Half a croissant, on a plate, with a sign in front of it saying '50c'
h a l f b a k e r y
Not so much a thought experiment as a single neuron misfire.

idea: add, search, annotate, link, view, overview, recent, by name, random

meta: news, help, about, links, report a problem

account: browse anonymously, or get an account and write.



Flowing traffic jams

Education in the art of leaving spaces
  (+7, -2)
(+7, -2)
  [vote for,

Many jams caused by sheer weight of traffic would not even come to a halt if people left bigger gaps between them and the car in front.

When the flowing traffic comes to a restriction and three lanes change to two you normally get everybody stopping before going through one at a time.

If everyone is educated in the art of keeping say a three car gap in front of them, the traffic could file in without even stopping.


spider, Dec 16 2000

Unjamming Traffic http://www.halfbake...Unjamming_20Traffic
Already Half-baked [blahginger, Dec 16 2000, last modified Oct 04 2004]

Highway cruiser of the future http://www.lowrider...um_straight01_z.jpg
I'd want a GPS unit in mine, of course. [eyeguy, Oct 04 2004]


       What he said. Duh.
thumbwax, Dec 18 2000

       That's the idea. People trying to change lanes without enough room can slow the traffic in their own lane as they try to wait for a slot or slow down traffic in the new lane as they cut in front of someone, who then has to slam on the brakes.
centauri, Dec 18 2000

       The more space there is between cars in a traffic jam, the longer the geographical length it takes up. If there is a constriction which allows through one car per second, the road before the constriction is two lanes wide, and two cars per second enter the road toward the constriction, then if there's a car per lane every 10.56 feet (bumper-to-bumper Yugos) then every 1000 seconds this goes on the backup will stretch for another mile. If the car's spacing was one per lane per 21.12 feet, then the backup would grow by a mile every 500 seconds. If they were spaced out one per lane per 31.68 feet as you suggest, the backup would grow by a mile every 333 seconds.   

       What matters at a traffic jam is the flow of traffic through the bottleneck. While other factors may affect the risk of accidents, or the obstruction of motorists who don't want to go through the bottleneck (e.g. if there's an exit 3 miles before the bottleneck but the back up extends out 4 miles, a motorist wanting to use that exit would have to wait a long time to reach it). Spacing the queued cars might reduce the risk of accidents, but it would increase the likelihood of the latter problem (motorists being delayed even though they don't want to go through the bottleneck).
supercat, Dec 19 2000

       Peter – yes, someone goes into the gap – that's the whole idea. Everyone lets one car in front of them. It's fair, it's the only way.   

       supercat - put your calculator down and get behind the wheel. The length of the jam is not the point, the fact that the traffic would remain flowing is the point. A jam is only a jam when it aint moving.
spider, Dec 19 2000

       Stop and go traffic is arguably more likely to result in other accidents or at least road rage which might exacerbate the jam.
centauri, Dec 19 2000

       Stop and go traffic also causes more wear and tear on vehicle components, more pollution (from engines and brakes both). It also requires more driver attention and certainly causes more accidents (a great many accidents occur in stop-and-go traffic when someone misses the "stop" part). These are not psychological advantages!   

       If the traffic jam is caused by simple congestion and not a fixed-rate bottleneck, continuous flow is more efficient than stop-and-go waves, both in terms throughput and in terms of space consumed. So supercat's objections are reversed; smoothing out the flow is a win for everyone involved.   

       Anyway, go read the links, we and others have discussed and analyzed this to death. If you think you understand the phenomenon, or are tempted to say "Duh" to anything here, you're almost certainly wrong.
egnor, Dec 20 2000

       Good link Waugsqueke, and well worth a read. I'd never realised traffic jams were so *complicated*. There must be an "-ology" name for the study of traffic jams.
Skinny Rob, Dec 20 2000

       It's better to travel constantly but slowly than 'stop-go' because then you have a rough idea of how long it will take you to get where you're going, for one. Also, when traffic stops, you don't know that the jam is not an accident. If it keeps flowing, you at least know you can persevere with the road you are on
spider, Jan 06 2001

       Traffic Jam Action-Driver Education?   

       Spiders correct, Police (UK) say use all lanes, where lanes merge. Well they do in road works where lanes disappear.   

       Drivers tend to feel that the bloke in the outer lane is cutting in, he is of course, he has to up the road, but if we used all lanes, and as a lane is merged, drivers followed advice to let ONE vehicle in, the one in front, the traffic would, surely, be more likely to flow? A car's in front if its even only partly ahead, ok? the gracious driver can even signal that he or she is giving way.That way drivers needn't hesitate, just FUNNEL-IN   

       Instead of the above scenario, traffic tends to stay in the inner lane, well it does in the UK, knowing that lanes merge ahead, or somethings happened out of sight up the road. Drivers' then try to keep the vehicle in the outer lane out. Reasoning that drivers who didn't move over into the inner lane were just trying to jump the queue.   

       These procedures, if valid, need to be added to a driving test requirement, in the form of a written test?
uked, Jan 06 2002

       A continuous slow flow of traffic is in some ways preferable to stop-and-go, but in other ways not. Vehicles with manual transmissions are often not well equipped to travel continuously at 3mph. Stop and go traffic isn't exactly great on a clutch, but it can be better than travelling so slow that the clutch must be allowed to slip to avoid killing the engine (letting the clutch slip continuously will wear it out quite quickly). Also, while stop and go traffic can cause more driver fatigue and aggravation than continuous slow traffic, slow traffic which seems to be moving smoothly can lull drivers into inattention. Given the way today's cars are made, a 5mph collision can do major damage (even a 3mph collision can cause significant damage if vehicles collide in the wrong way).   

       As for remarks about my putting down the calculator, if there is a constriction in the road which can only be passed by one car every two seconds, then as long a car is making it through every two seconds nothing behind that constriction will affect the total time for people to get through it. Anyone who drives in such a way as to get their vehicle through the constriction four seconds sooner than they otherwise would have done will either cause someone else's vehicle to wait four seconds longer or two vehicles to wait two extra seconds each.   

       While it's common for politicians to paint 'finite sum games' where they don't exist, in many ways getting through traffic constrictions is a finite sum game. The only way in which it isn't a fixed sum game is when either something changes the flow rate of the constriction itself, or when people exit someplace before the constriction itself (while someone who worked his way ahead of the traffic flow and then went through the constriction would slow down everyone he passed, someone who worked ahead through traffic and then exited before the constriction would not affect the travel time of those going through the constriction).
supercat, Jan 08 2002

       On a stretch of the A4 in West London's surburbs ,
there is a series of road juctions that can be negoitated,
without stopping as all the traffic lights,
are synchronised to change at the same delay time - a steady 40 miles an hour.

       Theres no point rushing up to a red light there,
. Trouble is you can't place a set of,
lights or other signals to control traffic where there's,
congestion (unless its at a spot thats regularly congested)
uked, Jan 09 2002

       Stop-and-go is also not preferable because of the delays between consecutive cars, causing that 'slinkie effect.' It's much more efficient to try and cruise along at the same speed.
RayfordSteele, Jan 10 2002

       <Anyone who drives in such a way as to get their vehicle through the constriction four seconds sooner than they otherwise would have done will either cause someone else's vehicle to wait four seconds longer or two vehicles to wait two extra seconds each.>   

       This is based on the assumption that the car ahead of Speedy is going to continue travelling at the same rate; Speedy will need to slow down to avoid rear-ending the slower car in front, and causing yet more problems. *however* if the driver in front sees Speedy's car speed up, then they may speed up as well, causing the next driver to speed up and so on. This won't go on forever, because people are going to take an average amount of time to look in their mirror and speed up, and over time, the effect will dillute going forward. However, the car behind Speedy will probably attempt to close the gap, thus creating a gap wave backwards as well, which will taper off eventually to an average speed determined by driving conditions, traffic density, etc....but for a short period of time, traffic will flow more quickly.
artgeco, Jan 05 2003

       I like your idea of using state troopers to smooth out traffic flow...maybe special teams could be put through a traffic flow class, and be sent out to basically do nothing but moderate the flow of cars. In dire situations, they could direct traffic, and would also be invaluable for providing advice to city planners on how, when, and where to do road improvements. These officers would KNOW TRAFFIC. I also think that a smoother ride would have lots of unintended effects...not only would there be less rubbernecking to slow things down to begin with, but stress in all categories would decrease. Suicides, alcoholism, etc. would probably see at least some drops. Public health would improve, as less fumes would be floating around, and less time would be spent being mad in a car.
artgeco, Jan 05 2003

       //Anyone who drives in such a way as to get their vehicle through the constriction four seconds sooner...>   

       This is based on the assumption that the car ahead of Speedy is going to continue travelling at the same rate; //   

       If one car can go through the constriction every two seconds, and if that fact isn't changed by Speedy's behavior, it doesn't matter what he does; my above observation will still hold (unless, e.g., Speedy causes an accident or creates another constriction which allows less than one car every two seconds).
supercat, Jan 05 2003

       It's common that after a "merge zone" the road is usually clear, so that after you manage to slowly merge, suddenly you can take off at high speed. In this situation the bottleneck is not caused by the fewer lanes ahead. Instead it is caused by the SLOOOOOOWWW merging. If traffic comes to a complete stop and people take turns, it can easily take 5 seconds or more for one car to pass (so each lane goes at under 2MPH.) However, if the cars had initially been going at 30mph with space between them, then the two lanes can "zipper" together without slowing down significantly. The "stop while taking turns mode" is easily 10x slower. Another way to picture it: it's the difference between a 4-way stopsign intersection versus a traffic light. A traffic light essentially forces huge spaces to appear in incoming streams of traffic. Remove a traffic light and replace it with a 4-way stop sign and the traffic flows WAY slower through that intersection. Taking turns sucks, it's incredibly slow,but unless people decide to leave space ahead of them so other cars can merge, it's the only "mode" that develops. (So if you should decide to bring a 20-car space into the traffic jam, it will momentarily function as a traffic light as opposed to a 4-way stop sign!)
wbeaty, Jan 20 2003

       Hey, I ran one of those Java traffic simulations and compared the flow of traffic with smooth flow versus traffic with an outbreak of "waves", of stop-and-go flow. I started things out with very dense but smooth flow. When the waves spontaneously arose the traffic flow suddenly decreased between 10% and 30% depending on the spacing of the "waves." It looks like traffic waves *DO* reduce the average flow. (So the question is... would traffic speed up again once someone smoothed out the waves? Or would it remain stuck in the 30% slower state even with the waves removed?)
wbeaty, Jan 20 2003

       One more thing to think about. Suppose traffic is sparse when approaching a merge zone, with huge spaces between the cars. In that case the two lanes can combine into one lane without any problem (I'm assuming that the spaces were so large that cars merging to become the single lane don't even need to tap their brake pedal.)   

       Now suppose a pedestrian crosses the road at the single lane so traffic backs up in the incoming two lanes. Now we have two solid-packed lanes of cars with more cars piling up behind, and the drivers start taking turns. They merge together incredibly slowly. In fact, they're so slow that incoming traffic piles up constantly behind the bottleneck, and the jam grows and grows.   

       See? This is a DYNAMIC bottlneck. The bottleneck didn't even exist until a temporary stoppage triggered it's formation.   

       THIS is where the finite-sum game stuff comes in. Merge zones are not a finite-sum game if they can function in "jammed mode" and "unjammed mode." A driver who brings a huge space into such a bottleneck can evaporate the bottleneck entirely. It's like paying one dollar to turn on a money pipe which doesn't turn off again; like creating a new resource from nothing, as opposed to fighting each other to win a fixed "prize." It's also a kind of "Stupidity of the Commons" game, where if nobody bothers to fix the situation, everyone suffers continuously, yet only a little effort could create vast improvement.   

       The "tragedy of the merge zone" is that we'd have to stop fighting: we'd have to give up our need to block interlopers and punish cheaters. I've talked to many people about merge-zone spacing, and the majority of them refuse to EVER bring spaces into the jam. After all, big spaces would give the cheaters free reign!
wbeaty, Jan 20 2003

       In Germany there is a road sign that reads "Reissverschluss" which means "Zip". The idea is that where two lanes merge into one or similar, each driver should let one car in in front of them. In theory this should keep the traffic flowing but as (nearly) all Germans are maniacs who are worried about being 2 feet late for work once they get behind the wheel this NEVER happens. I do not like driving here. Could you guess?
squeak, Jan 20 2003

       This problem is also exacerbated by the fact that it is a matter of national pride to drive as closely as is humanly possible to the car in front, flashing your lights and swerving out to indicate "I wish to overtake on this blind corner/brow of hill/double parked street" if the car in front is only doing 120 kph.
squeak, Jan 20 2003

       My tecnique for dealing with the propblem of people merging too soon:   

       I hate merging early because then there's always a bunch of jerks who go zipping by in the empty lane to cut in at the last second, slowing me down in the process. I also hate being a jerk since it's stressful to merge at the last second, and I know it makes others mad. ( I do this on occasion, especially when late, so I'm a jerk as well.)   

       My solution is often to be a second order jerk. I tend to be in the lane that is ending since it does generally travel faster, but as soon as I can see the end of the lane and verify that there aren't any exits on that lane that someone might need to use, I'll quit passing people in lane I will have to merge into, and just stay even with a gap (even a tiny one). Most of the time, since there aren't people cutting in slowing down the main lanes, traffic will have taken off again by the time I'm near the end of the lane, so the gap will have widened enough to merge. Even if the pace of traffic hasn't picked up, most people will open up the gap for me since I'm not being a normal jerk. (This may also have to do with the average politeness of the drivers where I live.)   

       Of course the people behind me who want to rush up to cut in front of everyone often seem a little upset. I've had people pass me on the shoulder several times, so this might not be a smart thing to do in an area with large road rage problems.   

       I suspect that this tactic encourages more people to stay in the lane that ends for a long time, reducing the speed difference, and giving less advantage to the jerks. I also theorize that it gives a boost to the speed of the cars going through the bottleneck since there is a period of time without anyone merging in. In any case, it is generally faster for me than merging early and less stressful than being a jerk...   

       So, am I an idiot (please explain), or should more people do this?
scad mientist, Jan 20 2003

       [scad mientist] it works for me - I'll try it next time I drive into London on a Monday morning.   

       I hate to go to the lowest common denominator here (check my other posts, I usually do!), but there's one problem with all the ideas put forward on this page: Joe Public as a rule is a blinkered maniacal egomaniac who will NOT let someone into the queue in front of them, because they want to get there first - a great example of this is the M1 in England - at any given time, once it starts to get busy you see about 70% of the traffic in the outside lane, even when the other lanes are free - this slows down traffic no end, causes road rage, and leads to stupid things like overtaking on the inside - "undertaking"... Having said that, the way the UK government is going, cars will be illegal in a couple of years anyway.   

       Sorry, did I get a bit passionate there? I do apologise.   

kmlabs, May 13 2004

       guys, just buy a bicycle
etherman, May 13 2004

       etherman the day i can get a 50 mile journey finished faster cycling than driving is the day i switch to a cycle as a main means of transport.
engineer1, May 13 2004

       Where am I going to find a $ 40 K bike to satisfy my need to show other people how much better I am than them ?
SystemAdmin, May 13 2004

       [SystemAdmin] I'm sure Lexus and Rolls would come up with an appropriate product strategy. I forsee a day of autobahns and Interstates filled with the "status" vehicles shown in the provided link.
eyeguy, May 13 2004


back: main index

business  computer  culture  fashion  food  halfbakery  home  other  product  public  science  sport  vehicle