Traffic Sim Predicts Jams Before They Happen

Via_Patrino writes "The New Scientist reports that: A traffic simulation system is helping drivers by predicting jams up to an hour before they happen. Traffic flow can be divided into three categories: freely flowing, jammed, and an intermediate state called synchronised flow in which dense traffic moves in unison. Physicists at University of Duisburg-Essen have developed 'the first model to reproduce all known traffic states.' Predicted conditions are displayed on the official website, and more than 90 per cent of the time, traffic density is predicted correctly."

Comments

What if people start using it?

[Zebidiah]

What if people start using it? They avoid the traffic jams, thus no traffic jams. Wrong prediction!

Re:What if people start using it?

[mphase]

Good point. It only applies if a large portion of the potential traffic was aware of the prediction but still a good point. It's not a matter of being a right or wrong prediciton though, it's a matter or decreasing traffic. If it makes it so people get stuck in traffic less than it is a success. The statistic of being "right" or "wrong" in its prediciton is pointless. Though obviously if every damn person was aware of it the predicted traffic would become the perfect driving conditions because everyone would avoid that area. So then people would start to purposefuly drive in the areas the system said would have heavy traffic, knowing that everyone else would avoid it. But then because so many people did that there would be traffic and so people would learn to avoid the areas...repeat until you come to the topic of a land war in China.

Re:What if people start using it?

[obli]

Some people still need to get to places in time, though...

Re:What if people start using it?

[Zebidiah]

So what you are saying that people would start to second guess each other. They would probably all go for the easiest route and cause a traffic jam. What we need now is for this program to be able to predict the effect this program will have on traffic and....well repeat continously.

Re:What if people start using it?

[obli]

It will be incorporated into the original program by skynet-style intelligence, which will eventually take over the world, but that's a different story.

Re:What if people start using it?

[12357bd]

Well, let the system consider the new traffic state, at the end you have a perfectly distributed traffic and a steady and boring prediction.

Re:What if people start using it?

[NanoGator]

"What if people start using it? They avoid the traffic jams, thus no traffic jams. Wrong prediction!"

Oh the horror.

Re:What if people start using it?

[name773]

...or is it...

Re:What if people start using it?

[Anonymous Coward]

If everyone was using it, you wouldn't need to predict traffic condidions, since you could then use the system to directly control the traffic conditions.

Re:What if people start using it?

[gilroy]

Blockquoth the poster:

Wrong prediction!

To steal a line from Minority Report, "The fact that you prevented it from happening doesnt change the fact that it was *going* to happen." (All hail the IMdb)

Re:What if people start using it?

[bairy]

What about the other side too.. Every sees that Street A will have a traffic jam so they all go alone Street B... which would then have a traffic jam

That would move the problem rather than solve it.

Re:What if people start using it?

[Neduz]

I don't think everybody will avoid the traffic jam, some people will think that if all the others avoid that place, there will be no traffic jam.

Re:What if people start using it?

[AlphaFreak]

Actually, this is happening (people uses it) and the effect is the opposite: more traffic jams. Read the article: people who uses the predictions modifies their planned route... unfortunately this somehow breaks the model and leads to new, unforecasted jams.

Re:What if people start using it?

[brunokummel]

well, I guess they are aware of that.. =D "But the website has already become a victim of its own success, admits Schreckenberg. Some of the 300,000 people a day who are visiting the site are replanning their journeys on the basis of its forecasts, and this is beginning to make the forecasts themselves less accurate. And soon it could get even worse when the website becomes available on 3G cellphones, he says."

Re:What if people start using it?

[jcuervo]

repeat until you come to the topic of a land war in China.

It all ends up in the Middle East.

Re:What if people start using it?

[Anonymous Coward]

you just described the problem of utilizing the intelligence gain from cracking WWII Nazi codes: if heard about a fleet of submarines and attack it, then they know you know the codes, so they change them...but if you do nothing, then what's the point of breaking the codes

Re:What if people start using it?

[Anonymous Coward]

I have a solution for this... the traffic simulator can give different information to different people, thus spreading out the traffic and preventing the jam. Yeah, sure it's lying but they don't need to know that!

Re:What if people start using it?

[Anonymous Coward]

Depends on how they use it. If they use it like the TRANSIMS group at Los Alamos National Labs, city planners would use it to identify areas that would cause traffic jams. They could then implement a change and test it in the simulation to see how it affects the traffic. Seattle and Houston had their traffic modeled by this system a couple of years ago.

TRANSIMS is an agent based model with each human accounted for in the system based on census info. So it is complex enough to handle any traffic pattern. It runs on LANL's ASCI Blue SGI cluster.

The TRANSIMS project has been around since at least 2000 so this latest project isn't the first to analyze traffic.

Re:What if people start using it?

[wideBlueSkies]

>>To steal a line from Minority Report, "The fact that you prevented it from happening doesnt change the fact that it was *going* to happen

What I take away from that statement is that even though we learn how to avoid, or run from a problem. The problem is still there, just waiting to be solved.

wbs.

Re:What if people start using it?

[serutan]

It only applies if a large portion of the potential traffic was aware of the prediction

Not necessarily. Depending on the situation, the threshold number of people whose behavior can change the situation will vary; could be large or small.

Where I live (Seattle) the state DOT puts up a GREAT online traffic map, which I check religiously before hitting the road. I hope it someday incorporates technology like this.

Re:What if people start using it?

[Hasdi Hashim]

sounds like a stock analyst's prediction.

"The price will go up", so people buy and the price goes up. "The price will do down", so people sell and the price goes down.

Re:What if people start using it?

[Anonymous Coward]

...this will be like weather forecasting:

accurate but occasionally off track

Re:What if people start using it?

[zonker]

until schroedinger's cat runs into the road and gets hit by a car...

Re:What if people start using it?

[Zebidiah]

It's just second guessing. Which way do you jump. As long as every body guesses it may work. If every body reads a 'forecoast' they everyone may jump the same way; gridlock.

Re:What if people start using it?

[Zebidiah]

Actually, this is happening (people uses it) and the effect is the opposite: more traffic jams. Read the article: people who uses the predictions modifies their planned route... unfortunately this somehow breaks the model and leads to new, unforecasted jams.

It needs to take its own preditions as input!

Re:What if people start using it?

[Anonymous Coward]

True story!

I was putting in an application for a job on the day it actually had to be in (before noon). I delivered by hand very early in the morning (before 7am). As I was driving down the coast road (dual carraigeway North East UK) to deliver it I noticed a dead cat on my left hand side, thought shame, but what the hell. After I handed in the application to the caretaker, I turned my scooter round and drove to work. On the way I passed a dead cat on my left hand side again (on the same stretch of road), thought 'bloody hell, deja vu', looked across to my right hand side and saw the other dead cat that I had passed originally, more or less directly opposite!

Completely offtopic but have anyother people in the NE UK noticed that roadkill is way higher than previous years. I'm drunk. Going to bed. Night all.

Re:What if people start using it?

[cagle_.25]

"The fact that you prevented it from happening doesnt change the fact that it was *going* to happen."

So says John Anderton, defending the use of pre-crime techniques.

Of course, the entire point of the movie (culminating in Anderton's boss's death, contrary to the pre-vision of the three precogs) is that there isn't "absolute metaphysics" as Anderton claims. At the end of the movie, pre-crime is dismantled because it turns out to be fallible and not absolute after all.

Of course, you probably knew all that and were just reveling in the irony and paradoxes of "pre-traffic". :-)

Re:What if people start using it?

[jcuervo]

*sigh* It's a joke.

It goes like this. "You cut someone off in traffic, and they get all pissed off, so he cuts someone off, who gets off the freeway and starts speeding, until he gets pulled over by a fat cop, so he makes a fat joke, and the cop gets all pissed off, until it all ends up in the middle east!"

I forget who told it.

What is the cause?

[lachlan76]

But the system still wouldn't be able to predict traffic jams caused by accidents and cars breaking down, so it wouldn't be perfect. And if everyone reads when there will be a jam, no-one will drive then, and the traffic jam will be at a different time.

Re:What is the cause?

[skraps]

And if everyone reads when there will be a jam, no-one will drive then, and the traffic jam will be at a different time.

Presumably, this is a closed-loop system, with a short loop. The predictions are based on very recent, near real-time data. If the traffic is affected by people reading the predictions, then the predictions would adjust to compensate very quickly.

Re:What is the cause?

[Wastl]

Accidents can (probably) be predicted with probabilities. Assuming that you have "synchronized traffic" of a certain length, it is (probably) very likely that an accident happens.

Besides, as far as I understand, the system currently only considers highways, which (at least in Germany) have at least dual carriage ways. If the traffic is not dense, most accidents or break downs do not cause jams, as the traffic simply passes by.

Sebastian

Re:What is the cause?

[Trent05]

Good call.

I'm sure only a small pecentage of people would be checking the website. Heck, in my town they publish where the sobriety checkpoints are going to be set up for the weekends and holidays, and the cops STILL have a bumper crop of DUI/OMVI arrests.

Oh well, I guess it pays to pay attention...and drink at home.

Re:What is the cause?

[s.fontinalis]

On American Motorways the probability of a vehicle breaking down in the traffic lanes is directly proportional to the nearness of a major holiday, as is the ability of the driver to shift the vehicle from the travel lanes.

Re:What is the cause?

[The FooMiester]

Just like the old "gridlock alert" days in NYC. They stopped doing them back in the 80's, because of that very effect. They announced a gridlock alert day for one of the games in the '86 world series, and my uncle said the traffic was so nonexistant that he didn't do under 55 until he got to Shea stadium.

Re:What is the cause?

[Bozdune]

Right.

Note that this system is infinitely better than the radio "traffic reports," like those in Boston on 1030 WBZ. The announcers already know from experience where all the slowdowns are likely to be, so they just repeat the same B.S. every morning, true or false, until they get lucky and their helicopter spots something, or the State Police radio in an accident.

I used to commute the Mass Pike eastbound from 495 into 128/95, and I can count on the fingers of one hand the times that the "traffic report" was actually accurate.

As far as people paying attention to the web site and changing their route based on 30 minute or 60 minute advanced prediction, thus screwing up the "prediction," that's only an issue if you need advance warning before you leave the house. What's really needed is a real-time decision when you come up against a junction point. Which way? Tell me now. Should be easy enough to do with a real-time feed from your car to the model.

Not that there's that many obvious alternate route choices around Boston anyway, but it sure would be helpful to have precise information on the jam-ups for those of us who know the back roads.

Re:What is the cause?

[Lars T.]

Quote the article: "But the website has already become a victim of its own success, admits Schreckenberg. Some of the 300,000 people a day who are visiting the site are replanning their journeys on the basis of its forecasts, and this is beginning to make the forecasts themselves less accurate. And soon it could get even worse when the website becomes available on 3G cellphones, he says."

Re:What is the cause?

[jrl87]

So to get this real time information, some idiot is going to put another screen or something in the dash of cars to display it and people will be watching it for traffic problems ahead of them instead of watching the road, causing even more traffic problems.

There really is no good way predict and then relay the information.

I am not sure what it is like up there in Boston, but in Texas the radio announcers give the same BS, but I am familiar enough with traffic flows to know what to expect. For example, parts of 820 are always bad because there aren't enough lanes, I30, especially going west, is really bad from about 4 to 6 mostly due to Lockheed employs getting off, but other places as well. Going into Dallas on 30 is almost suicide no matter what time of day it is because there is construction and they decided to make the road ressemble a slalom course and there are several locations along the way that have wrecks over 60% of the time when I go by .....

Not quite as accurate as this system, but it is effective and it doesn't add any danger other than the fact that I am out on the road with all the other idiots ..

Re:What is the cause?

[Anonymous Coward]

ah, so if I just use the old reverse psychology, everyone will try to avoid the jam and I'll just cruise on by :)

Re:What is the cause?

[iantri]

Wow.. traffic reports must be infitely better here, then..

In Toronto, you can watch the major local TV morning show, BreakfastTelevision on CityTV, and hear a traffic report based on, and with live video, since the Ministry of Transport has cameras set up all along the 400-series highways. (Tinfoil hat folks: The cameras are shit quality, frequently fogged up or out of focus, and way to far away to read plates..)

Re:What is the cause?

[aggiefalcon01]

I'd like to see this kind of (near) real-time information appear in my own car when I want it, so I can make an instant decision, as opposed to deciding to avoid a certain area in 15 minutes when I'd be getting there. Take the onboard DVD/nav in modern cars, and get it to retrieve info from this system in realtime, and I'll be happy.

Houston appears to be the same story. The traffic reports on the radio appear to be based on hour-old information, when they do reflect reality (and that itself is rather rare).

Pre-emptive analysis?

[skraps]

Do city planners have access to this sort of technology? It stands to reason that if the people designing the roads had a better understanding of traffic dynamics, they could design better road systems, and increase capacity proactively.

Re:Pre-emptive analysis?

[mocm]

There were and are several projects in Germany that develope traffic models in conjunction with models simulating the route choices of people depending on their lifestyles. The problem is that there are people that do not believe that people can be simulated by cellular automata and reject those simulation. But there are also city planers who use those models when planning new structures like stadiums, airports and others that need road connections.

There is a sourceforge project thatsimulates multimodal traffic with cellular automata. Also have a look at this link for more information about traffic simulation.

Re:Pre-emptive analysis?

[ratsnapple tea]

Yeah, it's pretty common practice nowadays to use computer models to assist in urban planning, not just for designing new roads, but also to see where best to widen lanes, where to add additional lanes, and even how to time traffic lights. One problem with prediction, over the long term, is that it turns out no matter how much road you build, it isn't enough. So building in 5% additional capacity staves off traffic problems in the short term, but within a decade or two you've got 10% more drivers, the additional traffic coming from the fact that your original (planned) increase in capacity made it that much easier to drive. Of course, this is why governments are more and more turning to congestion tolls as a way to thwart this cycle. More here.

Re:Pre-emptive analysis?

[supretigre]

City planners don't need access to this sort of technology in places with traffic jams. I am one... in a city with really bad traffic. Like a few astute individuals have pointed out below - the traffic jams happen at the same times everyday, on the same routes! I saw a nationally-recognized traffic engineer speak once - he was a real renaissance engineer, for his age. In his presentation he showed us data that suggested that regardless of the capacity of a road, the dynamics of vehicles and drivers in the US results in a average maximum throughput of about 35 mles per hour of any typical arterial or highway. Basically, we get to choose how many cars we want stuck in traffic by choosing the number of lanes (and the capacity at the ending points) in our arterials and freeways. Adding lanes is only a very temporarily successful measure. Another interesting point he made (I've GOT to remember this guy's name... senior moment) is that at 30 mph or less, it is much more effficient to distribute traffic over a grid system, even at lower speeds, than it is to try to puch them all through a big pipe (freeway). Sounds familiar, eh?

Re:Pre-emptive analysis?

[Anonymous Coward]

what the f*ck makes you think that is redundant?
do you not understand how to order the comments chronologically?
please, please.. point out the previous post that has even remotely the same comment.
there should be some form of iq test before slashdot gives out mod points to complete shit-for-brains idiots.

Hooks law

[Anonymous Coward]

I remember a few years ago being told by a teacher that traffic flowing under 28mph on Motorways obeys hooks laws for compression waves travelling through a spring almost perfectly...

The compression waves travelling through the traffic are the reason that everything goes stop/start once traffic slows below a certain speed...

Re:Hooks law

[value_added]

"The compression waves travelling through the traffic are the reason that everything goes stop/start once traffic slows below a certain speed..."

Interesting. I think I'll make it into a bumper sticker.

Check back in a few days and I'll let you know how effective it is calming down the idiot honking his horn behing me. ;-)

Re:Hooks law

[synesis]

This stuff is old. I attended a freshman course in differential calculus in 1972 at King's College in London given by Professor Kilmister. He developed a set of non-linear equations describing traffic flow. The equations showed the build up of jams and demonstated wave-like behaviour.

Re:Hooks law

[Anonymous Coward]

It is this sort of behaviour that makes roundabouts a better idea than 4 way stops: bunching induced by the junction in slower traffic is much less severe. On the other hand I have no clue why on quieter roads in the UK there is an obsession with chicanes which cause bunching (and also force cars to go head on).

There seems to be gulf between theory and practice.

The dumbest junction I saw was a 4 way stop at the end of an off ramp. When the road at the stop got clogged traffic would then rapidly back onto the freeway and stop that. With a traffic light controlled junction there would be at least some chance of filtering traffic from the freeway, although in this case by causing 'inverse bunching'.

A note

[pyth]

For those people concerned that releasing the predictions will interfere with the results, please remember that not many people will actually use this tool.

Re:A note

[Zebidiah]

I agree with what you say, but as I said above. How accurate will the predicitions be if it were commononly used?

Can it predict its own interference/effect it will have on traffic jams?

Re:A note

[Zen Punk]

Also note that most people don't have web access whilst driving down the freeway, and thus won't be able to get the updated predictions anyways.

Wow....

[supertsaar]

It must be an ultra-sophisticated system to be able to predict a traffic Jam on the Amsterdam A10 direction Coentunnel at 17:00 every weekday :P

Re:Wow....

[Anonymous Coward]

no shit, I can make a program like that too:

10 Print "Houston 59 between downtown and 610 loop during rush hours."

This is one of my pet peaves

[rsilvergun]

I hate seeing people speeding to red lights, when if they'd just follow the flow of traffic and the lights they'd get where they're going just as fast, and without causing traffic jams. I'm a terrible driver, but I figured this simple thing out pretty early on. Why can't the rest of drivers?

Re:This is one of my pet peaves

[Anonymous Coward]

Because a single driver's illusion of getting there 30 seconds faster is far more important than keeping traffic flowing smoothly for everyone (including other similarly-deluded drivers).

Re:This is one of my pet peaves

[willpall]

Or a corollary to that:

I drive a stickshift and consequently cannot stand stop n' go traffic on a freeway. So I even it out and am able to maintain a consistent speed regardles of the sporadic flow ahead of me. This of course pisses off the person behind me because I have "too much space" in front of me. They usually don't realize that I'm doing them, their car, and their gas mileage a favor. I wish people would just think about these things. The easiest lane to be in in this type of traffic is usually the lane with all the trucks, as they too cannot stand the constant shifting and braking of stop n' go traffic.

Re:This is one of my pet peaves

[12357bd]

How true!
In my experience, you have to drive a slow car to learn what 'protective drive' is. People with 'normal' (fast) cars never realize how easily the flow is changed. Drive a slow car for a decent period of time (one year) should be part of the trainning of any driver.

Re:This is one of my pet peaves

[transient]

I do this in my stick-shift too, though not on the freeway (I live in a small town). There's a leg on my drive home that's always backed up at a stop sign -- stop n' go like nothing else. If I shift into first and let the engine idle, I catch up to the car in front of me right as it starts accelerating. It makes people behind me crazy and I can't understand why.

I guess it's just more evidence of the superiority of stick-shift drivers. ;-)

Re:This is one of my pet peaves

[arcade]

Oh how I hate people with that attitude. :-)

I'm not a driver (well, I've got a driving licence, but no car). I'm a cyclist. There is nothing more annoying than a slow moving car snailing along at 20mph when I'm in a training session - treading along at 30mph. It invariably causes me to be pissed of and just pass the slow driving sucker.

Who far to often doesn't like being passed by a fast moving cyclist.

Yeah, I know, I'm a dangerous asshole cyclist in the traffic, but eyh, it's my health - and I've never had a traffic accident yet. :-D

Re:This is one of my pet peaves

[thogard]

30 seconds at the start of a long trip can add up. Remember that stoplights compound at more than 50%. The result is that a few seconds through a light might mean hitting the next one in time and that saves two minutes. In an area with bad traffic, a 2mph drop in speed at 35mph can increase your trip time far more than the the 6% of the speed decrease would predict.

Re:This is one of my pet peaves

[garcia]

See, here in the fine State of Minnesota you have to be right behind the person in front of you at all times. You cannot even leave 1mm of space. If you do, some cheating asshole will try to move up the free flowing lane and cut over at the last possible moment to avoid the 35 minutes of backed up single lane traffic that you had to endure.

Now, this may not seem that bad but... This cheating asshole fucks up the speed of those in the free flowing lane, causes people to mash on their breaks in the traffic lane, and pisses me off that a) this asshole cheated and b) that some cell phone talking dumbass (I don't know who these people talk to at 7am but it better be damn good) who has been waiting 35+ minutes to get to this point has let the cheater in!

Re:This is one of my pet peaves

[Lars T.]

Well, cyclist don't care about traffic lights anyways, so they are exempt from that traffic flow rule.

Re:This is one of my pet peaves

[Lars T.]

No way. In Germany almost everybody drives stick, and they still don't get this flow thang.

Re:This is one of my pet peaves

[MurphyZero]

I do this many times with an automatic. Unfortunately the people in the next lane like to ruin the space by moving into it. Slowing me down as I increase the space again. But then I live in Florida which has three types of drivers: tourists, who don't know their way around (think sudden turns and multiple lane changes to reach exits), Snowbirds (think retirees, mostly during winter months, slow drivers) and natives (At least 15 mph over the speed limit, and switching lanes constantly to get around snowbirds) In general, I hate Florida driving.

Re:This is one of my pet peaves

[pipingguy]

the superiority of stick-shift drivers

Clutch much?

Most of the key in driving is keeping the car in a state of balance.

http://tisc.planet-f1.com/drivers/results_by_dri ver.phtml/tis/ENG?driver=STEWAR

Re:This is one of my pet peaves

[TheLink]

You're smoothening out the wave, but it does delay traffic flow. Your proposed method is fine for very long roads with few intersections, but not so good in places with lots of intersections, because it means fewer cars will make it through the intersection - which means slower average traffic speed and people taking longer to reach their destinations. For roads with lots of traffic lights, it's better if people accelerate at reasonable pace in lock-step but they must not overfill the road and block the intersection. By accelerating at a reasonable pace - I mean taking into account safety considerations - the mildly stupid/careless pedestrians. Nevermind about the darwin award nominees. In some intersections here, they have started putting countdown displays indicating the number of seconds the traffic lights have before they change - e.g. from red to green, from green to yellow, from yellow to red. This is very useful - it is easier for people to not overfill a road, and at the same time not _underfill_ a road.

I doubt there is a single correct solution. Even if there is one, not everyone drives the same, nor can they, nor can they be expected to. After all it is a public road for diverse members of the public, not an race track with professional racers.

Fuel consumption isn't a big issue. Most people with cars earn significantly more than their fuel bill - otherwise they shouldn't have cars. I also forsee that trying to maintain consistent speeds, smooth acceleration/deceleration to save fuel would become less important when cars with regenerative braking become more common. Right now, most vehicles brake by converting kinetic energy to waste heat. To me this is one of the main reasons why an electric/hybrid car is more efficient than conventional cars in practice - it's all because of regenerative braking. For the same amount of money, they're all about as efficient in converting fuel to motion, it's just what happens when you slow down. If you had a way to add decent regenerative braking to a conventional car (even better - bus), you'd improve efficiency significantly.

I'm not sure if gasoline fuel cell cars would be more efficient.

Re:This is one of my pet peaves

[molo]

That only works when traffic lights are synchronized so that they turn green when you approach at the speed limit. There are far far more intersections where the lights are independent, and you have a reasonable chance of squeeking by on a green or yellow if you speed.

-molo

Re:This is one of my pet peaves

[Anonymous Coward]

There is one more point to consider: cities like Austin where the stop lights are not timed for smooth flow. In these cities, hitting the gas a little more aggressively will get you through the next light where the others are stuck at the now-red light.

If only the traffic lights were timed, it would make life a lot more simple and reward those who flow nicely.

Re:This is one of my pet peaves

[ZiggyM]

Culture also plays a role here... I did the exact thing back when I lived in Seattle. Luckily, people there are polite and at most they would flash their headlights telling me to move my car. I live now in Lima Peru, and such strategies are impossible here. As soon as you leave a couple meters [~10 feet] in from of your car, another car will cut you off and steal your spot.

Re:This is one of my pet peaves

[Magius_AR]

I hate seeing people speeding to red lights, when if they'd just follow the flow of traffic and the lights they'd get where they're going just as fast, and without causing traffic jams. I'm a terrible driver, but I figured this simple thing out pretty early on. Why can't the rest of drivers?

I don't see how speeding towards a red light would cause a traffic jam. Now if you're referring to "speeding towards a green/soon-yellow light to beat the red", that's another story, and YES there is a time difference. I can't count the number of times I've beaten the light and left the poor sap with nowhere to go and nothing to do sitting back at the light. In a town where these damned reds last upwards of 1-2 minutes, the cumulative amount of time saved on subsequent lights/traffic flows/etc becomes quite substantial.

All that is irrelevant anyways. My time is precious just as yours is. Who are you to claim what a minute of my time is worth? It's simple courtesy...just get the hell out of the way of faster drivers. Why is that so hard?

I drive a stickshift and consequently cannot stand stop n' go traffic on a freeway. So I even it out and am able to maintain a consistent speed regardles of the sporadic flow ahead of me. This of course pisses off the person behind me because I have "too much space" in front of me. They usually don't realize that I'm doing them, their car, and their gas mileage a favor. I wish people would just think about these things. The easiest lane to be in in this type of traffic is usually the lane with all the trucks, as they too cannot stand the constant shifting and braking of stop n' go traffic.

Do it in a passing lane and I'll certainly care. Leaving a gap of open space in any tight traffic is just _begging_ people to cut into your lane. I don't know how you haven't seen it. Though you may think this is not "slowing" anyone down, remember that for every car that fills the space you left, you need to slow down to leave the same gap. What this leads to is your lane of traffic losing alot of speed/progress (rightfully pissing off those behind you) simply because you want your own comfortable drive. Hey man, you didn't have to buy a stick. And if you want to do that, at _least_ have the common courtesy not to do it in a passing lane.

Re:This is one of my pet peaves

[rsilvergun]

Speeding to a red light causes trafic jams because it takes time to get cars moving again. When it's one or two, the time is negligible. But Once you start gettting 10 or so (common at popular intersections in my town), then the 11th guy doesn't make it through. Now the next light still only lets 9 through, but there's one more stopped. So at the start of the next light cycle there's two people already waiting at the light. Each light cycle adds one person to the line (more probably, this is just a really simple example, I'm not smart enough to actually model this stuff out, if I was, I'd be designing roads for a living :). The end result is a hugh line of cars from 5pm - 6pm or so. If people followed the flow of traffic, everyone would just cruise through lights and this wouldn't happen.

I can't remember the guy's name, but the guy who writes (or used to write, I haven't seen the books in years) sun's core java series had a program that modeled this on his web site.

What happens next...

[lewko]

But the website has already become a victim of its own success, admits Schreckenberg. Some of the 300,000 people a day who are visiting the site are replanning their journeys on the basis of its forecasts, and this is beginning to make the forecasts themselves less accurate. And soon it could get even worse when the website becomes available on 3G cellphones.

...and the system fails to take into account for drivers futzing with their cellphone instead of concentrating on the rear of the car in front of them..... Crunch!

Re:What happens next...

[HungSquirrel]

As someone said in an above thread, fender-benders aren't taken into consideration. It's a denisty model, not a human error model. :(

Re:What happens next...

[Zebidiah]

As someone said in an above thread, fender-benders aren't taken into consideration. It's a denisty model, not a human error model.

Or perhaps even 'It's a density model not a human density model :-). Shame, if we could only model human stupidity.

GPL IT PLZ

[HungSquirrel]

That is all. :)

Re:GPL IT PLZ

[mocm]

ok

Re:GPL IT PLZ

[boaworm]

Here is the source code:

#!/usr/bin/perl -w

$now = `date +%H`;

if($now == 7 || $now == 8 || $now == 16 || $now = 17){
## Genius part!!
die "Rush hour, take another route!\n";
}else{
die "Roads are clear!\n";
}

## Please note, code is not optimized yet, but it's 90% accurate!

Re:GPL IT PLZ

[HungSquirrel]

No shit. :D

Re:GPL IT PLZ

[Anonymous Coward]

I don't think Maxis will GPL thier software anytime soon. :)

well here in minneapolis

[JeanBaptiste]

you can count on going about 15mph anywhere during peak hours. most highways are only 2 lane, and they tend to go to one lane through anything important, like crosstown (hwy 62) merging onto 35w. Predicting jams before they happen? I can do that.

doh

[pyth]

I should have read the New Scientist article. Apparently is IS being used by many people, already!

Measuring Theory Of Traffic

[uncreativ]

An idea borrowed from quantum mechanics...
Measuring the system changes the system.

What if everyone takes the same alternate route to avoid the "busy" route?

As the story points out:

"But the website has already become a victim of its own success, admits Schreckenberg. Some of the 300,000 people a day who are visiting the site are replanning their journeys on the basis of its forecasts, and this is beginning to make the forecasts themselves less accurate."

Re:Measuring Theory Of Traffic

[skraps]

From their web site, it appears that the system was originally designed as an open-loop system, which explains their concerns about feedback. Taking an open-loop system and mis-using it as a closed-loop system often results in rapid oscilations due to the prediction model over-compensating for the feedback.

Assuming they have a good enough understanding of the way traffic works, they could change to a closed-loop prediction model that would converge on a correct answer even in the face of feedback.

Re:Measuring Theory Of Traffic

[gilroy]

Blockquoth the poster:

What if everyone takes the same alternate route to avoid the "busy" route?

Yep, it's psychohistory all over again... *shaking fist* Curse you, Isaac Asimov!

Re:Measuring Theory Of Traffic

[Anonymous Coward]

You sir

areafuckingmoron_____==+++=WWWWWWWWWWWWWWWWWWWWW__ -WWWWWWWWWWWWWWWWWWPReason:Pleaseusefewer'junk'cha racters.Reason:Pleaseusefewer 'junk' characters.Reason: Please use fewer 'junk' characters.Reason: Please use fewer 'junk' characters.LEASEDFSDF__
Disevidence
Preferences
Subscribe
Journal
Logout

Sections
Main
Apache
1 more
Apple
4 more
Askslashdot
Books
BSD
Developers

Re:Measuring Theory Of Traffic

[Random832]

Disevidence
Preferences
Subscribe

Seems like there's been a rash of these "accidental identity disclosures" lately... wonder if they might be joe jobs?

Re:Measuring Theory Of Traffic

[betelgeuse-4]

Chaos Theory is more appriate here than quantum mechanics. Just measuring the flow of traffic doesn't change the flow of traffic. However, making predictions based on your measurements can create a feedback loop which can affet the system.

Isn't this a statistical problem?

[anubi]

It seems to me its nearly as impossible to predict a traffic jam as it is to predict stock prices.

Both are fundamentally chaotic.

Sure, you can calculate expected probability based on past performances and expected flow... but we've all seen freeways humming right along at 70MPH and no problems until just ONE driver makes an error... then all hell breaks loose.

I don't think even predicting the weather is as tricky as predicting traffic flow, as at least the weather patterns follow known laws of physics for at least near-term before losing out to the chaotic nature of weather patterns. People are just flat out unpredictable.

Re:Isn't this a statistical problem?

[ratsnapple tea]

While you might not be able to predict which driver makes an error, you can say with near certainty that somebody is going to munge up the traffic flow. That's what you base your predictions on. Remember that chaos theory indicates that although the underlying systems may be chaotic, the emerging patterns can be predicted with a fair degree of accuracy.

Re:Isn't this a statistical problem?

[anubi]

Very true.

Some very interesting things come out when one starts using statistical analysis of chaotic phenomena. The patterns, as you say, are pretty consistent.

I guess its kinda like predicting what the noise level next to the freeway at any particular hour will be - and be pretty darned accurate. But try to identify exactly where you will be in the measured acoustical waveform and you will be completely out of luck.

Chaos gets interesting, doesn't it?

Re:Isn't this a statistical problem?

[12357bd]

Both are fundamentally chaotic

That's mixing levels...
Yes at the lowest level both are inherently similar (human dependent), but at higher levels they are very different, ie, cars have an almost fixed size (related to the highway), stock orders can be any size, that's a strong difference. Also cars are phisical objects, so they move in a predictable (relatively simple) way, fluid mechanics and compression waves explains a big part of traffic behaviour.
It seems that the authors just make the model more realistic, not specially different than other computer simulations.
Weather is more an example of a complex system, than a 'chaotic' one, you just need a lot of information to produce a sensible prediction, but it's not a coin flipping problem (brown movement), again, is all about observation/modelation levels.

Re:Isn't this a statistical problem?

[ArsenneLupin]

Yes at the lowest level both are inherently similar (human dependent),

Yes, and this human dependancy is what makes them unpredictable. As soon as a reliable prediction algorithm is found, the actors (who are aware of the prediction) act on it, falsifying that same prediction.

In the case of traffic, drivers tend to move to an alternative route, which is shown to be less congested, making it more congested in the process. In the case of the stock exchange, people will buy the stocks which are predicted to rise in advance, which makes them rise or fall earlier than predicted. In both cases (traffic and stock market), actors seeking their personal gain have an active interest of "messing up" the prediction. Predicting human-dependant events is inherently difficult, unless the event is of the kind that gives a (perceived) advantage to "going with the flow", such as fashion or US presidential elections. In that case, its easy to predict, because the prediction even helps the event come true (self-fulfilling prophecy)

cars have an almost fixed size (related to the highway)

Can you spell T R U C K?

Weather is more an example of a complex system, than a 'chaotic' one, you just need a lot of information to produce a sensible prediction, but it's not a coin flipping problem (brown movement),

Chaotic is not the same as random. Even a perfectly deterministic system may be chaotic, and thus unpredictable unless initial conditions are perfectly known. In a chaotic system, tiny differences in initial conditions lead to huge differences in final conditions (the "butterfly effect"). As the initial conditions are never perfectly known (measurement errors, finite density of measurement point mesh) a long range prediction is impossible, even if the system obeys deterministic laws.

Re:Isn't this a statistical problem?

[anubi]

Yes. Cars have a fixed size, but just as stock orders can be any size, the effect of a driver error can range anywhere between a minor pullover to the side of the road, or taking the whole freeway down.

And yes, it sure seems weather should be more of an example of a complex system, but I still consider that the nonlinearities involved force it to chaotic behaviour. We can get close... and the shorter the time frame, the more accurate we can get. As we increase the resolution of the complex system of weather models, we can predict further into the future. I hold that the chaotic nature driven by these un-modelable nonlinearities which extend beyond the molecular level will mean we will never develop accurate long-term forecasts, although our statistical predictions should hold out fine.

Its just I don't think I will be able to predict with much degree of certainity how much rain will fall on Atlanta, Georgia on May 12 of next year, but I think I can pretty accurately tell you how much rain to expect in the State of Georgia for the week or month a year from now.

Statistics seems to me a weird mistress. She teases me a lot.

Re:Isn't this a statistical problem?

[12357bd]

Can you spell T R U C K?

I can try... :)
Yes, trucks are bigger than cars, but is not that significative, both are a minuscule part of a highway, and both run on a single track, so dinamics are similar. On the other hand, stock orders easily vary 2 or 3 orders of magnitude.

About chaotic not being random, true, absolutely true, but the idea was that wheather can be reasonably predicted if you consider enough information for the desired detail level, it only becomes unmanegeable once you try to predict particular events (ie, tornado's path) or too future conditions, that's where the question becomes 'chaotic' or 'random' or simply 'too complex' if you prefer.

Re:Isn't this a statistical problem?

[12357bd]

I hold that the chaotic nature driven by these un-modelable nonlinearities which extend beyond the molecular level will mean we will never develop accurate long-term forecasts, although our statistical predictions should hold out fine.

yes, stats are clearly limited as a modeling tool, I prefer the 'let us simulate realiity' or modeling approach (ie: cellular automata), it gives a better understanding of the problem at hand. Stats are just as good as data set are, no more, but models reflect our own (lack of?) knowledge of the modeled thing.

Re:Isn't this a statistical problem?

[anubi]

Interesting link you have!

Oh yes. Models.

When I worked in the Aerospace industry a few years ago, I had lots of models for the devices and fabrication methods so I could try to predict their electronic behaviour using the SPICE simulator.

The SPICE simulator itself is mostly set up to simultaneously solve lots of linear differential equations with connectivity represented by placement in the matrix. A big linear algebra array of differential equations. We could see how the math worked. But trying to get an accurate - I mean a really accurate - model of the real world??? It seems no matter how fine I tuned something, after comparing what I was calculated to get against what I got, there was always some interraction that went unaccounted for and screwed up the entire calculation. Many of the designs had chaotic elements that had large nonlinear outcomes to small control inputs, which meant even a small unaccounted for coupling would royally screw up the works.

I worked for years on some of that until a new management team came in and mandated we all go to closed-source simulators. At that point, I lost damn near all understanding of what I was doing, as sans understanding how the machine handled my simulations, I could no longer get anything to simulate properly. But management liked the pretty reports, even if I had no idea how accurate they were. I ended up losing my job over my bickering over the loss of my machine.

But then, I could see their point of view too... when one is getting paid what those guys got paid, they had better things to do than listen to whining engineers.

Re:Isn't this a statistical problem?

[12357bd]

Unless things change and change soon we will not live the arrival of another astronaut to another planet, it's a shame. Space and not war should be the human goal, after all, space still is the last frontier.

Re:Isn't this a statistical problem?

[anubi]

My sentiments exactly.

I get the idea that you, like me, were active during the fun years of the 60's and 70's before Dan Goldin and his suite of tie-guys ran the tech force into the ground.

Personally, I think that guy set the technology back in this country by at least one half a century, if you consider the loss of all the integrated knowledge we acquired by doing what we did when his suite of tie-guys did all their "right-sizing" by laying off droves of their own scientists and engineers.

I consider that man the equivalent of the adolescent pizza truck driver which tells management he can deliver the pizza five minutes sooner, but wrecks the truck in the process... while the more experienced people would sit back and analyze the economics of marginal benefit over marginal cost and come to the decision that it wasn't worth wrecking a $25,000 truck over a $10 pizza. But marketing and engineering, who base the concept of "value" from different perspectives, seldom - if ever - arrive at the same conclusion as to how to proceed. Whether or not we will succeed follows the laws of physics, which the engineer follows, but whether or not we will be allowed to even try follows the laws of marketing. Unfortunately for us, the engineers were placed below executive management types in the corporate hierarchy, rendering our engineering skills rather moot. And for many years - although funding was not a problem, it was damn near impossible to get any US payload into space.

Enjoy Cassini. It's the last of the old-school technology we all loved.

The latest stellar performances of the Mars Exploration Rovers though gives me hope that maybe our technology hasn't been completely flushed down the toilet.

Re:Isn't this a statistical problem?

[abreauj]

It seems to me its nearly as impossible to predict a traffic jam as it is to predict stock prices. Both are fundamentally chaotic.

Predicting chaotic systems may be impossible, but often the introduction of a periodic perturbation can reduce the chaos significantly, leading to better predictability.

so what your telling me is

[Anonymous Coward]

So what your telling me is, the traffic jam sim is a jedi? And he can see jams before they happen.

And I guess he has an evil father right? "Luke, I can feel the road rage in you, you must let is out"

And I guess news stories about traffic jams have to use lucas/powerpoint diagonal slides instead of hard cuts?

And Han Solo is the guy in the riced up car that speeds? Chewie likes "phat bass beats" in his WQRX p27983tyXX extreme ear-bleed SUB WOOFER SYSTEM.

And I guess there will be massive traffic jam merchandising? Possibly even a prequel traffic jam series with JAR JAR BINKS who is the pedestrian who always presses the walk button even when he has no intention of crossing the street?

And all the people who were caught in the traffic jams ages ago are like "this traffic jam sucks compared to the originals".

Anyway, enough of this, may the green lights be with you.

Predicting the traffic is easy doing something

[Bob Bitchen]

about it is the hard part. Changing behaviors is the most difficult thing to do. If people staggered the time they leave work and if they could tolerate a little bit of inconvenience of car pooling or riding a bus or train or subway then we might see the traffic ease up. Building more freeways and adding more lanes to existing freeways doesn't provide a lasting solution. Most people would reap the benefits of walking more to get mass-transportation and getting out of their cars more often. Of course there are the lucky ones that don't even need a car and can commute soley on public transit. Or just on foot.
I predict that traffic will get so bad and car ownership will become so expensive that people will figure out ways to get around without a car and possibly even change their lifestyle in the process. But it will happen gradually, I think it is already.

Re:Predicting the traffic is easy doing something

[JaredOfEuropa]

Changing behaviors is the most difficult thing to do. If people staggered the time they leave work and if they could tolerate a little bit of inconvenience of car pooling or riding a bus or train or subway then we might see the traffic ease up.

It's quite easy, actually. Give people an accurate prediction of a traffic jam on their route, and they will detour or postpone the journey. And most people already have an accurate prediction... that nasty jam on the way to work has been there every day for the last 10 years, so why wouldn't it be there today?

Over here, people are already staggering their travel times... the jams used to occur only around 8am, now it's packed solid from 7am to 10am. Transport companies are wising up and scheduling runs in the early afternoons... but even around noon it's not unusual to get stuck in traffic these days. Of course you could go by train... except that they are jam-packed during rush hour (it's more like rush morning anyway).

The best change in lifestyle would be to telecommute, or move closer to work... but it also means that there should be suitable housing near your office. In the 60s and 70s, our government planned exactly the opposite: work in the cities, live in the suburbs (or in 'sleeper towns').

Re:Predicting the traffic is easy doing something

[rsilvergun]

Car pooling and busing are not "a little bit of inconvenience" in most parts of America. Public transportation is all but non existant, and employers have come to expect employees to own cars and as such can locate their businesses pretty much where they damn well please. It doesn't help matters that in any community car retailers are bound to be among the most wealthy and influential; and they're not gonna take kindly to public transport. Seriously, what sane person would want the hassle, expense, and extra work hours at their crappy job to own a car if it wasn't a) necessary and b) ground into their skulls since birth by marketing Gods.

Re:Predicting the traffic is easy doing something

[MtViewGuy]

The best change in lifestyle would be to telecommute, or move closer to work... but it also means that there should be suitable housing near your office.

Alas, there are only two cities in the USA that could fit that mold: New York City and Chicago. Most other cities in the USA are so "spread out" that mass transit is not as viable as people think.

Interestingly enough, London in the UK grew by following the Underground (subway) lines as they spread out from the center of London. But then, that growth happened before the automobile became widely available to everyone.

Re:Predicting the traffic is easy doing something

[TheLink]

"If people staggered the time they leave work"

Hey if bosses are flexible about the time you arrive to work that'll help to. Then we'll only have probs if _everyone_ held a 9am meeting on the same day.

It's not the inconvenience of car pooling etc that stops people from driving their cars. It's more like the "insurance" problem - you drive your own car "just in case" stuff happens and you need independent transport you can rely on. Taxis can help, but where I live they're not that reliable - they often refuse to pick you if you're going somewhere that's jammed, or there just aren't enough taxis around... Maybe we could do what Singapore does - reduce the number of private cars (the COE/license to own a car can cost more than the car itself and expires after a few years) so the taxis can go everywhere, but there are other factors - even in S'pore at certain times there aren't enough taxis around, and the buses and trains don't go everywhere...

Re:Predicting the traffic is easy doing something

[demonlapin]

I would. People tend to forget that, except for the hassles during commuting, cars are really great. I can travel where I want to, whenever I want to, and spend as much or as little time there as I choose. Public transit is useful only in a few situations - long-distance commuting, you can't afford a car, you live in a monstrous city that just can't handle the traffic, etc. You don't need a marketing division to make people want them (as opposed to, say, soft drinks).

I just wish I was rich enough to afford a private jet. That would be total mobility.

Re:Predicting the traffic is easy doing something

[Anonymous Coward]

The problem with mass transit in the DFW area is all the goddamn negros.

Here's the algorithm they use:

[phandel]

I believe the algorithm was leaked a few days ago:

if ((time == 8am-ish) || (time == 5pm-ish))
double jam =1; // double jam sandwich

if (city == LA)
long long jam = 2^64;

Re:Here's the algorithm they use:

[rush22]

damn you! I spent all this time reverse engineering the algorithm and someone makes the joke ahead of me.

Traffic Jams inGermany aren't the same...

[Anonymous Coward]

Some segments of highway are only two very narrow lanes as the highways haven't been re-done since the 1930s. In fact they are huge slabs of concreted put end to end with sometimes gaps in them. Exit turns aren't inclined cause these were made for cars that didn't exceed 30 Mph back in those days. These are, funny enough, the ones where you can speed without limit :)

On top of that, in southern germany, drivers never leave more that 2 inches in front of them as if they won a price if they can push you. I have never seen more unsafe driving practices. The whole German driving idea is to overtake you as soon as possible and then driving at the same or even lower speed than before.

So when cars collide, you get 40 or 50 smashed up together. And a traffic jam for miles and miles. It always happen every year, check the news evey winter.

Now, I don't think these correspond to most higways in the world, most other European countries have things like wider lanes, continuous asphalt, and drivers who have heard of the word "safety cushion" and respect the concept. Also for a highway system to work you need cars and cops. In Germany, all drivers think they are the cops and that the laws apply to the others.

So before you adopt this algorithm (90%!! That's VERY VERY Goog, what are you all complaining about?) make sure they model other than German traffics on it.

Also: US traffic is particular. If a cop car is there doing the speed limit, you get something like 80 cars behind in a herd and in most states people don't use the fast/slow lane systems to overtake or let faster cars have their way. It's a less dynamic system in a way and thus easier to model.

Re:Traffic Jams inGermany aren't the same...

[Anonymous Coward]

Some segments of highway are only two very narrow lanes as the highways haven't been re-done since the 1930s. In fact they are huge slabs of concreted put end to end with sometimes gaps in them.

Hmm, either you have been on a German highway the last time in the early 1990s in East Germany, or you simply have no clue what you are talking about. ;)

On top of that, in southern germany, drivers never leave more that 2 inches in front of them as if they won a price if they can push you.

You have regularly problems with drivers who don't know the concept of keeping a savety distance, that is true. Although I would say that is common in all Germany, not only in the south...

Re:Traffic Jams inGermany aren't the same...

[Anonymous Coward]

Also: US traffic is particular. If a cop car is there doing the speed limit, you get something like 80 cars behind in a herd and in most states people don't use the fast/slow lane systems to overtake or let faster cars have their way. It's a less dynamic system in a way and thus easier to model.

Thats because a lot of the assholes on the road in
the US are individualists. They don't give a fuck
what happens outside the bounds of their SUVs.

Too many times I've seen idiots in their SUVs
preventing faster cars from passing. When you
pull parallel to them you'll discover that
they're lauging into a cell phone or stroking
the bimbo in the passenger seat.

The result:
A huge tail of traffic behind that gets slower
and slower finally coming to a stop because
people get too close to each other then over-
rect when they see a brake lights ahead ...

Don't get me started on the soccer mums in their
Cadillac Escalades and Ford Expeditions.

I live in NY, btw.

- Moomin

Re:Traffic Jams inGermany aren't the same...

[anynameleft]

Well, okay, take a look at the Netherlands. We have a speed limit, wide lanes, continuous asphalt and all other things modern highways should have. And what is the result? Still about 40 traffic jams each day, so much that most news bulletins only talk about "the usual traffic jams". On the other hand, near Liege in Belgium you have strange concrete highways, sometimes with same-level crossings and people cycling on them, yet I have never seen a traffic jam there when we went to France.

Re:Traffic Jams inGermany aren't the same...

[Anonymous Coward]

>as the highways haven't been re-done since the 1930s. In fact they are huge slabs of concreted put end to end

this is only true for less than 5 percent of the whole autobahn. don't know where you got your information, but e.g. in my state there isn't one km of "old" autobahn...

Re:Traffic Jams inGermany aren't the same...

[Magius_AR]

The whole German driving idea is to overtake you as soon as possible and then driving at the same or even lower speed than before.

*blink* Dude, this is how passing should be done. Pass quickly as not to obstruct faster passing traffic and then get out of the way quickly.

The lower speed thing you said is wrong, yes.
If you pass somebody, you damn well better be going faster than them afterwards, even if that speed difference is only 1 MPH.

Kinda makes me think back

[OverlordQ]

to this page which was linked on slashdot a while back.

Re:Kinda makes me think back

[OverlordQ]

It's an obvious fake, look at that post, it says Subscribe . . I'm already a subscriber, so thus it should of said Subscription.

Are you kidding?

[la belle femme]

"in which dense traffic moves in unison, like marchers moving in step"
Some important facts, which the system designers seem to have forgotten: Real systems veer towards chaos and orderly behavior is anathema to all nationals except Germans. That apart, any traffic simulation design could be considered a success if it can be put to use in all conditions. I am quite sure that this system will come a cropper on Indian roads where cars have to jostle for space with cattle, cycles, pedestrians, three wheelers, two wheelers and so on. (Digression: I always find the sight of a stray dog crossing the busy road a lesson in traffic manners. The dog looks either side, ensures that there are no vehicles in vicinity and then crosses: A case of animal adapting his behavior to real time situations.) Human beings are not so adaptable and would do the best they could to defeat even the most beautifully designed systems.

Re:Are you kidding?

[anubi]

Yeh... the thing that makes it seem awful hard to me to predict traffic flow is not knowing how severely any given error will impact the general flow.

Its a highly nonlinear system... and I cannot even begin to guess the "gain" of the feedback loop. One driver mistake may end up as a matter discussed alongside the freeway as traffic continues largely unaffected... another ends up taking the whole freeway down for hours.

Every time I have ever played with such a system, I could use statistical strategy to predict what things should be, but I was usually quite a bit off from what things would actually come to be for any particular instance.

hehe... your dog analogy... this too is a system that weeds out dumb dogs that run out in the street without looking. Statistics says the dogs that run out in the street without looking will soon cease to exist, ergo the ones that still exist have likely acquired the art of looking before leaping.

I like your reference to the Germans. In my earlier years, I was subordinate to a fine German Engineer. He left me with a tremendous respect for the German ways of doing things. To this day I remember his wisdom and try to come up to his standards. He was by far the finest Engineer I have ever met, and I felt very honored to have worked under him.

this is more a social science than a science

[konekoniku]

because of that human factor involved, and all social sciences are inherently unperfect because of it. this isn't something unique to predicting traffic flow: psychology, sociology, political science, etc. can never make 100% accurate predictions because of the human factor involved.

No wonder!!

[rxmd]

I see why they did that in Duisburg and Essen, two of the five or six larger cities in the Ruhrgebiet.

The Autobahn A40 that connects the Ruhrgebiet cities from Dortmund via Bochum, Essen and Mülheim to Duisburg has the reputation of being the worst motorway in Germany, at least wrt/traffic jams. Guaranteed traffic jam 8-11am and 3-7pm.

It was probably just a professor's idea how to get to work more quickly in the morning.

Predicts jams BEFORE they happen?

[Anonymous Coward]

Wow... what'll they think of next.
In my town they only have a system
that predicts jams AFTER they happen.

Synchronized flow isn't stable

[KjetilK]

Hm, I suspect that the synchronized flow state isn't stable: The main reason: People are driving too close to each other.

If there is just a small change in velocity of one driver, the next guy is going to respond to it by hitting the breaks. The next guy is going to panic and hit it harder, and so it goes. I've seen this happen in real life many times: Just a small riple can make a jam, three or four cars involved is sufficient.

I fact, I think I saw an article about this too, it could have been long ago in Europhysics News or something. They like publishing stuff like that.

I'm trying to keep a lot of distance when I'm driving: Three seconds in normal situations (just count), and up to five seconds if I'm in synchronized flow. That way, I can absorb many ripples if the three or fours cars in front of me is slowing down. I think it helps, but surely it doesn't help a lot if it happens further ahead.

It is of course important not to lag too much when the cars starts moving again, so I generally speed up to follow in the start, but then try to build up some good distance, when the flow gets going.

But then, I'm just speculating...

Re:Synchronized flow isn't stable

[rritterson]

Your ideas about keeping more space on the road are sound, but if everyone were to do that (which, in my estimating and own experience is about twice as much space as most people give) the density of traffic would be much lower so fewer cars could fit on the roadway at a time.

Re:Synchronized flow isn't stable

[jrumney]

The density of traffic would be lower because the traffic is flowing more smoothly, not because less cars are using the road. This is a good thing.

Re:Synchronized flow isn't stable

[winwar]

Actually, a two second following distance is just fine. In good weather and conditions. More is nice, but people tend to move in... (granted, I have seen people move from wide open outside lanes into congested center lanes for no apparent reason, other than basic stupidity I guess...)

Now, back to reality. In periods of high traffic, it is NOT possible to keep such distances (at least without the liberal use of a grenade launcher). There are TOO MANY CARS. To keep a three second or greater following distance, there WOULD have to be fewer cars.

Re:Synchronized flow isn't stable

[jrumney]

Remember we're measuring following distance in time, not in distance. In periods of high traffic, following distances are usually much higher than 2 seconds, due to the extremely low speed of the traffic.

Re:Synchronized flow isn't stable

[Magius_AR]

Hm, I suspect that the synchronized flow state isn't stable: The main reason: People are driving too close to each other. If there is just a small change in velocity of one driver, the next guy is going to respond to it by hitting the breaks. The next guy is going to panic and hit it harder, and so it goes. I've seen this happen in real life many times: Just a small riple can make a jam, three or four cars involved is sufficient.

You say the main reason of this is people driving too close. I say it's poor drivers.

Just as often I've seen people randomly hit their brakes for no reason. All the time I'll see people overreact on braking (aka when those 3 people in front of me hit their brakes, I continue coasting and never need to hit the brakes. Anyways, if they kept the damn passing lane clear of traffic, there wouldn't be such an interruption in flow.

The main bane/cause/continuation of those jams are the self-righteous pricks that say "hell, traffic isn't going anywhere, I'll sit in the fast lane". They're probably the same people that sit in the fast lane when it's raining and they (or their car) is incapable of going faster than the speed limit, so they assume the rest is true of everyone else. These types of people are the causes of jams.

Just look on the road...whenever a cop is around doing the speed limit, traffic piles up. You think is due to speeding or following close? Hell no, it's due to traffic obstruction, namely people blocking all lanes and refusing to move.

Re:Synchronized flow isn't stable

[AK Marc]

The density of traffic would be lower because the traffic is flowing more smoothly, not because less cars are using the road. This is a good thing.

At a given speed, lower density means lower capacity. Lower capacity is a bad thing. Higher density is a good thing.

They just need to take this into account...

[ebyrob]

Use the "effects" of the prediction to get the desired results. For example, if you want clear sailing on Route 9 going north at 4:00pm, predict bad traffic ahead of time.

If you want more predictable traffic patterns, fudge prediction differences downwards so that less people will take action based on the predictions.

It's all one big feedback loop, dampen accordingly.

Re:They just need to take this into account...

[Anonymous Coward]

For example, if you want clear sailing on Route 9 going north at 4:00pm, predict bad traffic ahead of time.Problem is, then there is a major traffic tie-up on Route 7 at 4:00pm as everyone tries to avoid Route 9...

Wonder if they will predict...

[SalsaDot]

...the slashdot effect on their site

oops, different flavour of jam

Other anti-jam techniques

[Bushcat]

Some of the anti-jam techniques I've seen around the world tend to be based on getting information to the driver as quickly as possible.

The ring road around Eindhoven has recommended speed indicators that show what speed you should be travelling at to hit the next light on green. It seems to work quite well.

The Hanshin Expressway network around Osaka has video processing equipment that can automatically recognise traffic congestion, including the characteristics of traffic accidents. It then alters roadside information boards to route drivers around the congestion. Other areas do the same thing.

adio bReacons update vehicle navigation systems in real time to show time to destination, congested roads, and if you're following a planned route, will re-route you as traffic conditions change.

Unfortunately most signals in Japan aren't load-sensitive, but set to 1 minute green in each direction irrespective of time of day, day of the week or class of road. So circular routes around Tokyo, for example, become major barriers to traffic entering in the morning and leaving in the evening.

On Japanese freeways, the major cause of congestion is the humble tollbooth.

Re:Other anti-jam techniques

[VCAGuy]

It's funny that you should mention the tollbooth. Here in Orlando, Florida, the local Expressway Authority teamed up with the University of Central Florida and had this tool called TPSim (that's Toll Plaza Simulator) made for them. The tool allows them to simulate traffic through the plaza to determine the optimal number of attendant/cash lanes, change lanes and electronic-only lanes to best serve both directions of traffic. It really has helped the congestion around the Holland East (which was really bad before they started using the tool).

Results?

[Anonymous Writer]

A traffic simulation system is helping drivers by predicting jams up to an hour before they happen.

Let me guess- it predicts that traffic jams will happen on weekdays, around 5pm.

Re:Results?

[supretigre]

AND don't forget weekdays, at about 7:30 am!!! How f-ing hard is it to predict something that happens every damn day in a lot of places?

Re:Results?

[winwar]

Well, it is interesting research.

But, for any city of any significant size, it's essentially:

We can rigorously predict traffic jams with less certainty than the average driver.

The research is just not terrible useful at the present time.

Academic proof...

[johannesg]

Since this problem OBVIOUSLY reduces to the Halting Problem, it is easily proven that you cannot, in fact, predict traffic jams.

Re:Academic proof...

[TheLink]

Well Traffic jams definitely are to do with halting... Often they're caused by halting problems too :).

Re:Academic proof...

[johannesg]

I was hoping that the apparent juxtaposition between the actual truth of the statement, and the fact that "Halting Problem" does also refer to a traffic jam, would create the illusion of humor... Until you shattered it.

;-)

Re:Academic proof...

[Zebidiah]

Well done!

You've proved that by its own success that it doesn't work! :-)

Re:Academic proof...

[TheLink]

Sorry couldn't stop myself ;)

Traffic jam solution

[reboot246]

Just some ideas.

1. Reduce the number of redlights needed and make the ones left more intelligent
2. Use more public transportation, especially for commuting
3. Get the idiots off the road
4. All exits should be to the right, because exits to the left mix fast and slow traffic
5. Restrict 18-wheelers to the right lane only and make them use by-passes when available
6. Get the idiots off the road.
7. Move wrecked vehicles off the highway as soon as possible, even if it means coming back to tow them later. Why shut down a whole interstate because of a vehicle in the median waiting for a tow truck?
8. Make not using turn indicators a MAJOR fine, at least $500.
9. Extend onramp and acceleration lanes
10. Get the freaking idiots off the road!

Idiots include - people talking on the phone, reading the paper, putting on makeup, sight-seeing, watching DVDs or TV, . . . you get the idea.

Re:Traffic jam solution

[Anonymous Coward]

> 2. Use more public transportation, especially for commuting

Eh. People smell bad. Sometimes I smell bad, too.

> 4. All exits should be to the right, because exits to the left mix fast and slow traffic

So do some exits to the right. Particularly one where I live. People hit 60 right into a red light.

> 8. Make not using turn indicators a MAJOR fine, at least $500.

Oh, c'mon. We *all* forget to signal sometimes, and sometimes it just isn't neccessary.

> Idiots include - people talking on the phone, reading the paper, putting on makeup, sight-seeing, watching DVDs or TV, . . . you get the idea.

A friend of mine actually bought one of those airhorns (y'know, cans of compressed air with a loud-assed horn on the nozzle). He was out of air in an hour, blasting it at idiots.

Re:Traffic jam solution

[The FooMiester]

4. All exits should be to the right, because exits to the left mix fast and slow traffic

5. Restrict 18-wheelers to the right lane only and make them use by-passes when available

I agree with #4, but #5? You'd never get on the highway around here. Not only that, but we have these things called "hills", and some trucks can only do 35 or 40 up them. Imagine if the right lane were solid with trucks moving along at 35mph! If you could get on, you'd still have to deal with the blue-hairs driving along at 45 in the other lane.

Re:Traffic jam solution

[reboot246]

Oh, c'mon. We *all* forget to signal sometimes

Well, my first thought was the death penalty, but I backed away from that.

Re:Traffic jam solution

[TheLink]

Heck just 3)6)10) will get rid of 90% of traffic, you and me included ;). Voila - no traffic jams. Be careful what you wish/ask for.

Seriously. I dunno about you, but in my experience nearly every driver is an idiot at least once in a few months. It only takes 1 idiotic action to screw up traffic for 20 to 10000 cars (depending on the idiocy of the act).

If everyone were highly skilled and knowledgeable drivers, concentrating 100% on driving, and equiped with top notch well maintained vehicles, I bet the traffic densities and average speeds could go up very significantly.

Of course it helps if you get rid of the drivers who are _consistently_ incompetent, or driving cars that just aren't reliable or can't achieve the average _desired_ speed...

Re:Traffic jam solution

[reboot246]

A couple of added thoughts -

Our main problem on major highways, other than interstates, is that we've turned them into shopping aisles. The U.S. and state highways going into and out of cities are lined with malls, shopping centers, eateries, etc.. It makes for traffic jams possible even not in rush hour. I've seen bypasses built only to be crowded with shopping centers in only a couple of years.

The problem with interstates is that they go THROUGH cities!! Whoever came up with that brilliant idea should be drawn and quartered. It would have been SO much better if they had been built to go around downtown with spurs built to go into town.

If we're going to insist on commuting by automobile, then the highways going into cities are going to have to be limited-access highways. Leave the interstates to through traffic. Go look up the word "interstate".

It would help matters greatly if traffic engineers actually got out and drove on the streets and highways they try to manage. I imagine that their commute to and from work is quite problem-free.

Re:Traffic jam solution

[AnimeFreak]

Vancouver has it right when it comes to freeways in a sense.

http://www.sfu.ca/~falun/images/large.gif

The four freeways that feed into Vancouver proper do not even go near the city centre. They just feed the city and that is it.

Rubbernecker screen

[abrinton]

Why has nobody made a "Rubbernecker screen"? After an accident or incident, clear the roadway and put a screen up. Nobody can see the accident, and therefore they start moving again once the road is clear.

You could even pay for it by selling ad space!! (I can't beleive I suggested that!!)

Re:Rubbernecker screen

[LordLucless]

So your idea is to put something designed to get people to look at it on something designed to stop people looking. I'd patent that one real quick!

Dumber than a very dumb thing

[Anonymous Coward]

Why predict tomorrow's jams when you can measure today's?

I *know* (yeah, real psychic, me) where Monday mornings jams will be. Same place as last Mondays.

This is right up there with predicting the weather - "same as yesterday" gives you 80% accuracy.

We do not need geeks like us to help design the road networks. If you want to solve congestion with a match-to-demand model, just make them wider where the jams are now.

imagine yourself...

[brunokummel]

..and other 300,000 users in the "prediction" system at the same time: "hmm I-95 is gonna be jammed in one hour from now.I better go home a little bit earlier today." tada!! =D we have succesfully jammed the road with one hour in advance!! What an improvement! Thanks again modern physics!

I think the problem is too many cars on the road

[Stevyn]

Well, yeah, call me captain obvious. I saw this show on the history channel a little while ago that did a computer simulation of traffic on a highway. It actually showed that up to a certain point, a lot of cars on the road is fine. However, once it hits a certain point, the flow dramatically drops.

Living in New Jersey, traffic is a part of my daily life. I've noticed that even when there are a lot of cars on the road, the traffic can still move along above the speed limit in rush hour. But there is still that point when all the traffic just seems to stop and you're crawling along at 10 miles an hour.

The show also mentioned a study the government funded. They spend several million dollars to try and figure this problem out. Their conclusions: "there are too many cars on the road." Now, for the low low price of $20,000, I could have told them that.

Traffic Waves - research and animations

[WRXFiles]

In 1998 William Beasley posted a paper on Traffic Waves in Seattle, subtitled "SOMETIMES ONE DRIVER CAN VASTLY IMPROVE TRAFFIC".

The site has great animations and excellent explanations of the impact of different drivers actions on the overall flow.

Worth a look: Traffic Waves

I can predict Traffic myself.

[jellomizer]

i87 Between 8:15-8:45 High Traffic south bound from Exit 8-6
i87 Between 4:45-5:15 High Traffic North Bound between Exit 4 and 5 and exits 7-8

It is easy to predict in Albany NY because all the state workers work outside of albany and many of them live up north. In Saratoga County. So when they go to work they tend to go south towards Troy, Latham and Albany Plus there is a big bridge between exit 7 and 8 and people become stupid on the bridge (There are no tools and the same amount of lanes) and slow down. and Exit 4 is where the airport is. So the traffic barring any accidents is like that almost every work day. Except in August where the traffic is Stop and go going north from exit 1 and up. Because Saratoga is the August place to be. Luckily I have an alternate route.

Re:I can predict Traffic myself.

[thomasdelbert]

It's even easier in Rochester, Minnesota:

TH-52 until that FREAKING CONSTRUCTION is complete.

- Thomas;

Man i am so tired...

[zors]

...of Slashdot being so euro-centric.

Re:Man i am so tired...

[Anonymous Coward]

It's the cutting edge of politics in a very extraordinarily boring way ;))

Not news

[Jugalator]

SimCity did this ages ago!

The view can be seen here in the most recent version.

Re:Not news

[kisielk]

Exactly! SimCity has the solution... Just replace all your roads with railroads! No traffic :)

My tired eyes

[Anonymous Coward]

I was somewhat confused when reading the prediction of James. ...we seriously don't want James to happen

Predicting Jams is Easy

[Anonymous Coward]

To prove it, I'll go on record predicting a major jam in downtown D.C. tomorrow night.

M25

[bobblebob]

On the M25, the road that circles London, grid lock is avoided with variable speed limits and discouraging drivers from excessive changing of lanes. The road is constantly monitored by CCT and the speed limit set accordingly. The speed limit is enforced by speed cameras. In the rush hour the speed limit is reduced to 50 mph as apposed to the usual speed limit of 70mph and this, some how, reduces the risk of grid lock. Also if there's congestion at say junction 15 the speed limit will be reduced prier to the congestion, say at J13-J14. This help to reduce more trafic build up at J15 and gives it time to clear.

Screw all the simulations!!

[hlygrail]

In a former job many years ago, I did a lot of highway driving (deliveries and field service) and spent a good deal of that time thinking about traffic patterns, both as a mathematical system as well as a design process. My final conclusion after 2 years was that there are too many cases of poor road/intersection design and WAY too many cases of pathetic traffic light design, and even more cases where traffic lights hinder traffic flow in a very severe way, while not providing any of the originally-intended safety of said traffic light. The prevailing "wisdom" seems to be: if there's a traffic problem, put up a stoplight to control it. Not.

Traffic dynamics is fascinating, and certainly deserving of some studious focus, but none of this means a single, blessed thing to me unless people will:

at least drive at or near the speed limit, conditions permitting -- if you can't, get out of the way learn how to *accelerate* on an on-ramp so they merge at or near highway speeds -- when you don't, you cause massive traffic problems; stop tailgating so badly that traffic becomes a wave of stop/go -- backing off just a little bit means no more brake-light domino effect; stop driving 20MPH (that's 32.18KM/H for you Euro-geeks) below the speed limit in the left lane (or the right lane for your Euro-geeks) when the prevailing traffic is clearly ALL having to go around you -- you are a huge HAZARD if you're driving too slow in the left (Euros and Bahamians: right) lane;
-or-
get out of my way and let me do the above.

I'd make the common observation that older folks tend to drive slower and tend to do so, totally oblivious, in the left lane. But then there is at least an equivalent problem in the soccer moms and high-strung business suits on cell phones driving SUVs and mega-SUVs who pay even less attention to what they're doing. Heck, I've seen people driving down the interstates here (I-40, I-95, I-85 -- pick one) during morning traffic, travelling at over 80MPH and reading the morning paper!

That said, I think traffic problems tend to be less a mathematical/system problem than a brainless person problem, in many cases. Not all, but very many. Sadly, you can't "in-idiot" a driver, or a person for that matter.

I've always wished for a traffic law that gave every driver a dart gun. When someone does something obscenely stupid or hazardous (e.g. driving in reverse on a 65MPH intersate because they missed their exit), you shoot them with the dart. Three darts means your vehicle is incapacitated for 30 days. (Hmmm... I think RFID tags would be perfect for this!! :) If your vehicle is tagged more than 6 times in 2 years, you lose your license for 1 year.

Harsh? Definitely. But consider the *costs* of traffic in lost time, productivity loss, frustration/rage, increased fuel consumption, vehicle wear and tear, air pollution, etc. Pulling one person or one thousand people out of the traffic system to improve the flow for the masses sounds like good planning to me.

Oh, and please direct any comments about my tendency to drive well above the speed limit to /dev/get_out_of_my_way...

Re:Screw all the simulations!!

[The FooMiester]

The problem with this are the people who are for the 55mph speed limit. They think that "slower is safer" for some reason. I don't understand how that works, since when NJ raised the speed limit fatal accidents went down.

What people don't realize, is that if every stupid speed limit were obeyed, no lights jumped/ran, all stops made completely, traffic would be many times worse than it is.

Re:Screw all the simulations!!

[gokeln]

I saw a fellow in Dallas a while back who had in huge red letters on the back of his car, "COURTESY SAVES LIVES". He always had a big space in front of him, where anybody could merge, and a smile on his face if you passed him. And, I still try to emulate his attitude. He made a difference in the traffic each day he drove, and in the attitudes of many people who saw and followed his advice.

Re:Screw all the simulations!!

[rush22]

What people don't realize, is that if every stupid speed limit were obeyed, no lights jumped/ran, all stops made completely, traffic would be many times worse than it is.

Yeah traffic lights are stupid, I never pay attention to them, I just go whenever. It helps traffic flow.

Now, anyone who agrees with me, we can just take your license away right now, because you're obviously too immature to be on the road. And if any of you seriously want to reply to argue the merits of running red lights, just kill yourself now before you kill someone else.

Re:Screw all the simulations!!

[The FooMiester]

I'm not talking about blasting thru a red light at full speed out of the blue, I'm more talking about the 1/4 mile line of traffic at the mall waiting to get out, and the last few people going thru after the light turns red, and before the other side gets the green. I've sat and watched it many times, and the bottom line is that the lights just aren't set up properly. More cars would build up at the exit than would get thru on the green light.

Re:Screw all the simulations!!

[AK Marc]

Yeah traffic lights are stupid, I never pay attention to them, I just go whenever. It helps traffic flow.

When the lights are timed poorly, two people running the red at the end of every light cycle will reduce the backup by miles. I have seen a number of lights that are green for an insufficient time to allow the traffic through. The result is that the line grows by a about two cars every cycle. Over a two hour rush hour and a 3 minute light, that makes for an 800 ft backup or so. I've seen it worse than that in practice, so I know my numbers can't be too far off. If a car jumped the green and one or two ran the red, then the traffic backup would be 0 ft.

I'm not arguing about whether that is safe or recommended, but it would make traffic worse if no one ran the reds, and that is all they said. Of course, the real solution is to properly time the lights, both in activation and duration, in order to minimize traffic. But if all the traffic laws (and the signs and signals enforcing them) were set with traffic flow on an equal basis as safety, there would be so few violators that the income from traffic violations would drop. To prevent that, many laws are set with revenue above safety, followed by efficiency far below that.

Re:Screw all the simulations!!

[rush22]

OK I see what you're saying. Sorry about that, it was kind of unwarranted. I've been exposed to so many outrageous nutcases on ./ that I often tend to overreact and become one myself. :P

They can use route selection

[asterix_2k1]

The problem with prediction systems is that they people use them to thwart the 'unwanted future' and in turn create an unknown, potentially more chaotic situation.

In this case, as the story points out, all drivers accesing the prediction may use the same exit roads, or the same alternate route - which defeats the purpose of the system.

One useful alternative is : make a centralized system, say to monitor a city traffic. Whenever, a driver perceives that he is entering a slow traffic/jam situation, he contacts the system. The system, dynamically assigns one of the alternative routes to the driver. Here it is important that there be atleast two or more alternate routes (because in case of only one alternate route, it will become jammed). The route selection by the system can be done probabilistically so as to get highest 'expected' utility, i.e. best route utilization and throughput. Also, it should be easy in principle, say using GPS, for the system to monitor the exact traffic at any point in the city.

Infact, I did a similar computer experiment in my undergrad days - comparing static route selection vs dynamic route selection and no guesses as to who came out to be the winner by miles !

Devils Dictionary

[Anonymous Coward]

This reminds me of Ambrose Beirse's comments on Barometers being a wonderful invention for telling us what kind of weather we are having.

It doesnt take a super computer to tell me that there will be a major traffic jam if an accident happens on a major artery during rush hour.

Left lane?

[Benm78]

Your traffic tips and hints seem mostly correct, bad driving is something that definitively causes many problems such as traffic jams. However, bad driving can be modelled to a certain extent so it does not make all models useless per se. About the left/right lane and Europe: ALL european countries on the mainland drive on the right (as opposed to left) side of the road, just as Americans do. The left lane is therefore the fast lane in Europe as well. On highways, nearly all off-ramps are on the right side. There is 1 country in Europe that drives on the left site: the UK. All other countries that I've visited that drive on the left side are in Asia.

Makes sense

[Anonymous Coward]

I know where traffic jams will be just from knowing the route and the environmental conditions. Traffic is a realtively simple related rates type fluid dynamics differential. From what very meager Diff Eq. I have had, the techniques used to set up a system of equations, makes it easy for the mind to make such predictions. Part is weather, part is the time of day and the type of driver that is there, part is events, part is road geometry.... af few impulse factors... I can usually guess when to leavee early, and which route to take... speaking of which... there is a light mist right now... and I need to get to work... why the hell am I writitng this instead of....ah the slashdot virus effect! stealing clock cycles from the brain...

Traffic Waves

[pipingguy]

See also this site.

Have you ever been driving on an interstate highway when traffic suddenly slows to a crawl? You inch along for many minutes while waiting to see the accident which must have caused the jam. At the same time you also curse the "rubberneckers" who are causing the whole problem. But then all the cars ahead of you take off at high speed. The jam is over, but no accident, no police cars, nothing. WHAT THE HECK WAS THAT! A traffic jam with no cause? In the rear-view mirror you see all the poor saps behind you still stuck in the jam. But why? If all those people could just speed up at the same time, the whole traffic jam would evaporate. Why don't they ever do that? What caused the mysterious slowdown in the first place?

Re:Traffic Waves

[KjetilK]

Wow! That's exactly the same observations I have done! I have too noticed that I can actually clear up a jam by keeping like five seconds up to the next car, slowing down slowly.

There's a funny Traffic Wave Generator in Drammen, Norway. Unintended of course, but nonetheless. It's a longish bridge going from northeast to south in this picture, and at the northwestern end, there's a lot of traffic coming in, and at a relatively high pace. The speed limit is something like 90 km/h, which means the average speed is probably well in excess of 100 km/h. Then the limit on the bridge goes down to 70 km/h, and at the same time, it merges to a single lane in either direction. Bound to be trouble as it is... But to make matters worse, shortly after the 70 sign (perhaps 50 meters), there is a photo box, that, if it has film it in, will shoot pictures of anybody speeding.

But the true sign that nobody in authority has the faintest idea why this is the most hated persistent traffic jam in the country is a big, official sign saying "In case of a jam, follow along!" I mean, WTF are you going to do, I'm not sitting here for fun (or profit) you know! :-(

What they officially seem to be advocating is the fast acceleration. But not everybody can. For example with my mother's little engine, I can't... There is very little you can do to assist evaporation, as he well argues in the paper.

So they created the worst, high amplitude traffic wave in the country by putting a traffic control camera in exactly the worst thinkable spot. There would necessarily be a traffic wave there anyway, but it is making matters so much worse.

I admit that there is a thing I do not quite understand. The jam often extends the whole bridge, and does not dissolve before a km afterwards. It would be interesting to study this from the air...

Re:Traffic Waves

[pipingguy]

Ugh. Norway, like Canada, Sweden, the Netherlands, and France just do not count in this forum. NIH.

I'll bet that most Americans don't even know what Norsk Hydro is or what it does.

Ignorance is bliss.

Re:Traffic Waves

[KjetilK]

Actually, apart from the occasional troll, which is easily ignored and soon modded into oblivion, I think slashdotters in general are quite receptive and interested in what people from outside the US has to say, and generally acknowledge that /. is very much an international forum.

Re:M25 - Couple of typos in parent

[memoriesofgreen]

On the M25, the car park that circles London, grid lock is caused be idiots who ignore variable speed limits and encourages said drivers to put their foot down and lane jump. The road is constantly monitored by CCT and the "your taking the piss, if only" speed indicator set accordingly. The resulting crush of commuters is a source of amusement by the operators who watch from air-conditioned offices. In the rush hour the average speed is reduced to 5 mph as apposed to the usual speed of 50mph and this, some how, is called a transport policy in the UK. Also there's usually congestion at junction 15 and this will cause problems for Junction 16 - 18,19,20 - 24, and 25 - 14 the speed limit will be reduced further to 0mph. This helps us Brits to remember the scene in turnin during the Italian Job.

IC19

[Anonymous Coward]

IC19 fails to work for this system, since the road that heads from Sintra to Lisbon is jammed at any hour you pick making impossible to predict jams.

we use to say having a Ferrari on IC19 is like having a fastest linux amd64 puter on a 56kbps connection

Atlanta, GA is a unique situation.

[CausticPuppy]

In recent months, most of Atlanta's big traffic tie-ups (even worse traffic than usual) are caused by people threatening to jump off of overpasses.

So now for this traffic prediction to work, it basically has to predict bridge jumpers. Fortunately, as the government profiles us more and more and tracks us via RFID tags or whatever, it may be possible not only to predict when a person gets suicidal, but possibly also to predict the most likely date/time and even the correct overpass they will use.

</TongueInCheek>

No, it's chaotic Re:Isn't this a statistical probl

[WolfWithoutAClause]

just ONE driver makes an error... then all hell breaks loose.

That's just one butterfly though. It's the jam spreading out from there that causes the problems.

Both are fundamentally chaotic.

Yes, but the principle of chaos is that small changes to the initial conditions lead to an exponentially divergent change to the result.

So, the earlier you are able to detect a single car crash, or the traffic jam around it, the quicker you will be able predict a city-wide gridlock- and perhaps arrange for your own butterfly to manage the problem perhaps you change some traffic lights so as to prevent gridlock entirely.

People are just flat out unpredictable.

So are air molecules.

Wasted time?

[Anonymous Coward]

Recently calculated the number of days I've spent in the past year in traffic (i.e. day's counted as 12 hours of awake time and 12 hours of leisure time). If it weren't for commuting I could've worked an extra ONE AND A HALF months in a year! That is pretty ridiculous. I'm sure most people spend at least an hour or two driving!

I wish I could just work from home

Avoiding Traffic Jams

[aismail3]

Bill Beaty investigated how to avoid traffic jams by recognizing the intermediate state of synchronized flow and undoing the damage six years ago. Apparently, a traffic jam can be stopped, even once started, by a single car.

Traffic Waves

Color-blind mode!!

[beej]

Their applet has a color-blind button on it! This absolutely rocks!! I mean, when I can't see the difference between "light" and "vicious" traffic, I'm pretty much screwed. This simple addition made the applet useful to me again!

When building software like this, you do a great service to take us R/G colorblind people into account. We're not as rare as you think! BWHAHAHA! ;)

Re:Color-blind mode!!

[TheLink]

Am I wrong or would carrying a piece of green or red transparent film help? That way red or green could look black/darker to you whilst letting the other colours through.

It's easy to be right 90% of the time

[Len]

It's easy to predict traffic correctly 90% of the time. Traffic jams tend to occur at rush hour on certain roads. You can predict the traffic 90% of the time just using a clock.

The other 10%, when traffic jams occur unexpectedly, is the hard part. But it's what people need to know.

Re:I think the problem is too many cars on the roa

[TheLink]

Once you have a certain density it just takes an "event", and then traffic will slow. An "event" could be someone just braking a bit too much, or doing something that causes someone else to slow down dramatically. Once that happens you get a wave of slowed down vehicles propagating down the highway. Without a high enough density, the "event" can be avoided or may not even affect other vehicles significantly.

The worst are events that cause chokepoints. Then you end up with the "hourglass" effect. Lots of near stationary particles all waiting for their turn for free flow.

RE: road design plays a big role

[King_TJ]

Here in St. Louis, it's my observation that with relatively few exceptions, the big traffic jams always happen around the same times and places. Predicting a traffic jam with good reliability is possible primarily because inadequate roads/freeways for traffic flow during rush hour means really good odds of a problem in those key areas.

Basically, we have a situation where a large number of people need to travel from suburbs in the west to their workplaces in the city itself (east). On the ride home, you get the reverse. There are only 2 major interstates running east to west that will be dealing with most of this traffic flow, so they're almost a "given" for a traffic jam. (Technically, there's a 3rd. one - Interstate 44, but it's far enough south that it doesn't get the traffic we see on I-40 or I-70, which are more direct shots back to the majority of the suburbs in the west.)

You can also currently expect traffic jams heading to the northern parts of St. Louis, mainly because road construction has restricted traffic flow on our major interstate taking people north/south (I-270). in that section of the highway.

Cool!

[Anonymous Coward]

I knew some people working on agent modeling stuff like this. The basic idea is to get "good enough" results. Good stuff. :)

The Traffic Wave site..

[bmantz65]

He said that by leaving a large gap between him and the car in front of him, he was able to ease jams a bit. Umm, I don't know if Seattle is different but if I left ~1000 ft. between me and the car in front on the Parkway here in Pgh during rush hour, the people behind me would likely honk their horn and get pissed. Not to mention, a few cars from the other lane would jump in that space, thus negating the effect of the 'experiment.'

Reminds me of a story.

[NarrMaster]

I remember reading a story about the law of averages breaking down... everyone decided to go to a ball game at the same exact time, everyone decided to shop at the same time, cause they figured no one would be shopping then, etc. If anyone knows what I'm talking about, please, give me the source, I can't stand knowing this story without knowing what it is called.

Missing a critical one.

[juuri]

11) No stop lights within 1/4 mile of a high traffic exit. The spill over from lights such as these on loops cause much of the traffic congestion.

Ripped off again

[Tjp($)pjT]

I did similar work in college in the mid to late seventies. I did not use "live data" but had I access to it at the time, it would have worked well. One of the very interesting items is that a single car on a two lane Interstate (four lane if you are in a region that counts both sides, kind of like deer antler points being regionally different too), and if that car moves 5 miles an hour slower than the ambient traffic, and if the road is at over 50 percent capacity, the slowdown will cascade and the result will be slow to stop and go traffic at around 30-35 miles per hour.

Then I moved to Taxachusetts in the end of the seventies and discovered Rt. 3 north from MA to NH at 4:30 to 7 PM accurately was predicted by my traffic model. BTW ambient can be over the posted speed limit and it still slows the whole flow down.

Strange Attractor

[Randym]

It seems to me its nearly as impossible to predict a traffic jam as it is to predict stock prices. Both are fundamentally chaotic.

As I read your comment, it struck me that the flow of cars past a given point resembles the flow of water from a drippy faucet, and that both can be modeled as a 2D strange attractor. Just as you can map the time differences between Drips [ D1, D2, D3, etc. are absolute time of drips; T1=D2-D1; T2=D3-D2; etc.], you can map the space differences between cars. Then map it on cartesian coordinates thusly:

Point1: T1, T2

Point2: T2, T3

Point3: T3, T2; etc.

When traffic is flowing smoothly, the difference in time between cars is essentially random, so the map will just be a bunch of randomly scattered dots. But when jams began to occur, the average length between cars shortens, and the dots began to describe a spiral inward. At total jam, the length between cars is, essentially, zero, leading to a dot centered at 0,0.

It's the point when the average length between cars begans to shorten that is when the graph becomes most interesting. Also, this is where intervention might occur "on the ground" so to speak.

One drawback of this graph is that it describes only a single location along the roadway. You'd need many of these graphs to simulate a large traffic situation.

And, as to the poster's note that People are just flat out unpredictable., that's only true in the *individual* sense. In the aggregate sense, people are *remarkably* predictable. This traffic problem happens to be an interesting problem because it involves *both* aggregate (so it's fairly predictable) AND individual (ONE driver makes an error...) behavior.

I know exactly how to predict traffic jams...

[LamerX]

Look at the clock. I predict that every weekday at 7:00am-8:00am and 5:00pm-6:00pm traffic will be at its heaviest. Avoid driving at these times and you will be set. Also, I predict that road construction will cause traffic congestion as well. In the foreseeable future, I also predict unknown random traffic accidents to cause congestion. Use your radio tuner device in your car to listen for reports of these strange incidents and avoid those routes!

This is exactly why I hate good samaritans...

[IOOOOOI]

... you know: the people who slow down not because they have to, but specifically to let someone in from an adjacent lane, or driveway, or side street, whatever. They think they are being a courtious driver, sharing the road, etc.

They are oblivious to the fact that for the one vehicle they let in, they are causing at least one other car behind them to have to slow down and potentially miss the next traffic signal. Or worse, the person they let in does something to aggrivate the situation further:

• drive too slow
• deciding that they really want to be in the left lane so they hold up the first lane while they force their way into the next lane, causing that lane to have to slow down and so on
• how many times has someone in front of you let a person in from say a gas station only to have that person make an immediate right turn and you all wind up waiting for the pedestrians?

In each one of these cases, if the good samaritan had simply followed the rules of the road and maintained their right of way it would have been better for all. The few seconds of time saved by the benefactor is a fraction of the compounded delays of the people behind the samaritan.

What aggrivates me even further, is that driving schools preach that this kind of behavior is good and will help reduce the number of accidents. BULLSHIT. The rules of the road were designed to prevent accidents by increasing the predictability of events that occur on the road.

Don't get me wrong, there are times when you are compelled to give up your right of way or even blatantly break the rules in order to maintain your safety and that of others. Unfortunately, people have been taught by authority figures (driving schools, often as a condition of getting a ticket dismissed) that this should be a matter of practice. This is reinforced by the fact that if you sound your horn (as a safty warning of course :) at someone who suddenly decides to follow this advice, you are considered an asshole.

Unless it is to mitigate an unsafe situation, follow the rules of the road people!

El Farol bar problem, minority games

[Anonymous Coward]

What if people start using it? They avoid the traffic jams, thus no traffic jams. Wrong prediction!

Your question is reminscient of the El Farol bar problem. In that problem, there is a popular bar (the "El Farol") that everyone wants to attend. However, it's so popular that it gets very crowded. People want to attend when it's not crowded (thus making it an example of a "minority game"), but the problem is, if everyone uses the same deterministic model to predict when the bar will be least crowded, they'll all go at that time and invalidate their model. This is a well-studied problem in the dynamics of complex systems.

Re:El Farol bar problem, minority games

[Zebidiah]

Shame you posted this as AC. It sound very interesting. Was/Is there a solution?

Re:El Farol bar problem, minority games

[Anonymous Coward]

There are a lot of attempts at solutions, depending on exactly how the problem is phrased and what solution is considered acceptable (what constitutes "maximizing the satisfaction of all parties"). Some approaches have included evolutionary adaptation, inductive behavior, bounded rationality, and statistical mechanics. You can search the physics e-print server for "El Farol" or "Minority Game" in titles or abstracts for some physics-based approaches (mostly in the cond-mat archive); Google will get you other approaches. Look around at the Santa Fe Institute; that's where the actual El Farol bar is that spawned the problem.

Re:El Farol bar problem, minority games

[Nurseman]

People want to attend when it's not crowded (thus making it an example of a "minority game"),

There is an old quote attributed to Yogi Berra "Nobody goes there anymore because it's too crowded" I always liked that one :-)

Of course.

[Rufus88]

Didn't they teach you that the Halting Problem was unsolvable?

This is exactly why I hate good samaritans...

[IOOOOOI]

... you know: the people who slow down not because they have to, but specifically to let someone in from an adjacent lane, or driveway, or side street, whatever. They think they are being a courtious driver, sharing the road, etc.

They are oblivious to the fact that for the one vehicle they let in, they are causing at least one other car behind them to have to slow down and potentially miss the next traffic signal. Or worse, the person they let in does something to aggrivate the situation further:

• drive too slow
• deciding that they really want to be in the left lane so they hold up the first lane while they force their way into the next lane, causing that lane to have to slow down and so on
• how many times has someone in front of you let a person in from say a gas station only to have that person make an immediate right turn and you all wind up waiting for the pedestrians?

In each one of these cases, if the good samaritan had simply followed the rules of the road and maintained their right of way it would have been better for all. The few seconds of time saved by the benefactor is a fraction of the compounded delays of the people behind the samaritan.

What aggrivates me even further, is that driving schools preach that this kind of behavior is good and will help reduce the number of accidents. BULLSHIT. The rules of the road were designed to prevent accidents by increasing the predictability of events that occur on the road.

Don't get me wrong, there are times when you are compelled to give up your right of way or even blatantly break the rules in order to maintain your safety and that of others. Unfortunately, people have been taught by authority figures (driving schools, often as a condition of getting a ticket dismissed) that this should be a matter of practice. This is reinforced by the fact that if you sound your horn (as a safty warning of course :) at someone who suddenly decides to follow this advice, you are considered an asshole.

Unless it is to mitigate an unsafe situation, follow the rules of the road people!

Why estimate when you can get live updates?

[Anonymous Coward]

This site has very good reports of traffic conditions all over California. They're very good but tend to lag by about 20 minutes.

Re:Why estimate when you can get live updates?

[JazzManDRP]

"...live updates..."

"...tend to lag by about 20 minutes."

*ahem*

Get rid of the StupidPeople.

[rice_burners_suck]

I have an idea. How about, instead of predicting traffic jams and then working around them, let's find all those StupidPeople, with no place to go and all day to get there, who drive too damn slow on the road when there is NO REASON to do so, like going 50 in a 65 when there are NO cars ahead of them for 10 miles, and if there are 3 lanes on the road, then you can rest assured that three such StupidPeople will find each other and drive side-by-side, too slowly, when the whole world is behind them trying to get to where it's going; how about, let's find all those StupidPeople, take away their licenses to drive, and then we can enjoy the road when it's moving along at the correct speed, as opposed to crawling along, making a 20 minute drive into a 2 hour one, and using up eight times the quantity of gasoline, which costs way too much.

I was going to suggest shooting the StupidPeople, but that is a somewhat mean thing to do.

What happens if you mix both: September surprise!

[ArsenneLupin]

In the case of traffic, drivers tend to move to an alternative route, which is shown to be less congested, making it more congested in the process. In the case of the stock exchange, people will buy the stocks which are predicted to rise in advance, which makes them rise or fall earlier than predicted. In both cases (traffic and stock market), actors seeking their personal gain have an active interest of "messing up" the prediction. Predicting human-dependant events is inherently difficult, unless the event is of the kind that gives a (perceived) advantage to "going with the flow", such as fashion or US presidential elections. In that case, its easy to predict, because the prediction even helps the event come true (self-fulfilling prophecy)

And what happens if you mix both type of predictions, the self-unfulfilling (stock market) kind, and the self-fulfilling kind (US pres elections).

Yes, you get a nice September surprise (due to all the plebs leaning hard on the dollar, in anticipation that the big guy does...). But, in this case, a September surprise is also a nice surprise ;-)

Hit counts?

[Baorc]

Maybe if the website would take into account the hit counts it has estimating the number of people checking out the traffic and then approximately estimating (yes I realise the innacuracy in these words) the ammount of people taking the back roads and do the calculation again...just a thought. But as always, good luck predicting human behavior!

Cause you need to slow down!

[bluGill]

Wrong answer: when something like an accident happens you need to slow down. In fact anytime you see flashing lights (cop car, tow truck whatever) you better slow down until you know what is going on and have figured out how/if it affects you.

Granted most people look at the wrong thing, but at least they are slowing down. You shouldn't be watching the emergency itself. You need to have a broader focus of how the operations might affect you!

Re:Cause you need to slow down!

[AK Marc]

People slow down significantly when a crash is on the opposite side of a divided highway with a wall and 100+ ft. of space between them and the activity. There is no reason to slow in such a case, but people do.

No shit, Sherlock!

[Anonymous Coward]

And calculators can be used to correctly add two numbers together!

What the hell are traffic simulators used for besides predicting traffic jams?

Bash

[Anonymous Coward]

#!/bin/bash

case $(date +%H) in
(07|08|16|17) echo "Rush hour, take another route!" ;;
(*) echo "Roads are clear!" ;;
esac