stretch with no ramps and you have an isolated lane restriction..... Okay, forget it, how about we have a 2 lane example of that simulation to prove this please :P
posted by Chuckles at 10:37 PM on March 24, 2012
I say "was weird" because it doesn't form anymore, at least during the few times I've been down there.
Still happens, I was stuck in it at 3:40pm last Friday. Anecdata: long-haul trucks will never merge into the express lane exit more than 100 yards before the off-ramp. Get 2 or 3 in succession and the flow is stopped.
posted by N-stoff at 10:39 PM on March 24, 2012 [1 favorite]
Still happens, I was stuck in it at 3:40pm last Friday. Anecdata: long-haul trucks will never merge into the express lane exit more than 100 yards before the off-ramp. Get 2 or 3 in succession and the flow is stopped.
posted by N-stoff at 10:39 PM on March 24, 2012 [1 favorite]
I mean, yes and no. Volume sets the stage, but the triggers are needed to set off the actual jam. The triggers can't be totally eliminated, but they can be mitigated.
The ring road video has no triggers. Other than the inability of humans to precisely behave in the exact same way and maintain the exact same speed; 30 km/h, not 29.8 or 30.1.
Oh, but "Robots!" Robots are cool and will solve our problems for us.
Except that GoogleBMW's response algorithm has a slightly different acceleration curve from DaimerFacebook's; and those YahooYugos! slow more in the presence of unpredictable roadside phenomena, and Apple iCars don't fully support roads without SmartDrive technology, so they don't always merge the same way. So there may always be a level of variation in response, and in high volume situations, these sort of shockwave traffic jams can occur.
Luckily, someone has invented a different technology that allows dozens of people to travel at speed down a road, maintaining a separation of only inches from each other, without ever breaking formation.
posted by Homeboy Trouble at 10:49 PM on March 24, 2012 [8 favorites]
The ring road video has no triggers. Other than the inability of humans to precisely behave in the exact same way and maintain the exact same speed; 30 km/h, not 29.8 or 30.1.
Oh, but "Robots!" Robots are cool and will solve our problems for us.
Except that GoogleBMW's response algorithm has a slightly different acceleration curve from DaimerFacebook's; and those YahooYugos! slow more in the presence of unpredictable roadside phenomena, and Apple iCars don't fully support roads without SmartDrive technology, so they don't always merge the same way. So there may always be a level of variation in response, and in high volume situations, these sort of shockwave traffic jams can occur.
Luckily, someone has invented a different technology that allows dozens of people to travel at speed down a road, maintaining a separation of only inches from each other, without ever breaking formation.
posted by Homeboy Trouble at 10:49 PM on March 24, 2012 [8 favorites]
"... dragging through a half-hour or so of bumper to bumper traffic to discover that the cause of it was a crash in the opposite lane which caused a few people to slow down marginally just to rubberneck."
10,000 accident free miles as a staff driver in Korea. Twice I remember feeling the flames of wrecks on MSR 1 through my window, twice more I saw bodies stretched out on the side of the road because MSR 1 is sort of narrow, twisty and crazy at points and you can come around a corner into traffic that's stopped for the next five miles. I was in a truck (not driving it) that ran into a situation like that. The hood went up over the top of the cab of our HMMWV we hit so hard. Three other people in my unit died from standing between two stalled trucks they could only get halfway off the road after a breakdown. A bigger truck rounded the corner, smashed into them, crushed two to death and decapitated a third.
One of the afternoon announcers on the local NPR affiliate used to practically *hiss* with disgust over all the rubberneckers when she was reading the traffic report. Her voice would get really tight and she'd say stuff like "traffic is slow on southbound 204 because SOME PEOPLE NEED TO SLOW DOWN AND LOOK AT THE ACCIDENT."
Then suddenly — after a year or two of me noticing that the woman was unhinged about all the rubberneckers and getting sort of agitated over her agitation — she suddenly stopped doing it. I was relieved and thought no more about it, because she really harshed my mellow with her carrying on. Sometimes you want the soothing WASP drone of a public radio professional, especially when you're stuck in traffic.
Then, one day I was headed west on I84, rounded a curve and saw some flashing lights up ahead. Because traffic was moving fast and I was coming from around a curve and the accident was, indeed on the other side, but looked like it was in the middle of my lane because of yet another curve, I slowed down. Suddenly, I could hear that announcer hissing in my ear about the rubberneckers and I came to the realization — made years ago in Korea and forgotten after getting to work a nice home office job for 10 years — that slowing down because you're going down the highway at 70 mph and suddenly see flashing lights ahead and feel the need to give yourself some extra room to assess a potential threat is really sort of sensible. And it causes traffic jams. Which might be better than the alternative.
posted by mph at 11:12 PM on March 24, 2012 [5 favorites]
10,000 accident free miles as a staff driver in Korea. Twice I remember feeling the flames of wrecks on MSR 1 through my window, twice more I saw bodies stretched out on the side of the road because MSR 1 is sort of narrow, twisty and crazy at points and you can come around a corner into traffic that's stopped for the next five miles. I was in a truck (not driving it) that ran into a situation like that. The hood went up over the top of the cab of our HMMWV we hit so hard. Three other people in my unit died from standing between two stalled trucks they could only get halfway off the road after a breakdown. A bigger truck rounded the corner, smashed into them, crushed two to death and decapitated a third.
One of the afternoon announcers on the local NPR affiliate used to practically *hiss* with disgust over all the rubberneckers when she was reading the traffic report. Her voice would get really tight and she'd say stuff like "traffic is slow on southbound 204 because SOME PEOPLE NEED TO SLOW DOWN AND LOOK AT THE ACCIDENT."
Then suddenly — after a year or two of me noticing that the woman was unhinged about all the rubberneckers and getting sort of agitated over her agitation — she suddenly stopped doing it. I was relieved and thought no more about it, because she really harshed my mellow with her carrying on. Sometimes you want the soothing WASP drone of a public radio professional, especially when you're stuck in traffic.
Then, one day I was headed west on I84, rounded a curve and saw some flashing lights up ahead. Because traffic was moving fast and I was coming from around a curve and the accident was, indeed on the other side, but looked like it was in the middle of my lane because of yet another curve, I slowed down. Suddenly, I could hear that announcer hissing in my ear about the rubberneckers and I came to the realization — made years ago in Korea and forgotten after getting to work a nice home office job for 10 years — that slowing down because you're going down the highway at 70 mph and suddenly see flashing lights ahead and feel the need to give yourself some extra room to assess a potential threat is really sort of sensible. And it causes traffic jams. Which might be better than the alternative.
posted by mph at 11:12 PM on March 24, 2012 [5 favorites]
>> waiting until the last minute before changing lanes
>
> a long stretch with no ramps and you have an isolated lane restriction.....
These articles don't clearly explain why the late-merge works, just that it does.
Anecdotally (meaning, from observing multiple instances!) I suspect that the problem with early-merging is that a tiny trickle of late-merging drivers can slow things way down. But the opposite is not true. First instance: a few late-mergers in an early-merging population. The drivers in the long queue perceive the late-mergers as cheaters who zoom down to the end of the empty lane. Both lanes then halt, with no "zipper merge." Opposite instance: if instead the signs encourage late merging, things could move faster because any handful of early-merging drivers are essentially invisible. They'd merge at speed and only slightly alter the difference between lanes. Two full lanes means things appear fair at the actual merge point, so a flowing "zipper" can develop.
I don't think any existing simulation would show this. It's road-rage psychology. Let a tiny percentage of simulated drivers become late-mergers who ignore queues and race down the empty lane. Let a large percentage always merge early, but also become aggressive "merge preventers" who become angrier and angrier when sitting in a slow queue and watching an empty adjacent lane with occasional cars racing past everyone else! :)
posted by billb at 12:18 AM on March 25, 2012 [2 favorites]
>
> a long stretch with no ramps and you have an isolated lane restriction.....
These articles don't clearly explain why the late-merge works, just that it does.
Anecdotally (meaning, from observing multiple instances!) I suspect that the problem with early-merging is that a tiny trickle of late-merging drivers can slow things way down. But the opposite is not true. First instance: a few late-mergers in an early-merging population. The drivers in the long queue perceive the late-mergers as cheaters who zoom down to the end of the empty lane. Both lanes then halt, with no "zipper merge." Opposite instance: if instead the signs encourage late merging, things could move faster because any handful of early-merging drivers are essentially invisible. They'd merge at speed and only slightly alter the difference between lanes. Two full lanes means things appear fair at the actual merge point, so a flowing "zipper" can develop.
I don't think any existing simulation would show this. It's road-rage psychology. Let a tiny percentage of simulated drivers become late-mergers who ignore queues and race down the empty lane. Let a large percentage always merge early, but also become aggressive "merge preventers" who become angrier and angrier when sitting in a slow queue and watching an empty adjacent lane with occasional cars racing past everyone else! :)
posted by billb at 12:18 AM on March 25, 2012 [2 favorites]
above a certain traffic volume, jams will form simply from random minor fluctuations ... we can't fix traffic jams by focusing on lane changing and 2-second rules
The conclusion does not follow from the premise.
The thing about the 2-second rule is that if drivers stick to it then that sets a maximum possible traffic density (i.e. a maximum number of cars per kilometre) for any given speed. 100km/h is 36 seconds/km, so if all drivers are obeying the 2-second rule then the maximum density at 100km/h is 18 cars/km which is comfortably under the 25 cars/km that the linked research cites as the critical density for jam formation.
If cars in a given lane are passing a fixed point at 2 seconds/car, that lane wouldn't reach 25 cars/km until it slowed to 50 seconds/km or 72 km/h. That's a pretty comfortable rubbernecking margin.
So it follows that if most drivers took the 2-second rule seriously then that would, in fact, prevent spontaneous freeway jam formation.
The linked research contains a vehicle density vs. flow rate graph based on one month's observation on a Japanese freeway. That graph shows flow rate topping out at roughly 200 cars per 5 minutes at a density of roughly 25 cars/km. This is 8 km per 5 minutes, or roughly 100 km/h - full highway speed.
200 cars per 5 minutes is 300 seconds per 200 cars, or 1.5 seconds per car, which is about what happens in practice when people try to follow the 2-second rule in a stream of traffic that has occasional lane switching and merging.
Road engineers can limit freeway traffic density using tricks like one-car-per-green-light-per-lane entrance ramp gating, but if the majority of us understood that the two-second rule is in our best interests for trip time reasons and safety reasons they wouldn't have to spend our money doing that.
posted by flabdablet at 12:18 AM on March 25, 2012 [1 favorite]
The conclusion does not follow from the premise.
The thing about the 2-second rule is that if drivers stick to it then that sets a maximum possible traffic density (i.e. a maximum number of cars per kilometre) for any given speed. 100km/h is 36 seconds/km, so if all drivers are obeying the 2-second rule then the maximum density at 100km/h is 18 cars/km which is comfortably under the 25 cars/km that the linked research cites as the critical density for jam formation.
If cars in a given lane are passing a fixed point at 2 seconds/car, that lane wouldn't reach 25 cars/km until it slowed to 50 seconds/km or 72 km/h. That's a pretty comfortable rubbernecking margin.
So it follows that if most drivers took the 2-second rule seriously then that would, in fact, prevent spontaneous freeway jam formation.
The linked research contains a vehicle density vs. flow rate graph based on one month's observation on a Japanese freeway. That graph shows flow rate topping out at roughly 200 cars per 5 minutes at a density of roughly 25 cars/km. This is 8 km per 5 minutes, or roughly 100 km/h - full highway speed.
200 cars per 5 minutes is 300 seconds per 200 cars, or 1.5 seconds per car, which is about what happens in practice when people try to follow the 2-second rule in a stream of traffic that has occasional lane switching and merging.
Road engineers can limit freeway traffic density using tricks like one-car-per-green-light-per-lane entrance ramp gating, but if the majority of us understood that the two-second rule is in our best interests for trip time reasons and safety reasons they wouldn't have to spend our money doing that.
posted by flabdablet at 12:18 AM on March 25, 2012 [1 favorite]
I've always thought of these traffic behaviours as compression waves, 'cos that's the name we were taught in physics for that kind of propagating wave. "Shockwave" seems jarring to me - traffic slowing down seems like the opposite of a shockwave to me. :-/
posted by -harlequin- at 1:20 AM on March 25, 2012 [2 favorites]
posted by -harlequin- at 1:20 AM on March 25, 2012 [2 favorites]
billb, i first saw your trafficwaves posts before I had a driver's license (which i only bothered to obtain around age 40, about six years ago) and I have to tell you, your stuff dramatically effects how I drive, if I am driving as mindfully as I wish to. A few years later I remember realizing that you were a Seattle driver, and watching your video upthread tonight for the first time, the northbound express lane exit, makes me realize we have likely driven together.
I seriously cannot thank you enough for this insight. Next time to see a random driver deliberately gapping northbound, just assume it is me, and I am doing it because of you. Thanks.
posted by mwhybark at 1:32 AM on March 25, 2012
I seriously cannot thank you enough for this insight. Next time to see a random driver deliberately gapping northbound, just assume it is me, and I am doing it because of you. Thanks.
posted by mwhybark at 1:32 AM on March 25, 2012
typically triggered by a single driver slowing down
Actually, I know that guy. His name is Robert Crenshaw, and he lives at 2320 Starling Lane.
I've tried to warn him about this.
KNOCK IT OFF, BOB! People are starting to notice!
posted by twoleftfeet at 1:36 AM on March 25, 2012
Actually, I know that guy. His name is Robert Crenshaw, and he lives at 2320 Starling Lane.
I've tried to warn him about this.
KNOCK IT OFF, BOB! People are starting to notice!
posted by twoleftfeet at 1:36 AM on March 25, 2012
I'm so glad I live and work out in the country.
Wait - forget I said that. Everybody...it's terrible out here in the country; very boring. You really don't want to move out here. Honest.
posted by tommyD at 5:48 AM on March 25, 2012 [1 favorite]
Wait - forget I said that. Everybody...it's terrible out here in the country; very boring. You really don't want to move out here. Honest.
posted by tommyD at 5:48 AM on March 25, 2012 [1 favorite]
> merging at the last minute work best
Yes, this should be promoted more. Merging at the last minute means everyone is making full use of the available road. If everyone merges half a mile early, it's no different from if the single lane section were half a mile longer. The only thing that slows traffic down above and beyond the constriction is people blocking merges out of some misguided sense of justice.
posted by lucidium at 6:23 AM on March 25, 2012
Yes, this should be promoted more. Merging at the last minute means everyone is making full use of the available road. If everyone merges half a mile early, it's no different from if the single lane section were half a mile longer. The only thing that slows traffic down above and beyond the constriction is people blocking merges out of some misguided sense of justice.
posted by lucidium at 6:23 AM on March 25, 2012
This is fascinating stuff. I am incredibly happy that I don't have that kind of daily commute, but I think about these things whenever I have to drive through Seattle or another city with traffic issues.
posted by Forktine at 7:01 AM on March 25, 2012
posted by Forktine at 7:01 AM on March 25, 2012
If I merge early and get the hammer down, I can get another lane over and skip the jam created by all the other mergers, as it has filled up the lane to the point where it has stopped, or nearly so. The problem generally is that the next lane over is quite a bit quicker, so it involves really mashing the pedal and getting you speed on in as few seconds as possible. In the case where I do this often, it is exacerbated by multiple ramps merging in, plus an off ramp, so at least three groups of cars competing for one lane. I prefer to be in and out of it as soon as possible.
posted by Bovine Love at 7:46 AM on March 25, 2012
posted by Bovine Love at 7:46 AM on March 25, 2012
I was a long distance commuter for years. Oscillation damping is a well known practice, and I respect it when I see people doing it. The problem is that if they are overzealous, then people will ignore them and fill the gap in front of them. That, and the fact that oscillation damping doesn't save any time, though it does save wear and tear on your brakes. And there is the minor thing that after the last oscillation, the damp-practicer will be left in the dust, waiting for the next wave, which doesn't happen. Not a big deal, but mildly annoying if you are the one doing it.
Wave propagation is the fundamental issue, but a critical thing to remember is what I call side loading. This is what happens at rush hour - the freeway is already loaded to capacity (not bumper to bumper, but freely moving 65mph), then you start injecting cars from the sides (the on ramps) forcing wave propagation at multiple location in the system. Los Angeles, anywhere, at 6 am-ish, you see this at work. Note that volumes will rise earlier the further you are away from the city center (or as I like to call it "the Dragon's Mouth" if you take 101/10/5/110 directly through downtown LA, you are going through the Dragon's Mouth. ) So the further away you start, the earlier you need to start, but it's not a linear time-distance relationship. The downside is, that if you start early enough to avoid traffic for traffic's sake (as opposed to avoiding traffic to be outright late) you will be quite early to wherever you are going.
I found going west to east in the morning, that if I made it into the Valley by 6 a.m. (specifically 101 at White Oak) it was smooth sailing through the city and out into east LA. Five or ten minutes late, and you'd see the other commuters loading the road heavily; the difference to City of Industry would be about a half hour to forty five minutes.
Anyway the secret to commuting in heavy traffic is to go slow enough so that you don't contribute to rapid backward wave propagation. Chill out, get in the number three or four lane (if you don't know how the lanes are numbered, you don't commute enough to need this advice) and let people pass you. It's not a race, and I guarantee you will see 90% of the people passing you again in a few minutes, sitting in their lane as yours compresses more slowly and you pass them. It's like professional poker; you don't play to win every hand or even every time you sit at the table, you play to win over the long term. Turn on the radio, and drive in a way that lets you enjoy life.
Oh and one more Xoebe-terminology thing. I call the number two lane the "Weenie Lane". It's where the weenies drive. They want to go fast, but not fast enough to go in the number one. Or, the number one lane isn't fast enough for them, so they use it as a passing lane. Add in the fact that people are traversing number two to get in and out of number one, and it's a bad place to be. Weenies drive there. Either man up and get in number one, and own it; or get in number three and float lazily down the river.
posted by Xoebe at 7:58 AM on March 25, 2012 [2 favorites]
Wave propagation is the fundamental issue, but a critical thing to remember is what I call side loading. This is what happens at rush hour - the freeway is already loaded to capacity (not bumper to bumper, but freely moving 65mph), then you start injecting cars from the sides (the on ramps) forcing wave propagation at multiple location in the system. Los Angeles, anywhere, at 6 am-ish, you see this at work. Note that volumes will rise earlier the further you are away from the city center (or as I like to call it "the Dragon's Mouth" if you take 101/10/5/110 directly through downtown LA, you are going through the Dragon's Mouth. ) So the further away you start, the earlier you need to start, but it's not a linear time-distance relationship. The downside is, that if you start early enough to avoid traffic for traffic's sake (as opposed to avoiding traffic to be outright late) you will be quite early to wherever you are going.
I found going west to east in the morning, that if I made it into the Valley by 6 a.m. (specifically 101 at White Oak) it was smooth sailing through the city and out into east LA. Five or ten minutes late, and you'd see the other commuters loading the road heavily; the difference to City of Industry would be about a half hour to forty five minutes.
Anyway the secret to commuting in heavy traffic is to go slow enough so that you don't contribute to rapid backward wave propagation. Chill out, get in the number three or four lane (if you don't know how the lanes are numbered, you don't commute enough to need this advice) and let people pass you. It's not a race, and I guarantee you will see 90% of the people passing you again in a few minutes, sitting in their lane as yours compresses more slowly and you pass them. It's like professional poker; you don't play to win every hand or even every time you sit at the table, you play to win over the long term. Turn on the radio, and drive in a way that lets you enjoy life.
Oh and one more Xoebe-terminology thing. I call the number two lane the "Weenie Lane". It's where the weenies drive. They want to go fast, but not fast enough to go in the number one. Or, the number one lane isn't fast enough for them, so they use it as a passing lane. Add in the fact that people are traversing number two to get in and out of number one, and it's a bad place to be. Weenies drive there. Either man up and get in number one, and own it; or get in number three and float lazily down the river.
posted by Xoebe at 7:58 AM on March 25, 2012 [2 favorites]
Where do you think people should pass? The shoulder?
posted by thelonius at 3:32 PM on March 25, 2012
posted by thelonius at 3:32 PM on March 25, 2012
I came to the realization — made years ago in Korea and forgotten after getting to work a nice home office job for 10 years — that slowing down because you're going down the highway at 70 mph and suddenly see flashing lights ahead and feel the need to give yourself some extra room to assess a potential threat is really sort of sensible.
Yeah, sure, if, as you say, you're rounding a curve and there's a mess of flashing lights and you're not sure which side of the road they're on, it's entirely sensible to slow down. But that's a pretty specific set of circumstances. There are lots of times you'll get the rubberneckers when the disturbance is in the opposite direction lanes, on a long straight section of the freeway, where no one has any doubt whatsoever about that, and there are no emergency vehicles involved. I've seen traffic jams caused by a stalled semi sitting six lanes over on the far side of the freeway.
posted by yoink at 7:26 PM on March 25, 2012
Yeah, sure, if, as you say, you're rounding a curve and there's a mess of flashing lights and you're not sure which side of the road they're on, it's entirely sensible to slow down. But that's a pretty specific set of circumstances. There are lots of times you'll get the rubberneckers when the disturbance is in the opposite direction lanes, on a long straight section of the freeway, where no one has any doubt whatsoever about that, and there are no emergency vehicles involved. I've seen traffic jams caused by a stalled semi sitting six lanes over on the far side of the freeway.
posted by yoink at 7:26 PM on March 25, 2012
When an accident happens in the opposing travel lanes, it's completely natural and inevitable that people slow down. This is a reaction to a disturbance that is unexpected and potentially life threatening, and it has to take place in a fraction of a second to do any good at all. Drivers slow down a bit before their brain sorts out that there are no accident vehicles crossing the median, for example. Have you ever seen a traffic accident at highway speeds? It is terrifying.
In busy traffic conditions, this is enough to create the compression wave that slows traffic for some time. People who don't understand this typically blame "rubberneckers", often with disturbingly savage anger. This mainly reflects a need to personalize agency for the inconvenience and annoyance of being stuck in traffic, in my opinion, and is erroneous. You really think people slow down so they can get a good look at a stalled truck?
posted by thelonius at 7:49 PM on March 25, 2012 [1 favorite]
In busy traffic conditions, this is enough to create the compression wave that slows traffic for some time. People who don't understand this typically blame "rubberneckers", often with disturbingly savage anger. This mainly reflects a need to personalize agency for the inconvenience and annoyance of being stuck in traffic, in my opinion, and is erroneous. You really think people slow down so they can get a good look at a stalled truck?
posted by thelonius at 7:49 PM on March 25, 2012 [1 favorite]
This mainly reflects a need to personalize agency for the inconvenience and annoyance of being stuck in traffic, in my opinion, and is erroneous.
I did find myself wondering how the public radio lady knew the minds of the motorists she was excoriating, but I'm a generous soul: Were I to accept the possibility she might possibly be wrong because she was thought she could read minds (which I do not believe is possible), I also have to accept the possibility she'd read some study I hadn't. Maybe with some special Nielsen box wired up to the study group:
Today at 5:39 you were westbound on I84 and decelerated rapidly near mile marker 5, where reports indicate there was a traffic accident in the eastbound lane. Did you:
a. decelerate out of a strong sense of caution
b. decelerate because you were curious and thought you might get to see dead people
c. decelerate because traffic had slowed
d. Other (please specify in space provided):
Coincidence. Spilled frappucino in lap while texting bowling team about slowed traffic
posted by mph at 10:56 AM on March 26, 2012 [1 favorite]
I did find myself wondering how the public radio lady knew the minds of the motorists she was excoriating, but I'm a generous soul: Were I to accept the possibility she might possibly be wrong because she was thought she could read minds (which I do not believe is possible), I also have to accept the possibility she'd read some study I hadn't. Maybe with some special Nielsen box wired up to the study group:
Today at 5:39 you were westbound on I84 and decelerated rapidly near mile marker 5, where reports indicate there was a traffic accident in the eastbound lane. Did you:
a. decelerate out of a strong sense of caution
b. decelerate because you were curious and thought you might get to see dead people
c. decelerate because traffic had slowed
d. Other (please specify in space provided):
Coincidence. Spilled frappucino in lap while texting bowling team about slowed traffic
posted by mph at 10:56 AM on March 26, 2012 [1 favorite]
Well, I may be wrong - I haven't like performed a study on this. But I've come to distrust the rubbernecker narrative.
posted by thelonius at 12:38 PM on March 26, 2012
posted by thelonius at 12:38 PM on March 26, 2012
I should have made clear that I don't trust it either. But leaving room for it to possibly be true made it possible to not spend all my time worrying that the public radio lady wasn't getting desperately needed mental health services.
posted by mph at 12:52 PM on March 26, 2012
posted by mph at 12:52 PM on March 26, 2012
When an accident happens in the opposing travel lanes, it's completely natural and inevitable that people slow down.
Yeah, of course. If you actually see an accident happening in the the opposing travel lanes ahead of you you would be a fool not to slow down. Who is arguing otherwise? Just because one can endlessly proliferate examples of situations in which it DOES in fact make sense to slow down does not mean that people do not slow down in situations where it does not make sense.
You really think people slow down so they can get a good look at a stalled truck?
Yes. They do. People back off the gas to have a gander at all sorts of trivial things if they represent something of a change to their usual expectations. They'll slow down to look at a cop pulling a speeder over, for example.
It's not as if we're operating purely on guesswork about these things. Often enough you'll hear on the radio what caused the backup in the opposing lanes before you hit the traffic jam in your own lanes--and no, it won't be a mass of flashing blue lights just around a curve on the freeway so that drivers can't tell which side of the road they're on, or a massive accident that drivers justly feared might spill over onto their side of the road or whatever. It will be "an obstruction in lane 1 is causing delays" or some other such utterly mundane event.
posted by yoink at 8:03 AM on March 27, 2012
Yeah, of course. If you actually see an accident happening in the the opposing travel lanes ahead of you you would be a fool not to slow down. Who is arguing otherwise? Just because one can endlessly proliferate examples of situations in which it DOES in fact make sense to slow down does not mean that people do not slow down in situations where it does not make sense.
You really think people slow down so they can get a good look at a stalled truck?
Yes. They do. People back off the gas to have a gander at all sorts of trivial things if they represent something of a change to their usual expectations. They'll slow down to look at a cop pulling a speeder over, for example.
It's not as if we're operating purely on guesswork about these things. Often enough you'll hear on the radio what caused the backup in the opposing lanes before you hit the traffic jam in your own lanes--and no, it won't be a mass of flashing blue lights just around a curve on the freeway so that drivers can't tell which side of the road they're on, or a massive accident that drivers justly feared might spill over onto their side of the road or whatever. It will be "an obstruction in lane 1 is causing delays" or some other such utterly mundane event.
posted by yoink at 8:03 AM on March 27, 2012
Last night I was on the freeway and I decided to test out the theory about breaking up traffic waves by driving slower than the wave. I didn't notice any effect at all. Perhaps it depends on the aggressiveness of drivers, around here they aren't very aggressive at all. But I did discover what I think is the optimal strategy: take advantage of traffic waves. When there is a gap, merge into that lane. This will slow the drivers behind you and break up the wave ahead of you.
posted by charlie don't surf at 10:31 AM on March 27, 2012
posted by charlie don't surf at 10:31 AM on March 27, 2012
« Older Aisle seat, please. | For once, clouds are a good thing Newer »
This thread has been archived and is closed to new comments
posted by Chekhovian at 8:03 PM on March 24, 2012 [1 favorite]