The one thing you seemingly ignore is that when you slow to create a space in front of you, the car behind you has to slow as well. and the car behind that. Human nature and the lack of skil of most drivers means they typically notice late and over-react, slowing too much. This is how traffic waves start and amplifies with each car behind you. You even mention it in another context in your FAQ about how one fast driver can cause a traffic wave, but don't seem to realise your behaviour also causes it. Other cars behind you will change lanes to get around the idiot not keeping up with traffic flow, or just 'get out of the slow moving lane'. Again, that makes the traffic in the lane he moved into slow a little and a wave can potentially start there. You're just moving the problem around, not solving it like you seem to think.
Australia - Sunday, October 12, 2014 at 20:48:56 (PDT)
USA - Monday, July 21, 2014 at 23:18:37 (PDT)
I saw some of this 20 years ago while I was driving between home and college. I was taking circuits classes and realized that by leaving a gap ahead of me I was acting as a inductor and filtering out the high frequency components of traffic. If you graphed the density/speed of traffic waves, you would get a high frequency signal. Running that signal through an inductor allows the low frequency, direct current, constant speed signal through, but not the high high frequency / rapid velocity change signal. The distance between cars can be modeled as a potential.
Alan O. <firstname.lastname@example.org>
San Jose, CA USA - Wednesday, June 04, 2014 at 20:00:31 (PDT)
Ha! I thought i was the only one doing this, good to know i'm not alone. keep up the good work!
Amsterdam, NL - Monday, June 24, 2013 at 03:29:05 (PDT)
Been doing this ever since I was 16. I'm 30 ;)
USA - Sunday, June 23, 2013 at 22:12:57 (PDT)
I call it the "Accordion Effect" it's all about following to closely... no room for minor adjustments...
JOHN L AHERN <email@example.com>
SAN ANTONIO, TEXAS USA - Friday, June 14, 2013 at 22:30:55 (PDT)
Very interesting stuff.
On very congested motorways in the UK we actually have light (and camera!) controlled variable speed limits to perform the role of the rolling state troupers and it works well. I also think that your adopted driving style i.e. leaving bigger gaps so not needing to break so much reduces not only your vehicle's fuel consumption but that of the other vehicles around you. Difficult to prove I guess...
Wayland Smithers <Smithers_Wayland at Yahoo.co.uk>
Notts, UK - Thursday, January 03, 2013 at 11:42:09 (PST)
"If you want to be punctual in the office you have to arrive late to the jam."
Hello, congratulations for your great work. I don't believe in phantom traffic jams, these jams are due to the fact that nobody to taught drivers to stabilize the flow of traffic.
The driver needs two distances: security and damping.
I would like to introduce the wavedriving technique. Please,visit http://www.youtube.com/watch?v=uM2EfPbFSCk.
MADRID, MADRID SPAIN - Sunday, December 16, 2012 at 08:17:13 (PST)
Interesting that you never mentioned keep right except to pass. [ It's in the FAQ. Also, that only works in light traffic, while my stuff all deals with all-lanes congestion. -billb ]
I still think this is the best rule to follow when driving in our highways. The key is to always have an open lane in your left for passing. Ideally if all drivers follow this rule, there's always going to be drivable lane/s on your left. The key is for all drivers to follow this rule all the time...Then we won't need to build express lanes, carpool lanes, etc...Also good for emergency vehicles...
Tracy, CA USA - Tuesday, October 09, 2012 at 22:31:08 (PDT)
I suggest a credo for today?s need to drive responsibly and to minimize high density traffic wave congestion:
?Every driver is responsible for the immediate driver behind them, keeping them safe and avoiding the need for sudden braking?
How This Rule Works In traffic:
1) The Rule Is Expansive and All Inclusive
If we followed this general rule, we would all naturally be responsible for leaving the proper space/time ahead of us. Space clearly gives us reasonable reaction time to perhaps "brake by simply coasting" and not forcing the driver behind us to suddenly have to brake hard. It assures that braking is something one does as a "last resort" and only when no other circumstance for safety and/or avoiding a collision is available.
2) Allows Trusting The Front Driver and Eliminates Reactive Braking
Keep in mind what normally happens to traffic when the first driver uses their brakes either idly or in protective response to traffic. We are conditioned to then over brake for two reasons:
a) Less Reaction Time - We have less time than the front driver to react and consequently need to over compensate by braking harder. This affect gets compounded, causing each upstream driver to brake harder until a collision or multiple collisions occur.
b) Don?t Trust Other Drivers - We are conditioned not to trust our fellow drivers because they haven?t yet been educated to abide by our Rule. This lack of trust causes us to be "overly" cautious and results in the over use of our brakes.
3) The Rule and Its Implied Driving Etiquette Are Required Learning During Driver Education
The rule and its derivatives must be taught as part of our driver?s education program, with their aspects tested as part of the licensing exam. Also included must be the responsibility to adhere to a driver?s etiquette code, and not just stop at obeying basic traffic laws as was done in former DMV years. Highways are becoming too congested to allow rouge drivers to ignore their social driving responsibility and road etiquette. Tickets for abusing social driving etiquette should be as prolific and costly as our traffic laws of speeding or not stopping at a stop sign, etc..
4) The Rule is a Prelude to Using Auto-Controlled Vehicles on Congested Roadways
Just as extended use of cruise control will incrementally help future traffic ?wave congestion?, this Rule and its implications of driving etiquette will take us to the next step forward. Of course our final step forward with impending technology will eliminate drivers form the equation and rely only on auto-controlled vehicles in regions of high density traffic. Ideally this will eliminate accidents due to all the ramifications of driver shortcomings , and only allow the possibility of technological glitches to become problematic.
Herman Vogel <firstname.lastname@example.org>
Danbury, CT USA - Friday, September 28, 2012 at 08:47:47 (PDT)
No way, YOU are the cause of traffic. The gaps you are creating and the brakes lights and swerving of other cars to get around you are exactly what is slowing down the people miles and miles behind you.
[Nope, not happening. This backup existed before I arrived. Notice that I'm driving at the same speed as the guy ahead, not "slow." Nobody swerves around me to pass (need second camera facing backwards to show this.) All those merging cars are coming in from a distance. These merging cars are drivers who missed their chance to get in line for the exit lane far back, and now I'm letting them merge at speed. Without my gap, they'd go down to the end and force their way in, same as the big truck in the first seconds of the video. See FAQ #8. And if you want to comment on a video, use youtube so everyone can compare the actual video with any comments. -billb]
People need to know their lane and maintain a consistent gap with car behind them. Otherwise, human nature will draw other drives to that open space causing the cars around them to slow down to allow the car to cut into that open area. Then you get a domino effect.
[No, you get a domino effect if you decide to sit in this exit lane and try blocking all merging drivers. Blocking them doesn't even work, it just makes everything grind to a halt. Merging drivers see the solid wall of aggressive tailgaters in the exit lane, so they just drift on down to the actual exit and force their way in as shown in the vid. No gear-teeth merging. No high speed zipper, instead it's people fighting for no reason at a merge zone, illegally blocking cars coming from another highway, while crawling at a few MPH and causing a huge backup. But at this location, sometimes one driver can unplug the whole thing. A big space triggers the flipflop to "un-jammed mode." The giant clot dissolves in front of you while you watch. (Didn't happen while I was filming though.) -billb]
Traffic is caused by oversaturation. If you add in giant constipated turds to the mix slowing down the fast lane, angering other driver, and creating large smell gas bubbles between your "waves", it makes the experience worse for everyone.
USA - Friday, March 23, 2012 at 10:28:34 (PDT)
Interesting read! I always wondered why people are tempted to fill up the empty space I leave ahead of me and when I tried to get the answer, found trafficwaves!
USA - Wednesday, September 21, 2011 at 17:38:59 (PDT)
Traffic in some ways resembles a market, with each driver conveying information about conditions ahead to drivers behind. Good drivers get rewarded with better gas mileage.
Steve Cobb <email@example.com>
Nashua, NH USA - Monday, July 04, 2011 at 05:02:22 (PDT)
I came across this same theory about ten years ago. I thought I was the only one. It's great to see that this actually works for others as well. Thanks for getting the word out. I wish more people would understand this practice. I used to drive for an airport shuttle service in the Milwaukee area. My passengers would often ask what I was doing and then would be amazed at how well it works. I had one regular passenger that was a skeptic until I showed him both approaches to jams on four separate runs. The time difference was negligible.
Tony F. Wisconsin
Tony Freiwald <firstname.lastname@example.org&*()&>
Manitowoc , wi USA - Wednesday, June 29, 2011 at 14:13:01 (PDT)
I've only only seen the video, which got me thinking. Wouldn't it be great if there were an occupation created just to eliminate traffic. Deploying traffic unjamming drivers to stabilize traffic?
[This is happening! Traffic in Belgium and the Netherlands is the worst on earth, and those countries have instituted the Blokrijden program. These are brightly-marked cars driven by wave-smoothing drivers, and it's illegal to pass them. Long ago people fantasized that if we all had robotic cars, the cars could organize themselves into high-speed "cohorts" which move like freight trains on the highways. But it turns out that all we really need is one single "train engine" at the front of the line, with a smart human driver. -billb]
toronto, ontario canada - Sunday, March 13, 2011 at 18:16:54 (PDT)
If on the Interstate, not only would I watch the brake lights of the car in front of me but the car in front of them. I might not have to brake b/c the reaction time of the car in front of me.
Brakes should get brighter the harder you push the brake pedal.
Nashville, TN USA - Thursday, March 10, 2011 at 03:50:13 (PST)
I love this. I've been trying for years to do my bit to combat road rage. I try never to get off the road without having done one nice thing for somebody, whether it is letting a driver merge, stopping for pedestrians to cross at unmarked intersections, or whatever. I had no idea that just slowing down and leaving room for folks to move (which is something I've always done) could have such an effect on traffic. Thanks.
CA USA - Wednesday, March 09, 2011 at 02:59:32 (PST)
Okay, I've been trying to hypermile in Jakarta and I can't succeed 100% in coasting because there are so many jams, so many cars, and small lanes, and I think I noticed that going hypermiling makes the driver behind me kinda mad (horning or flashing headlights), because I'm driving slower than normal cars (especially when I was coasting a few meters from a red light) and the lanes are mostly tight. So it is a harder challenge for me to drive, since normal speed on the road is about 30-40 km/h in jams (fastest is 60km/h on a straight lane). I kept at 40 when everyone was 60, but it's quite hard to keep at 40 when everyone is slow. What do you say about this? How should I use your hypermiling technique when speed is usually low? Because when speed is slower, everyone tend to get stressed and try to be the first by cutting others' lanes.
But of course, in one jam stream, I noticed that when I hypermile, there are only TWO cars cutting my road, a smaller number than my expectation (about 7-10). So I should say hypermiling definitely worked in increasing my MPG and the MPGs of the people behind me, and getting cut by one, two, or even three cars doesn't make a valuable time difference in getting to the destination. And I've tried to abandon hypermiling at one time, because I am really, really, really late but you know what? The stress and tension caused me to scratch my car twice! So even though we're late, it isn't worth to burn more fuel and cause accidents (= extra bucks to cover). From now on I'll hypermile, no matter what. Time difference is about 5 minutes anyway.
I have recently read about extreme hypermiling who risk their lives for more MPG, like tailgating a big rig, death-turning (turning without brakes), turning the engine off to coast, or removing additional weights such as extra tires and emergency tools. To me, THAT is not a good hypermiling technique. It's not advanced or expert; it's dangerous. It's almost just as good as removing the brake pads and discs, wipers, airbags, seatbelts, chairs, and replacing the steel frame with a recycled plastic one to totally lighten the car's weight.
Hypermiling is good, but not too much to risk one's life or disturb someone else's comfort. I hypermile, but if someone else is in my car, I'll let them have the AC and stereo comfort. It's much better than arguing with someone else in the backseat or with other drivers over who gets the lane. Trust me, hypermiling in the right time, and not too overly done, keeps you young.
Joshua Tan <JRTTS@@yahoo.com>
Jakarta, Indonesia - Tuesday, March 08, 2011 at 19:12:46 (PST)
How about a campaign to get this information into the DMV HANDBOOKS! You'd think it would be backed by ECO-Friendly to save Brakes and Clutches(Asbestos), and Gas.
Seattle, WA USA - Monday, February 21, 2011 at 15:41:20 (PST)
I call this cool. I've searched for how to save fuel, and your site comes as one of them in the list. I saw your amazing physics lesson created out of boredom, and I think, "Wow, this is a must-try" so I did. I tried on a road with no jams. When I saw a green light far away, I think it's turning red, and when it did, I stopped accelerating. It was a soothing ride, and everyone slept. But I had the feeling (and my dad also said) that people behind me must be mad at me slowing down waaay too early. Of course, they would usually pass my car and get caught in the same red light. But sometimes it was not how people would pass my car that worries me, but what they would think of me spoiling my time. What do you say about this? Oh, by the way, I hope I can post a video of my hypermile trip to compare to yours and to see if I got it right.
Thank you for the great site, and also the animations. Even though I also noticed this strange stop-and-go circumstances in jams rather than fluid flowing, as you say it, I have always thought that it's unchangeable. But you have opened my eyes.
Joshua Tan <JRTTS@yahoo.com >
Jakarta, Indonesia - Thursday, February 03, 2011 at 03:48:42 (PST)
This is a great site - and validates what I have practiced for years. I used to commute about 20 miles on the Washington, DC Beltway (I-495), which has horrible congestion. Once you open up a space, you create room for the traffic to "breathe". I totally agree with the advice on this site.
Now, I often ride a motorcyle in heavy traffic. Because of the manual transmission, and for safety reasons, I always leave a space in front of me. I always have a nice smooth ride with very little stopping. Ask any motorcyclist - we hate putting our feet down!
Chester Arthur <email@example.com>
Landenberg, PA USA - Sunday, January 16, 2011 at 12:49:47 (PST)
Brilliant! I'm commuting here in Germany on slightly smaller scale compared to your situation in Seattle. In the morning traffic congestion is happening nearly everyday. Also the rules on german highways (Autobahn) are slightly different, your observations apply here in Europe as well.
Avoiding to use the brakes whenever possible is very good rule of thumb for starters. Saves a lot of fuel as well. Thanks a lot and keep up that good work.
Peter Funk <firstname.lastname@example.org>
Ganderkesee, Germany - Tuesday, November 09, 2010 at 12:50:00 (PST)
Love your analysis. Some additional pointers, that I did not immediately see: Anticipation is key. It isn't enough to slow down early, you need to speed up early as well; yet another reason to leave that big gap in front of you. Watch the cars in front of the car immediately in front of you, and as much as possible:
1. Start slowing down *before* the car immediately in front of you starts slowing down. Stay off your brakes if at all possible--brake lights tend to cause a chain reaction of people applying brakes, which is another cause of waves.
2. Start speeding up *before* the car immediately in front of you begins speeding up.
Both of these actions require leaving a large gap between you and the car in front of you, and watching the traffic ahead of the car in front of you. If you're stuck behind a semi, it isn't always possible to see what the traffic in front of him is doing, but I've noticed that semis tend to be pretty good about leaving gaps in front of them.
Seattle, wa USA - Saturday, October 30, 2010 at 15:05:16 (PDT)
You are correct in all of your observations. Large spaces dissipate the waves and allow for safe merging.
I just wish other drives would drive this way in heavy traffic situations and life would be a little nicer. PSA's could help educate drivers to the large space in front and the left lane is for passing only on two lane highways...
Hyram Williams <email@example.com>
Chattanooga, TN USA - Wednesday, September 29, 2010 at 22:03:36 (PDT)
I know this theory works. I used it yesterday, on Labor Day on Interstate 90 West. Set my cruise control for 64 mph. and stayed in the right lane. About 80 percent of the traffic passed me, and there was always smooth flow ahead of me, going 65 mph or better with longer gaps. No long bunch-ups behind me either. On 90 Eastbound, on the other hand, there was nothig but stop-and-go traffic from Wisconsin Dells to Madison. Even the sections of rolling traffic were only rolling about 35 mph. I could see angry faces on 90 East. This strategy saves gas, saves a lot of brake pads. Eliminates road rage. You get home faster. Everybody wins.
La Crosse, WI USA - Wednesday, September 08, 2010 at 10:20:53 (PDT)
Loving the site - I brave Seattle traffic on 520 twice a day and have yet to learn patience with the terrible driving habits of Seattlelites and the USes unreasonably low speed limits.
That said, it's not only driving behavior that impacts traffic waves. There are certain locations in Seattle that have standing waves whenever traffic picks up a little (think rush hour). Typically, it's at points that have even a slight inclination.
My only explanation is that it's a consequence of the numbers of automatic cars and their trouble of keeping speed constant when faced with an uphill slope. Cars slow down, instant traffic jam.
It's particularly noticeable on the two elevated points where the 520 bridge connects to land.
Elimination of small bumps might go a long way towards improving traffic flow.
Seattle, WA USA - Monday, September 06, 2010 at 08:29:06 (PDT)
I'll say right up front that I probably fall into your category of "aggressive driver," but that doesn't mean I don't analyze traffic patterns effectively or that I'm not a good driver. I do a TON of traffic pattern analysis on my daily 1.5 hour each way drive, and even though I regularly cruise comfortably at 85mph, I'm a very safe and courteous driver, or at least I try to be. <more>
Bob <bpitas at gmail >
Taunton, MA USA - Thursday, August 19, 2010 at 09:35:22 (PDT)
Hi, I'm a long haul trucker for eight years now. I drive the roads for 11 hours a day and I too find it interesting to study traffic behavior. You seem to have got it right in most areas but there is one significant factor that you haven't addressed.
I call this "cheat merging". Cheat merging occurs when a vehicle moves from the thru lane into the dropped lane because the dropped lane moves faster and then merges back into the Thru lane at the point of constriction, in effect passing the congestion. This cheat merging adds to the total number of vehicles that must merge into the thru lane which greatly aggravates the congestion.
Sometimes cheat mergers use on ramps off ramps and even the shoulder as a cheat lane. Truckers are very familiar with this behavior and will sometimes stay in the dropped lane without passing the through lane. When the dropped lane moves at the same speed as the thru lane, then zipper merging works very well. When cheat merging becomes excessive, the thru lane comes to a halt. It's like we learned in kindergarten: When too many children butt in the front of a line or que, the line stops. I have found this to be a major aggravator to congestion. I hope you will consider this as a factor in your study.
Charles Brown <firstname.lastname@example.org>
Ogden, UT USA - Sunday, August 15, 2010 at 20:10:59 (PDT)
Great site! I've been doing the average speed technique for about a year now ,after reading an article about hypermiling and saving fuel. Just the other day I was stuck in rush hour traffic so I did what i always do. dropped the auto tranny into 2 and coasted. After watching the brakelights of every car in front of me flash on and off every 5 seconds, i realised what was doing to save fuel was also smoothing out the traffic behind me.
New Zealand - Thursday, August 12, 2010 at 07:40:29 (PDT)
A fantastic insightful read, and absolutely true as far as I'm concerned. I only wonder why this doesn't come naturally for most people.
I thought you might find this interesting (taken from http://kiwianarama.posterous.com/why-i-hate-bad-driving-in-new-zealand ):
On the Autobahns of Germany, drivers are given the freedom to travel as fast as their cars will allow. However, when there are roadworks, or lane reductions, speeds will be incrementally reduced in 20Kph chunks over a distance of several kilometres. Watching the many cars on an Autobahn slow down together, stay in lane and drive in even, tight formation at each speed reduction, thereby avoiding congestion, is almost hypnotic.
Auckland, New Zealand - Monday, August 09, 2010 at 21:23:04 (PDT)
This site is totally awesome! Very well laid out and the concepts are explained thoroughly.
I am from Southern California and I have been trying to tell my Colorado friends about the concept of merging speed and spacing on the freeways. At my high school in California, it was a requirement to take drivers ed. So for the past 21 years, I have been driving in accordance to the standards of the driver's manual.
However, I became somewhat lax on my driving standards since moving to Colorado in 1996. This is due to the fact that the freeways are called highways out here because most of them are only 2-3 lanes each way. Ok, so maybe that is not a fact. But I have become accustomed of not leaving that extra room for a car to jump out in front of me. Mostly out of fear that when cars make a mad dash to jump in front of you (may I add without their blinker on) they do so when traffic in front of you is coming to a complete stop. Since I am already slowing down, because I am always watching 2 cars ahead of me anyways, I am forced to stop shorter than what was calculated.
Another pet peeve of mine is when drivers don't understand the concept of every other car when merging on to the freeway, and to have that merging speed as well. Unfortunately some of the freeways in Colorado have a minimum to maximum driving speed of 45-65 depending on which part of the freeway you are on. And even in some cases, the speed limit can change from 55-60-65 and vice versa on the same freeway within a 10 mile stretch. A good way for the highway patrols to write more tickets if you are not familiar with which sections begin a reduction in the speed. But back to the merging speed issue. People don't realize that if this is not effectively executed, there will be a traffic jam as shown in your diagrams. It could be 10pm and there would still be a traffic jam due to "invisible accidents" probably due to drivers not being able to properly space their vehicle while driving on the freeway.
Just today (7/23/10), I heard on the radio that the Colorado State Patrol troopers and Denver police are using smart radar to ticket tailgaters. One lady called in to complain that she got ticketed after merging on to the freeway. How you cannot tailgate while merging onto the freeway during morning rush hour traffic is beyond my comprehension. But I understand the reasons behind the use of radar. According to a Sargent of the Denver police, "Ninety percent of all the accidents that my unit handles on the interstates are following too closely", (http://www.thedenverchannel.com/news/24362759/detail.html).
I'm glad I came across trafficwaves.org during my research into why the big push to ticket tailgaters. I wanted to make sure I was fully informed as to how to avoid a big fat $130 ticket!
Aurora, CO USA - Friday, July 23, 2010 at 20:03:34 (PDT)
i dey here keep off
ow, USA - Monday, July 12, 2010 at 01:47:59 (PDT)
I have been driving like this for years and have been thinking about ways to educate other drivers to do the same. Can we start a group, a web site, a company, to promote this way of driving with the goal of improving the flow of traffic throughout the USA?
We could sell t-shirts, hats, and bumper stickers to raise money and invest the money on building awareness with billboards and advertisements. We could work with governement agencies like CAL Trans in california to put up signage about merge areas, etc.
One key is spacing in the far right merge lanes where motorists are entering and exiting the freeway.
Bumper sticker and t-shirts with the logo of the word "traffic" and a circle slash. Sayings could include; my car is a "space ship", Space don't race, start "spacing and stop braking!, Don't give me a brake.
It is so obvious that "braking" causes slow downs. If you reduce the Braking and the wave that follows, you will reduce the traffic.
Call me to discuss,
fremont, ca USA - Monday, July 05, 2010 at 17:00:38 (PDT)
I've driven in ten countries. When I read the two loop summary scenerios I can't help but think: one is American, one is German. In Germany, no one hogs the far left autobahn space-- EVER. Pass and move. If someone needs to go around you, it's not personal. It's taboo, illegal, and deadly to pass on the right-- I've never seen it done. And the zipper technique is law. You MUST let one person merge. Everyone lets one person merge. So yes, the autobahn has no speed limit, though it is well-documented to be safer and (believe me first hand, here) much more efficient. France was the worst, it felt so slow. However (and outside of Paris/major cities) it still moved. This is fascinating to me because our driving aside, Americans are famous for our laidback culture. Maybe the etiquette should be bumped up just a tad for stricter driving laws. Traffic jams are an inconvenience, true, but furthermore-- they're dangerous. Great observations on your article, I'm definitely sharing with other military friends. :)
El Paso, TX USA - Wednesday, June 02, 2010 at 04:31:27 (PDT)
this site intrigued me, so i ran some numbers. it gets interesting. enlightened self interest should keep lane changing to a minimum when a section of highway is at capacity.
imagine a section of two lane highway at capacity, both lanes running at 60 mph. no entrances or exits. following space, S in seconds, and vehicle length L in feet.
vehicles per mile=
for a 2 second following distance and 15 foot vehicles, the vehicles per mile is 27.64.
now let enough turkeys(might not be one) change lanes to increase the left lane to 28.64 VPM and the right lane to 26.64 VPM, while keeping following distance and vehicle length the same.
the result is that the left lane slows to 57.73 MPH while the right lane speeds up to 62.44 MPH. changing lanes is the absolutely dumbest move you can make unless you have sky-cam capabilities. all you will do is slow yourself down.
USA - Sunday, May 23, 2010 at 12:24:06 (PDT)
an orbital view of slowdowns is essential for the thing to work properly. if traffic is stop and going at a net speed of 30mph, a wall of gappers moving at 30 mph will clear out the stop part. this may not improve overall traffic capacity.
however, a steady flow without the stops provides an opportunity for the flow to increase. anybody who has ever evaporated a wave with a gap and has a good rear view, can see the difference.
USA - Sunday, May 23, 2010 at 03:27:48 (PDT)
On some of our badly jammed motorways in the UK there are "Variable speed" traffic management systems which aim to solve the problem by doing exactly what you describe.
When in traffic I often find myself trying to maintain the average speed of the traffic around me. It's pretty irritating having to repeatedly get up to speed then stop. We're also taught to drive leaving a 2 second gap between us and whatever's in front. This is to prevent accidents, but I suspect it's also to try and prevent traffic jams.
Do you know about a website called "Waze" (http://www.waze.com). If everybody used their service, we'd all know where the jams were and could take appropriate action.
Can you now solve the problem of badly sequenced traffic lights in cities that seem timed to always make drivers reach red lights?
Wakefield, X UK - Sunday, May 23, 2010 at 01:05:16 (PDT)
Great site! I take the train to work and I've noticed that this works really well to get escalators moving again when they're blocked by that one person that just stands there rather than walking up. If everyone rushes up and stands behind the stationary guy, the escalator soon becomes a 'stand and wait' ride. But I've noticed that if I slow down when walking up the escalator to the point that the person in front of me is off before I catch up, then everyone behind me can keep walking up too. (I assume this works, I've never stuck around to find out:)
Sydney, NSW Australia - Wednesday, May 12, 2010 at 04:03:37 (PDT)
I salute you for having this site! I discovered this same method shortly after I moved to LA, and it's still AMAZING how one person can affect traffic not just in his lane, but in every lane around him.
You're exactly right, too, that in order to do this, you've got to let go of the winner/loser model in your head, and don't even fret that somebody else might not be "playing by the rules". Because, in essence, you will become the Pied Piper of the roadway, and other cars will start doing what you do, seeing how well it's working for you!
I think this stuff should be taught to every child in the country, and PSA's should be blanketing the airwaves, showing everyone the "way". ;-)
Thanks again for a great website.
Los Angeles, CA USA - Tuesday, May 11, 2010 at 16:20:20 (PDT)
The most important assertion you are making is that total throughput will be increased using your technique. That may very well be true, in which case I hope more people follow your advice.
However, I just wanted to point out that if throughput is equal, and I have to choose between alternating high-speed/zero-speed stretches and a constant 35mph on the highway, I'll take the stop and go anytime. It may be irrational, but the frustration that builds up when I feel I'm trapped behind slowpokes doing 20 miles an hour under the speed limit is undeniable and just can't be good for my health. Stop-and-go traffic doesn't bother me much because I at least feel like I'm going as fast as I can, and I don't worry about trying to pass a slowpoke because I can see there is nowhere to go.
I just wanted to point out that from a mental health point of view, not everyone wants a slow-and-steady 35 or 45mph. the mathematics of your argument may very well be sound...I'm just saying that what you are describing is the total opposite of what I would consider a relaxing highway ride for me. I'd go nuts! But then, my name is Speedy McSpeederson, so I might be a special case.
San Leandro, CA USA - Tuesday, May 11, 2010 at 14:25:05 (PDT)
Love the site, you should make bumper stickers "Don't like my driving? Then visit http://trafficwaves.org/ and realise you're the tool." :)
New Zealand - Tuesday, April 20, 2010 at 21:59:13 (PDT)
Nice stuff. I have notes on this subject (and many similar) I've compiled for my next book. The idea is to give people tools and knowledge to help them help traffic flow.
If you're looking for an additional perspective on traffic flow, I devoted a chapter of my first book "Ride Hard, Ride Smart" to it. Entitled "Reading the Road" (pp 118-129) it's meant to give motorcycle riders an edge in predicting traffic flow to maintain a safety cushion. The crux is I compare the flow of traffic to the flow of water in a river. Never have I had so much fun with the concept of deviants!
Pat Hahn <email@example.com>
St Paul, MN USA - Tuesday, March 23, 2010 at 11:42:02 (PDT)
Over here in the UK we know all about sitting in traffic jams. Virtually no motorway journey goes without a speed drop to lower than 30mph often with no visible reason. But anyway...
There is one aspect of a high traffic density, which is where do you fit the traffic merging onto a multi-lane highway that is already at full capacity with everyone driving nicely 2 second spaced?
If the nearside lane is spaced at 2 second gaps and the on ramp supplies a vehicle to calmly zip into each 2 second gap, we now have a nearside alne that is too closely packed. The other lanes are all at full safe, sensible denity,so how do these new vehicles get accommodated?
The vehicles have to slow down to reestablish the 2 second gaps. This then sends an accelerating bracking pulse back along the highway, eventually stopping the traffic well before the on ramp.
If the road is at full capacity this can not be avoided.
Even if you crerated a 1/2 mile space infront of yourself on the approach to the on ramp, this would quickly get filled by the traffic joining from the on ramp, and then you have a full road approaching the nest on ramp. Unavoidable British pessimism/reality.
If the traffic leaving at a junction is less than joining at this same junction you have a jam that cannot be avoided.
Simply put:- If the number of cars wanting to use the road is more than the road can handle if it were a carpark with everyone stationary, then you have the M25 motorway most rush hours.
But smooth driving with a reduction in aggression and quick moves will hold off the formation of a jam until the carpark density gets close.
j plant <firstname.lastname@example.org>
Bristol, UK - Sunday, March 21, 2010 at 12:53:31 (PDT)
Cool site! I have thought about this traffic jam busting behaviour and wondered if it would work! It does!
Note also the Green Light Surge, the light turns Green, everyone surges ahead then comes to a complete stop again, and then it's so slow to get going again, as some one surged then stomped the brakes hard and stalled.
There is a four lane road on my commute, buses, bus stops and traffic lights the whole gambit. I try staying in the bus lane going just below the speed limit and watch all of the others changing lanes to avoid the busses and slowly creeping to the rear as they panic brake and accelerate. I often cruise through and gain on a half dozen cars.
My favorite was a fellow who followed me for 15 kilometers onto and off of a freeway, changing lanes 26 times and panic braking who knows how many times. He started out behind me and ended up right in front of me.
I am glad to know I am not the only one who has seen this behaviour!
Dean Chesterman <dean(DOT)chesterman@shaw(DOT)ca>
Edmonton, AB CANADA - Tuesday, March 09, 2010 at 10:02:39 (PST)
Just found your site. I thought I was the only one out there doing this, now I don't feel so crazy. I drive the I-5 in the notorious Orange County "Crush" area - "brilliant" CalTrans planning.
The only tough part about Zen-driving are the super-angry folks directly behind you. I had a Mayflower big-rig last night that followed me for several miles (in the rain) with its high beams on because of the "huge" 60-80ft gap in front of me.
This big rig had to have been no more than 5 feet behind me at all times! The ironic part was the complete eating/elimination of traffic waves due to Zen-driving was allowing him to keep perfect pace with me. Maybe I'll get an apology in the after-life.
Anyhow one place I've used this was the I-5 south carpool lane where the 57 south carpool interchange merges in. The I-5 folks close ranks and the 57 folks have to fight to get in. Once while I was driving this on the I-5, I left a huuuuuge gap (very difficult to do) which let the entire backlog of 57 folks in - this resulted in a massive improvement in traffic following.
Fullerton, CA USA - Saturday, February 06, 2010 at 10:28:27 (PST)
After viewing the latest video, I can't help but wonder if local drivers have more effect than visiting drivers as the locals tend to know where trouble spots form?
Perhaps innovative signage will develop, warning or instructing drivers. Maybe someday, we'll see "traffic coaches" broadcasting jam busting instructions on roadside boards?
Come to think of it, my GPS has some feature to receive traffic jam reports, but I think it requires more hardware...
Pete Brown <''email@example.com''>
ottawa, on canada - Sunday, January 31, 2010 at 08:44:02 (PST)
Hi Bill - this is a wondeful site with great understandable explantions. I totally agree with your analysis. Here is a thought about why maintaining a constant gap ahead when you slow to enter a congested area works. I suspect drivers instinctively maintain a gap that keeps the time to traverse it ("headway") constant, so as you slow down it gets shorter. The size of the gap another car needs to change into your lane also gets smaller as you go slower, but it prevents lane change well before it reaches the limit of the following car. This is probably because of the physical size of a car not going to 0 (a fixed size), but also that an entering car needs space to brake (a variable size) as traffic slows. ie, the curves cross at a point that determines just how close a following car can get and still permit another car to chenge lanes and enter at that speed. Once traffic reaches the point where the distance between cars is less than the fixed and variable requirement of the entering car, the pattern quickly goes from synchronized to jammed. Does that make sense?
Cambridge, MA USA - Wednesday, December 30, 2009 at 12:32:09 (PST)
Being the laid-back person I am, I generally try to leave gaps, especially when I drive alone and am not in a hurry to get somewhere. But, usually, when others are in the vehicle with me, I have to deal with the pressure of "teaching the other reckless drivers a lesson" and not letting them merge in ahead of me, and that I should not be the one who is always the friendly driver; and besides, why should I be "late" by allowing others to "sneak in". I've been contributing in "wave erasing" for some time, but didn't know what to call it. Thanks for the extensive evaluation and insight!
I'll keep trying; and now I can explain to others what I'm doing...it's my civic duty as an interstate user, and I'll feel good about it!
MIKE ALLIGOOD <firstname.lastname@example.org_a_>
lawrenceville, ga USA - Monday, December 14, 2009 at 11:12:28 (PST)
Thank you for spreading this idea because not only does it help alleviate jams but it also improves the traffic efficiency as well.
My Dad used to get the New York Academy of Sciences journal and an article in there written in the 60's (!) has always stuck with me. They studied traffic going through the Holland tunnel. To break up jams they did what you do, and sent in slower pilot cars to open up spaces in the traffic flow, and they did this every few minutes. Cars traveled through the tunnel in flotillas lead by these pilot cars who maintained significant space ahead of them to act as shock absorbers which absorbed the speed fluctuations.
They found that not only did they almost eliminate traffic jams in the tunnel, but the number of cars getting through the tunnel per hour actually went up.
So not only do you help make the drive a more pleasant experience, you also help everyone get home a bit sooner.
The Stanley Park causeway here in Vancouver is an ideal place to practice this as it is often only one lane, and I enjoy timing my driving to try and eliminate the holdups. I find that it is possible to get the whole road moving steadily again.
jan d G <ruthandjanattheusualMSemailsitehotmaildotcom>
Vancouver, BC Canada - Thursday, November 26, 2009 at 23:28:22 (PST)
I posted your 'Traffic Waves' video on my blog this morning just ahead of the Thanksgiving travel rush.
Not sure if the massive L.A. holiday traffic can be affected by just one driver, but if the blogosphere would spread the word around holiday times, surely there would be more than one person using the method.
Whittier, CA USA - Wednesday, November 25, 2009 at 12:14:07 (PST)
Interesting, I too have been driving using this technique for at least 20 years. I believe that my (also independent) discovery is in part due to spending way too much time commuting.
One factor that has not been widely commented upon is what I think of as "a reduction in the perception of general aggressiveness on the part other drivers" when I am driving using this technique. This calming effect seems to reduce the amount of "jockeying for position" and guarding of their front end cushion that tends to cause a general reduction in the spacing between cars, which in turn means better traffic flow.
I did a little searching and came up with this very complete work on traffic flow theory http://www.tfhrc.gov/its/tft/tft.htm , that is well worth looking at if you're interested in the finer points of the traffic engineers domain.
John Trask <email@example.com my email>
Thousand Oaks, CA USA - Wednesday, November 18, 2009 at 09:53:04 (PST)
THE CITY HAS PUT UP A NEW TRAFFIC LIGHT ON SAN PEDRO DR AND PASEO DEL NORTE. THAT HAS GOT TO BE THE WORST IDEA EVER,ALL ITS GOING TO DO IS MESS UP TRAFFIC MORE.YOU PLANNERS NEED TO UNDERSTAND THAT MORE LIGHTS DOESN'T MAKE IT SAFE,JUST CAUSES MORE PROBLEMS.PEOPLE GET UPSET WHEN THEY HAVE TO STOP EVERY BLOCK OR INTERSECTION AT A DAMN LIGHT AND IT CAUSES ROAD RAGE AND ACCIDENTS.WHY DON'T WE JUST MAKE IT HARDER TO GET A DRIVERS LICENSE AND TEACH THESE STUPID PEOPLE HOW TO DRIVE, DON'T GIVE EVERYONE A DRIVERS LICENSE JUST CAUSE THEY APPLIED FOR ONE.SO GET RID OF SOME OF THE TRAFFIC LIGHTS ON ROADS THAT ARE NOT MAIN STREETS. And i bet you our traffic problem will lighten. TRAFFIC LIGHT EVERY BLOCK OR AROUND EVERU NEW BUSINESS IS NOT A GOOD IDEA NO MATTER WHO YOU ARE
ALBUQ, NM USA - Monday, October 19, 2009 at 08:43:20 (PDT)
About merging traffic -- a mistake I find many people make ist that they zipper into the continuing lane too soon, leaving hundreds of yards of empty lane in front of them, which encourage cars to leap ahead into that space and cheating. My solution is to stay in the merging lane as long as possible, adapting my speed to that of the continuing lane. This way no one accuses me of "cheating", the road capacity is maximized, and cars in the contiuing lane are encouraged to leave gaps without significantly slowing traffic.
Eric Trumpler <firstname.lastname@example.org>
Ludwigshafen, Germany - Wednesday, October 14, 2009 at 14:36:28 (PDT)
I would go along with your idea of introducing a wave absorbing gap between all the cars provided I had the adaptive cruise controls and higher speed lanes for those who use it....say 90 to 100 mph...to make up for the volume of traffic lost in the extra spacing.
[It's counterintuitive, but extra spacing improves traffic volume. Traffic flow is ruined whenever drivers get too close together. Fifty years of research has showed that best flow occurs when cars have 50ft to 100ft between them. At that spacing, drivers naturally go at around 40MPH. If they spaced themselves either closer or wider than that, total traffic flow would drop. When flow decreases like this, it causes a bottleneck, and cars coming in from behind make a jam start growing quickly. -billb]
Overland Park, ks USA - Friday, August 28, 2009 at 16:50:52 (PDT)
This is an excellent site. When learning to drive, my dad always taught me to leave lots of extra space in front me. This made common sense from a safety perspective. As I grew up and started driving on the freeways a lot, I noticed that if I kept a reasonable space ahead of me and slowly approached the stopped traffic, I could "dissolve" that small stopped wave ahead of me. I never thought much about this, until I just read your article. The experimentation and articulation of your site is very well done!
Thank you for providing this.
Even though I'm an aggressive drivers who likes to go fast, I'll use these common sense traffic breaking patterns.
LA, CA USA - Thursday, August 27, 2009 at 14:48:18 (PDT)
Bill Beaty, Leo Laporte tweeted about your video where you explain and demo your theory on breaking up traffic waves as a single operator. Awesome! You're the man! Thank you for validating me. This has been precisely how I try to drive in the notoriously jammed up Minneapolis/St. Paul metro. I picked up the trick for smoothing jams behind you by just leaving a big enough gap in front to allow a smooth average speed, intuitively. I just felt that smooth steady progress was better than coming to a stop. But the tricks about on/off-ramps were somewhat new. I'd let people in who were coming down on-ramps, but I thought I was just being nice. This is good stuff! Making traffic better! It'd be nice to have some telematics of some sort to give drivers prompting as to what speeds, if driven, might give the smoothest flow, etc. The advent of set-follow-distance radar-guided cruise control may make this sort of thing more commonplace. Thanks for the vid!
Andrew Skretvedt <email@example.com>
Grand Forks, ND USA - Wednesday, August 26, 2009 at 22:44:06 (PDT)
During two years of commuting to St. Louis from the Illinois side, I began noticing these exact patterns, and discovered the "use no brake light" rule for breaking up jams. I wondered if I was the only one realizing these preventive measures.
I noticed in my area a major cause of merging slow-downs is because of a large number of "Exit Only" lanes improperly used, and people waiting too long to get into the exit lane they need. There are jams every day going over the Mississippi into Missouri because of last minute mergers.
Thank you for getting the word out. I hope more people take notice to this and start making adjustments to their dirving behavior.
Jared Foley <@@firstname.lastname@example.org@@>
Granite City, IL USA - Wednesday, August 26, 2009 at 06:01:06 (PDT)
Your idea of Zipper merges is a bit simplified. The Zipper merge can only work as you have animated it so long as EVERY car that performs the zipper merge is comfortable driving at the new following distance for the ENTIRE duration of the lane reduction.
[No, you're assuming that both lanes are operating at maximum capacity. In that case the single downstream lane will become clogged, and a growing jam is unavoidable. But in the real world, jams at merge zones occur long before the downstream lane is overloaded. Also, jams at merge zones lead to empty downstream lanes. The jam itself is obviously the bottleneck. Remove the jam, and the flow increases. (Of course it can only increase to the point where the downstream lane can't handle any more cars. But that's not a traffic jam, that's just normal congestion. -billb]
What really happens I believe, the zipper merge occurs as the flashing arrows tell us to get over. One car travelling 70 MPH with a responsible 2 second following distance merges to the right, and is now going 70MPH with a 1 second following distance. Driver lets off the accelerator and drops the speed down to 65 MPH until the car in front of them has pulled ahead enough to allow a 2 second following distance again. Now the cars zippering behind this car will have to additionally slow until complete congestion happens.
I stumbled upon your site after reading up on hypermiling. My commute has always been aggravating and for a while I took Metra to work. The upfront cost of ~$140 really crushed my budgeting so I started driving again. It dawned on me I could alter my driving habits to increase my gas mileage. I started leaving gaps between the car in front of me whenever I got into a jam. I still haven't mastered being able to maintain a low speed and avoid some stop/starts but instead of "gunning it to the next car's back bumper" I stop with about 1-3 cars space between me and that vehicle instead.
The first full tank that I tried this technique on my mpg was 31. Normally it is about 23-27 mpg on the highway.
The first time I did this I also counted the number of cars that were able to merge either to or from my lane because of my gap. About 41.
Surprisingly I haven't gotten honked at or dirty looks from Chicago drivers. Some pass me or change lanes and I temporarily think this technique doesn't work...until I go right pass them again 5-15 minutes later. Their "cheating" didn't really advance them much further than the overall speed of traffic.
Nevermind the cheaters who use the side embankment as a lane.
Another side effect to doing this is that I'm not as frustrated from the lack of control. The critics of this technique aren't getting that by building and maintaining a gap you're decreasing traffic density.
Chicagoans are very aggressive and fast drivers but I noticed that Seattlites tailgate *really* badly.
Chicago, IL USA - Thursday, August 20, 2009 at 12:00:56 (PDT)
Do HOV lanes increase traffic jams?
Imagine a city that is 20 miles long and the HOV lane goes 15 of those 20 miles. If a traffic wave forms in the last 5 miles, won't the HOV lane simply add traffic more quickly to the back of the wave and not allow it to dissipate?
It might be better to have the HOV be an express lane with no exits for the 15 mile length and to have it controlled with a signal. If the average speed over this 15 mile length is close to the speed limit, don't allow entry to the HOV lane. If the average speed slows, then a signal would allow all cars who choose to enter the HOV lane for a minute or two which may shunt enough traffic around the plug to cause it to dissipate.
pueblo, co USA - Thursday, July 30, 2009 at 07:01:23 (PDT)
I find it interesting that you do not touch on the other, less traffic-related effects of replacing the "stop-and-go" with slower, more smoothly moving traffic. Firstly, the less you need to use your brakes, the less often you need to replace them. Secondly, the less often you need to accelerate your car as apposed to maintaining speed, the less fuel you will use (and believe me, this is noticeable). Additionally, less acceleration means less heating your (auto) transmission fluid, and the less it needs to be replaced, also the less strain put on your motor mounts, and the longer they will last. When you multiply the 2-3 mpg increase in efficiency across a thousand cars on the highway, every day, that adds up to a very large amount of fuel saved. There a multitude of effects of intelligent driving along with better traffic patterns.
Philadelphia, PA USA - Wednesday, July 29, 2009 at 23:01:39 (PDT)
I have been using the method you describe for about 12 years now. But I've never seen it as brilliantly explained as this.
Gershon <gershon_ben_franja@y a h oo.com>
Pueblo, CO USA - Wednesday, July 29, 2009 at 13:12:23 (PDT)
this is just an efficient way of wasting road capacity.:
MAINTAINING SPACE: traffic flow is highest at the speed where cars have about 75ft of empty space between. Drivers who close up ranks may think that they're helping stuff extra cars onto the highway, but instead they're ruining the system's throughput. As the number of cars per mile doubles, their speed decreases to less than half, so the average flow gets worse instead of better. Here's the rule of thumb: whenever you're driving smoothly at speeds above 35MPH, stuffing more cars onto the highway does help the overall flow. But in congestion, when you're moving at less than 35MPH, closing up the gaps causes a drastic drop in speed. The total flow rate is decreased where drivers pack themselves together, causing a bottleneck and leading to growth of huge traffic backups. If every single car maintained ideal space, then 75-foot spaces would give the highest flow. But if only a few drivers maintain space, then perhaps those who do it should try to inject much bigger gaps into the congested region ...but that's not been researched. (All this of course ignores benefits from better gas milage, fewer accidents from emergency braking, and fewer jam-triggering events at merge zones.)
TRAFFIC WAVES: When the "waves" appear, road capacity gets worse, and average speed goes down. This makes sense. Imagine instead that a clot of traffic caused capacity to improve. In that case the clot would dissolve even faster than it formed, and never grow at all. But if small traffic waves make capacity a bit worse, they collect more cars at their back end, which makes each clot grow and grow. -billb]
USA - Saturday, July 11, 2009 at 07:57:02 (PDT)
I have been taking space onto the highway for over a year since I first came across your article and I have to thank your for more peace of mind, better fuel economy and safer following distance. Your prescription is right on. I cannot believe how one person's actions can make so much difference in the usual problem spots on the road. Driving this way sure beats trying to reform the ill-mannered traffic stalling tactics of others. Thanks!
jim <jamesefa(double argh)e(double ell)at gmail dotcom>
mpls., mn USA - Wednesday, July 08, 2009 at 12:36:19 (PDT)
Is there a bumper sticker for the trafficwaves site? I would like to put one on my car.
NJ USA - Sunday, June 28, 2009 at 06:45:56 (PDT)
Here's a link to an article describing a recent study that pretty much confirms your ideas about the causes of traffic waves:
Alas they conclude that the phenomenon is unavoidable.
The readers' comments section is interesting - I took the liberty of posting a link to this site there. One commenter is a trucker who (consistent with your comments) said "we've known about this forever."
Jamie Hanrahan <email@example.com>
San Diego, CA USA - Sunday, June 14, 2009 at 01:59:43 (PDT)
I had to laugh when I read this. I used to live in Bellevue and the traffic 'waves' on 405 used to kill me, especially heading south. I used pretty much an identical method as you and would frequently notice behind me that traffic moved steadily and smoothly, while the traffic in the other lane would be at a near stand-still.
If only we could get more people to follow this.
Akron, OH USA - Tuesday, June 02, 2009 at 12:55:36 (PDT)
Your a douche bag. If your going to drive slow get out of the left freakin'lane
IL USA - Thursday, May 14, 2009 at 09:09:04 (PDT)
There is an interesting solution to this wave problem that has been deployed to great effect in Stockholm. They have automated signs that tell you how fast to drive to keep these waves from forming as you drive into Stockholm from the North. If you maintain the suggested speed, the cars will not bunch up. If you exceed the suggested speed by even a few miles an hour, then in about 10 seconds you will see it bunch up and you will have to slow down. I am not familiar with how they calculate the speed, weather it is just the average speed or something else. But I did play with it when I was there and it seemed to work well.
Another contributor may be automatic transmissions, in that they tend to want to go a minimum speed with your foot off the break. I find when driving a diesel (which has lots of back pedal) using a manual transmission, I tend to maintain a more uniform speed by being in the correct gear and averaging out the speed so that I don't have to stop and shift down into first and then second. It is kind of an incentive to keep it moving at a steady speed in order to not pay the stop and go penalty.
I wonder if it may be better in Europe where most cars are small diesels with manual transmissions compared to the US where almost everything has an automatic with a gas engine..
Steve Reese <sareese@earthlinkkkkk$$%.nettt>
Seattle, WA USA - Tuesday, May 05, 2009 at 23:24:19 (PDT)
My posting is not a question, but an answer to one in the Traffic Waves article. As a trucker we drive with a large empty space ahead of us because it takes us almost that entire space just to slow down, let alone stop. The average car can go from sixty to zero in almost no time, a loaded semi (even if he doesn't lock the tires up) can rarely go from 55 to zero in less than 300 feet. Truckers use an "eight second rule" whereby we watch the car ahead of us pass a sign, mile-marker, or other roadside object and if we reach the same object in less than eight seconds we're too close. In the event of an emergency stopping situation it takes roughly eight seconds for the brain to react, hit the brakes, and stop a loaded truck. If we didn't leave that massive space we'd kill you :) But you're right, it does help in a heavy traffic situation as well, allowing other traffic to merge, distort, and swap positions around us like a boulder in a stream.
Medford, OR USA - Thursday, April 02, 2009 at 04:00:31 (PDT)
This is absolutely fascinating. Thanks for sharing! I will be trying some of your tips on my commute home tonight.
Brisbane, QLD Australia - Tuesday, March 24, 2009 at 14:45:54 (PDT)
Thanks for the interesting analysis. I read recently that ants do not experience traffic jams, even though their tracks are linear, like highways. The main observation was that ants don't pass, ever.
Do a google search on ant traffic jams if you're interested.
Madison, WI USA - Tuesday, March 24, 2009 at 11:28:42 (PDT)
Your amazing video was blogged here:
Thanks so much for making this great piece available to share.
Zen Moments <firstname.lastname@example.org>
Devon UK - Thursday, March 19, 2009 at 10:35:20 (PDT)
I have a 50 minute commute in Pittsburgh (city of bridges and tunnels). Depending on my mood I will either lay back and open a gap in front of me, or I may just be a "cheater". I was thinking about your observations and they fit right in with my drive and the book on chaos theory I am currently reading.
Another way to think of the non-linear behavior is to consider populations of two types of animals, one predator and one prey. Intuitively you would think that eventually some happy balance is created. But that is not how it works. The system is non-linear with lots of what seem to be unexplained dips and peaks in the population. The populations never settle into an equilibrium.
Anyway, my point is that "cheaters" are like the predators and "gappers" and like the prey. The more gappers there are, the more incentive for cheaters. When there are too many cheaters they will begin converting into gappers.
That is why I am convinced that in the overall scheme whether I am a cheater or gapper, my influence will not really impact the system greatly. The only constant is that my commute experience is different every day.
Btw, your youtube videos are great. They make me think the theme song to MacGyver every time. http://www.youtube.com/watch?v=uxEq7dy3-Ok
USA - Friday, February 20, 2009 at 23:06:48 (PST)
This is all pretty obvious stuff, man. You've made no major discoveries here.
USA - Saturday, January 24, 2009 at 13:04:19 (PST)
I've been doing this for a few years between 520 and Federal Way on a regular basis. I started smoothing out the waves simply because the constant 'jackrabbit starts' to catch up with the back of the next wave put too much wear and tear on the 35 year old V8 in my work van. And the fuel economy benefits are a nice side effect when you only get 8 miles to the gallon!
Seattle, WA USA - Monday, December 29, 2008 at 16:23:29 (PST)
Thank you for the great work, many years I am thinking and practicing the same about traffic, but here in Italy I am regarded as an heretical - we build Ferrari, must drive as all of us have one.
It was heartening to me finding your site.
renato <email@example.com character>
grosseto, italy europe? - Wednesday, December 24, 2008 at 06:02:31 (PST)
Your a bunch of fools!
It's only going to make it worse behind you! You make everybody pissed and angry! If you watch the video, you can see traffic is moving in the other lanes! Your just holding everybody up behind you and causing people to bunch up and create more of a danger...your stupid!
Eric Soderlund <firstname.lastname@example.org>
poway, ca USA - Sunday, December 07, 2008 at 08:16:21 (PST)
It doesn't happen often, but leaving space in front of you would help in making room for emergency vehicles.
USA - Wednesday, October 29, 2008 at 10:05:46 (PDT)
Great work! One idea, though: I think you could sell the fuel-saving aspect harder. Cars use a lot of fuel when they accelerate but very little when they maintain a constant speed. Creating a steady flow of traffic will reduce fuel consumption and fuel emissions dramatically, and this would be the case even if there were no improvement in road capacity. More uniform car speed also reduces break pad wear.
[I should re-do it. I wrote it years ago, back when everyone was buying huge SUVs. -billb]
Knut Olav Homlong <khomlong att online dott no>
Stavanger, Norway - Tuesday, October 28, 2008 at 06:38:41 (PDT)
I like your thinking and analysis. I concur with your thoughts on the overall affects of facilitating better merging and it's positive affects on traffic behind you.
However, in terms of waves from rubberneckers or an errant brake light flash - the drive slower, keep a gap, minimize use of brakes does only two things ... it gets rid of the stop and go traffic, which you define as the traffic jam, and it allows you to relax and have a "better" commute. It does not get you to your destination any faster.
[I've seen traffic jams break up ahead of me. In certain situations the gap very definitely does get you there faster. Ask this question: when traffic "stop-waves" arise, does the average flow of traffic get changed? I've found several sources saying yes: the waves cause a significant drop in average traffic flow. And so various cities across the world have provided systems to wipe out traffic waves (e.g. variable speed limit signs, etc.) Second, if a "wave" gets trapped at a merge zone, the merge zone jams up. This jam creates a long upstream backup. It also creates very sparse traffic downstream, so once you get past the jammed merge region, you can drive fast. Clearly this jam is acting as a major bottleneck, and if one or a few drivers could wipe it out, the flow would increase. So... waves and jams really are bad, and if we could remove them, traffic flow really would improve. -billb]
I've thought long and hard about this - the simple fact is that what you do to the wave does not affect the average speed of the traffic in front of you. The average speed of the traffic in front of you is the only thing that controls the average speed of your commute. Would the world be a better place if we all had more relaxing commutes? Absolutely yes. Would there be fewer accidents? Probably. Would fuel and brake economy go up? Yes.
Now - if someone way up in front of you - say someone who was following someone who taps their brakes to change a lane would use these techniques, and some very low percentage of cars between you and them would also use them I think we might see an improvement in traffic - but the key person is really only the one right behind the idiot that only cares about them and not the rest of the traffic.
So what should we do ... start by adopting your suggestions. The most beneficial thing you can do for traffic, in my opinion, is to move over to allow cars to pass you as long as by moving over you do not cause the new driver behind you to brake or slow. This effectively creates additional road capacity by moving cars from behind you to in front of you. Second, do not change lanes with your brake pedal - use your eyes, mirrors and turn signals instead. Third, drive with space to minimize your use of your brake lights (they come on before you even start to slow). The cars 2 or 3 places behind you will react to your brake lights by slowing (or braking) even when you don;t really slow much - this is a implied safety/uncertainty thing. Fourth, maintain speed and direction - an old sailing trick. If everyone knew that everyone else was going to drive predictably we could merge, change lanes, not brake, etc. with comfort and implied safety. This also means that when you go around a corner or up a hill you should (safely) maintain your speed as well. And finally, be polite - it's the culmination of all of the above. (And never, never rubberneck - not that this helps but it feels good!)
Just my 2 cents - thanks!
Dean Suhr <trafficwaves - att - deansuhr - dott- us>
West Linn, OR USA - Wednesday, October 01, 2008 at 11:01:41 (PDT)
I have tried this idea and it really works. I now incorporate in my driving style both in my car and in the Big Rig. Cheers
Martin Davies-Roundhill <email@example.com>
Brisbane, QLD Australia - Saturday, September 27, 2008 at 16:27:04 (PDT)
I had no idea there were so many others that think just like me. It's refreshing and kind of exciting. I need to get out more.
Thomas May <knobs@roadrunner DOT com>
Chesterland, OH USA - Saturday, September 13, 2008 at 08:26:55 (PDT)
"KEEP RIGHT EXCEPT TO PASS" "DO NOT CROSS SOLID LINE"
Ever see either of these signs? They are the law on the Interstate Highway System.
[My usual policy is to ignore angry comments. But I'm making an exception. Please cool down and watch/listen to videos before criticising. Did you see the signs? That's not a passing lane, that's an EXIT ONLY lane going to the express-lanes. Note that *everyone* ahead of me is taking that exit. This is a famous Seattle traffic jam, "Northbound I-5 Expresslane Left Exit, South of Seneca St." It's a jam where all commuters are jockying for position in the left-hand exit lane. Usually during rush-hour that left lane is needlessly packed solid and crawling very slow, with occasional stop/go waves. I let many drivers merge, and by the time I arrived at the source of the jam, it was already flowing. I barely had to stop. Second, if instead I had closed up my forward space and blocked all merges, I could be arrested for *tailgating,* because I'd be driving unsafely with less than one second between cars. Yes, everyone else in the jam is illegally tailgating too. And finally, the solid white line is marking the HOV lane. See those diamonds painted on the lane? -billb]
Next time, why not show us what your pious selfishness
[ All these drivers are trying to merge into the exit-only lane, and I'm the *only one* letting them in ...and you think that's SELFISH? No, my behavior is called "Trucker's Politeness" or ""Gravel-pit Etiquette." Watch how the professional truckers behave in congested conditions. They have an overhead view, and they can see which selfish behaviors are triggering the congestion. If we merge like gear teeth,the jam will drain away. If we block all merges, the "gear teeth" come to a halt. -billb]
is doing to the traffic BEHIND you.
[ Please stop and think. Obviously the the traffic behind my car is moving as I do: it's flowing totally smooth with no stop/go waves (well, except for that one clot down at the actual exit.) By letting ten cars in, I did pull the entire region of slow traffic backwards by about 10-20 seconds. I might have made everyone's trip last maybe 20 seconds longer. But then the jam was unplugged! The jam slows everyone down by several minutes. That's how one driver can un-jam a merge zone: pull the jam backwards, so down at the far end, fast merging can commence. What I *really* needed to show was a split-screen view, with the second screen being my camera on the overpass (that's the same view as the first five seconds of my video.) Here's what you'd then observe: first some periodic "cheaters" are forcing their way into the solid-packed left lane and perpetuating a long backup. But then I arrive and the supply of cheaters would mysteriously dry up. That's because I'm letting them all in early. Next, with no cheaters to halt traffic, the line speeds up and the long backup in the left lane would rapidly drain away. Finally I would arrive, and the cars behind me would be flowing fast and smooth, with spaces between cars which allow merging. In that video it *almost* worked, but there was still a small backup remaining when I arrived at the end, which halted everyone for a few seconds. (A single traffic-wave still remained after I'd gone past. ) But how long would my unplugging last? There's nothing downstream to re-trigger the jam, so it might remain smooth for the evening. Or perhaps two angry drivers would intentionally block each other, "get stuck in the narrow doorway," and re-start the whole thing again. -billb]
altoona, la USA - Saturday, August 30, 2008 at 09:22:45 (PDT)
You know, I've actually had some of the same thoughts about "pressure waves" of traffic as you talk about here... but I've let myself become so irritable and competitive that I've never taken full advantage of the phenomena I've observed. You've given enough credence to the hypothesis that I may be able to get out of the rut. Thanks!
Schaumburg, il USA - Sunday, August 24, 2008 at 15:53:09 (PDT)
In Michigan, you have to drive slower on the right and pass on the left which makes sense since the driver sits on the left side of the car. Driving below the speed limit, not passing anyone, and leaving open lanes to the right will force people to pass on the right. If a traffic jam sneaks up on them because they can't see through your car to the right side to make the pass, then it could cause an accident in those lanes and tie up traffic. Why don't you do this in the right lane? You also crossed a solid white line in your video.
[Look at the overhead signs: that's an exit-only lane and not a passing lane. Look at the diamonds painted on the lane with the solid white line: that's not a "cars not allowed" lane, that's an HOV lane. -billb]
royal oak, mi USA - Monday, July 28, 2008 at 10:15:33 (PDT)
Maintaining a minimum safe follow interval is the key to maximum vehicle thru-put. For example, if 18' LOA cars drive 60mph at 2 second (176')intervals, 27 will go by per minute. Vary the speed in the range of 40 to 80 mph and the thru put stays at 26-28! Why? THE SAFE FOLLOW SPACE GOES UP WITH THE SPEED. Cut the follow time to 1 second (88' or ~5 carlengths at 60mph) and thruput doubles to 46-52 cars per minute. At 3 sec. intervals only 18 to 19 cars go by. Shorter intervals don't work well if larger, higher, wider vehicles obstruct forward vision leading to unnecessary braking, passing, and increased follow distance. Not to mention driver fatigue and frustration. What can we do? Where possible, open the left/ HOV lanes to any CARS of compatible dimensions. NO TRAILERS, CAMPERS, SUVs, VANS, TRUCKS, BUSES, ROOF RACKS. Keeping the bulk of the cars out of the way should make things easier and less congested for the truckers, especially with the move to smaller personal vehicles.
Williamsburg, VA USA - Saturday, July 26, 2008 at 10:13:05 (PDT)
i too am from seattle and have been pondering this same situation, but have yet to put it to pape. hats off to you for doing so.
marysville, WA USA - Thursday, July 24, 2008 at 21:30:26 (PDT)
I have been a long time advocate of this idea. I even had it published in our local newspaper (Philippine Daily Inquirer) in 2004. Had it named Greenzone Traffic Scheme. Sadly, it came on deaf ears. People can't seem to grasp the idea that keeping a large space in front as a headway decreases the chance of congestion. The fear of being cut into is more a concern than having themselves a hand in removing the congestion. You're site is heaven sent for me. You're FAQ's answered every questions asked at me everytime I tried to present it to Traffic Managers in the Philippines. It is the solution that can be used worldwide. Tailgating is such a bad habit that causes these traffic jams all over. Thank you for letting me know that I am not alone.
mars de jesus <firstname.lastname@example.org>
marikina, Philippines - Thursday, July 17, 2008 at 02:15:30 (PDT)
Good read. Haven't thought through the "math" but generally agree that in most circumstances cooperative behaviour increases the size of the pie while competetive behaviour decreases it. If you believe life is a zero-sum-game it will be...
Recently I was in a miles-long standing wave jam on a rural stretch of I-75 and noticed that two semi trucks ahead of me would park side-by-side for a minute or more and then slowly roll up to the tail end of the jam. This did not have the effect of busting the jam (maybe it was too large), but did allow (force) those of us behind the trucks to take our foot of the clutch and relax.
Long story short... I think the trucking industry is probably a better change agent than state troopers since they are out there anyway (free to taxpayers), have 2-way communication, and are extremely interested in highway efficiency.
As for urban signalized arterials, I understand that in Germany, some traffic signals have an LED display showing the progression speed of the green wave (which may differ from the posted speed limit). This provides feedback to drivers to maximize the throughput of the road.
Demian Miller <email@example.com>
Tampa, FL USA - Wednesday, July 09, 2008 at 03:53:44 (PDT)
Interesting article.. I unwittingly have been doing this for about 6 months now. Strictly for economic reasons I slowed down and started hypermiling a bit. Which leaves those huge gaps sometimes. I drive a diesel pickup. I often wondered why I wasn't seeing the huge jams I was used to dealing with. And my commute time had not really increased that much. Strange.. Then I read you article. And it dawned on me. I was actually eliminating those jams by driving just 5 mph slower than anyone else and employing hypermiling techniques! My stress level is lower, my mpg is increased, and I'm helping the overall traffic situation.
Can one person make a difference. YES! By forcing the Lemmings behind us to conform to the traffic pattern, we ALL benefit. Lemmings line up on the left, now in one lane instead of 2, and smooth drivers are now on the right. Maintain that space and all is well. Very cool article! Thanks! very enlightening!
Cumming, Ga USA - Tuesday, July 08, 2008 at 08:43:16 (PDT)
Looking at the traffic around my area I'm constantly amazed by the similarity between it and 1D non-steady internal compressible gas flow!! You see the compression waves and expansion waves travelling through the medium. Traffic lights are a source of unsteadiness in the flow; the further away you are away from them the more like steady-flow appears. Stop the flow and, as you show, a compression wave is created.
I'm glad I'm not the only mad man that sees these things!
Graham C. Williams <firstname.lastname@example.org>
London, UK - Tuesday, July 08, 2008 at 04:15:34 (PDT)
Hey, someone else did the wave-eating experiment independently. And on I-520 too!
Don't Stop Moving (on Seattle 520)
seattle, wa USA - Friday, July 04, 2008 at 09:51:25 (PDT)
Hey, I just read your piece for the first time. Exactly right! I've been using these techniques for several years. At the worst jam on my commute in San Jose, CA, there are just two lanes plus a commuter lane. Usually nowadays someone will cruise alongside me with a huge space in front of us so we reach the jam just as it is breaking up. No stopping! I imagine it evens out the traffic flow behind but I never stopped to check it out. Certainly much safer.
USA - Friday, July 04, 2008 at 01:11:46 (PDT)
We love your website. My husband and I traversed the 520 for about 6 years. We used to try to eleminate the jams by going very slow. It worked really well. I did want to comment that your zipper/merge theory does really work. In California it is law that you must merge every other car. Merging is fast and effortless and there is rarely a slow down. Most of the time we merged at relatively high speeds. I was suprised when I got my Washington Drivers License that it was not a law here. Then I noticed that people here feel it is doing a favor to let a vehicle merge. It then becomes a waiting game with many cars starting to merge too early, stopped with there blinkers on hoping someone will have pity and let them in. I think it would be great to get the "merge every other one " on the books and make it state law. At least the new drivers will know how to merge like a zipper!
Thanks again for your site, it was an enjoyable read.
Mrs. Juliet M.
Seattle, WA USA - Thursday, July 03, 2008 at 14:58:16 (PDT)
Bill, we need to talk. For the past three years I have been formulating an outline on a new driving theory. The theory is based on psychological, environmental and other driver variables. Please email me as soon as possible.
USA - Sunday, June 29, 2008 at 05:54:33 (PDT)
You should check out beatthetraffic.com and add it as one of your links for live traffic information.
Michelle Rodrigues <email@example.com>
Campbell, CA USA - Thursday, June 26, 2008 at 12:16:01 (PDT)
I also naturally developed this behavior (and amusingly I'm also an electrical engineer), but I am especially pleased that you are actively spreading the technique. It's good that you reassure readers that the weird feeling ... of doing something visibly different right in the middle of the ''traffic herd'' ... will go away as a driver raalizes it's actually working better than expected. I started driving the average speed initially for my own benefit: as an engineer who prefers to optimize most things I am repeatedly faced with, it felt unpleasant & wasteful to have to keep starting and stopping (and shifting too, if driving a manual transmission). [Two site suggestions: consider adding a checkbox option to keep commenters' email addresses hidden; Firefox (3.0 at least) doesn't seem compatible with the Security Number field.]
USA - Thursday, June 26, 2008 at 11:42:10 (PDT)
Hey man, great article, musings, collection of speculations. If people could only know...
But, I had written with a specific topic. I was required to go through a AAA safe driver training course for a job I used to have, and the instructor brought up a great point, and one which has haunted me for years since. He told all of us on the first day, while we were sitting around venting frustrations on the lack of everyone else's ability to drive, that if everyone were to actually maintain the 3-second window (between you and the car ahead of you), there would be no traffic jams, no accidents, nothing. Unfortunately, it would seem that our more basic instincts drive us to, well, drive like asshats (with a rare exception here and there). And it's true, a lot of people absolutely suck at driving. Obtaining a DL should be a lot more difficult, and the testing a LOT more in-depth, but that's just me ranting my own personal opinion.
Just thought I'd throw that so-close-but-will-never-be-reached goal out there for the people who, apparently, care enough to post this kinda stuff. Enjoy, and don't let the thought of how easy life would be if people didn't suck so bad at controlling their own f*cking speed and direction get you down too much.
Joe Novak <firstname.lastname@example.org>
Sacramento, CA USA - Wednesday, June 25, 2008 at 01:37:12 (PDT)
Last week I was driving across lower Manhattan during rush hour - something not for the feint of heart. I was driving from the Williamsburg bridge (at the East River) to the New Jersey-bound Holland Tunnel (at the Hudson River), a distance of only two miles. I noticed that the traffic control agents were pushing cars through the intersections as I arrived in Manhattan, and as I got closer to the West side and the Holland Tunnel, the congestion at the intersections became progressively worse until it took nearly twenty minutes to proceed across one intersection. It ould have been better if the traffic control agents had been in touch with each other and directed the agents closer to the Williamsburg Bridge to hold or slow incoming traffic so as to prevent the buildup on the New Jersey-bound west side with its resultant gridlock. The point is clear...there are only four lanes of traffic exiting that part of Manhattan, and those four lanes have a discrete capacity. Therefore, the incoming traffic rate must be limited to equal to or less than the outbound rate. If this is done, gridlock will be lessened and traffic will flow smoother even in an extremely congested urban setting.
Len the Pharmacist <LenRx1@aol.com xxx>
New York, NY USA - Wednesday, June 18, 2008 at 01:55:28 (PDT)
I think that the backups where, for example, 3 lanes merge down to 2 lanes, are a perfect illustration of Bernoulli's Law, in two dimensions. And in order to eliminate the backup, you would have to reduce the incoming traffic speed and thus the input volume to match the output volume. This would require constantly updated speed advisory signs for the approaching traffic, perhaps half a mile ahead of the merge. Has this ever been tried?
Clyde Revilee <email@example.com>
Santee, CA USA - Friday, June 13, 2008 at 10:51:31 (PDT)
I was going through this site and reading the experiment along with some comments. What struck me is the common sense of it all and the fact that it really boils down to simple courtesy. I have often wondered why, when there is a sign that points out this lane ends, no one takes the time to merge BEFORE the lane ends. When the lane ends, they have to slow down to merge and they get "punished" for not merging sooner. I have often slowed down before such events to allow those in the lane that ends to move over but they apparently don't think that far ahead.
Larry Hanson <firstname.lastname@example.org>
Houston, TX USA - Friday, June 06, 2008 at 09:24:26 (PDT)
I have been watching traffic waves in Sacramento for about 8 years now. I thought I would share a few observations:
1) The sun can cause a wave to start. This is especially true in the morning comuute on I-80 eastbound. There is a location where the road changes direction, and during the right time of year, that direction directly faces the rising sun. Drivers can't see as well and naturally slow down. The wave that starts can last for quite a while.
2) The traffic waves that start can move quite fast under certain circumstances. I was driving and there was a traffic wave started by an accident on the freeway in the opposite direction of traffic. I was driving roughly 65 m.p.h. For about 2-3 minutes I was driving alongside the wave departure point of the wave (where cars were leaving the wave and finding a big open space). The wave was moving backwards at 65 m.p.h.!
3) I design computer chips that move data quickly and I have been able to apply my observations of traffic patterns to my job to make the computer chips work more efficiently. Neat Huh!
Sacramento, CA USA - Thursday, June 05, 2008 at 12:23:36 (PDT)
Like you, I have have an interest in traffic, and like you, I have spent a lot time trying to educate people about the wave-like motion of traffic congestion... although not on the web. Compaction waves cause much of the traffic that we experience, but people are quick to blame only the rubber-neckers or slow drivers. The truth is more complicated. I wish you had addressed it, although admittedly, your web site is focused more on the "cure" and one form of prevention. I feel like you've missed a major point: if most people traveled the speed limit, most traffic congestion would cease to exist.
To illustrate the point, consider your animated graphic on page 2 (Traffic "experiments") that shows a typical traffic compaction wave, where the input equals the output -- thus a pinned wave that doesn't move from its origin. You do an excellent job of addressing the properties of such a wave when the stopped cars are moving at 0 mph. Now imagine a situation where you are moving at 65 mph and observing those "stopped cars," which would therefore have to be moving at 65 mph (given relative motion). What causes such a wave? It's simple. When people travel varying speeds, faster moving traffic will ultimately run into slower moving traffic. If there are enough cars on the road, then enough fast-moving traffic will hit slow-moving traffic and create the compaction wave your graphic illustrated. Since most sane people do not drive 70+ mph while only a foot or two off someone else's bumper, you don't simple create a pinned wave moving at 65 mph down the highway (relative to your observational speed of 65 mph). Instead, the speeders at the rear start slowing down, which causes the people behind them to slow down, and so forth. Eventually, the wave moves slower and slower until (if there's enough continuous traffic) the wave stops relative to an observer at rest (or moves backwards relative to the observer traveling 65 mph).
Many people like to claim that slow drivers are the cause of slow-downs. As the example above illustrates, it's actually a combination of both slow drivers and fast drivers. The fast drivers pile up behind the slow drivers, and as passing opportunities dwindle (based on the carrying capacity of the road) and inter-car distances shrink, traffic slows down. Obviously, the best solution is to homogenize speeds -- have everyone travel the same speed. Impatient drivers think that slow drivers should just speed up, but to what speed? One fast driver may wish everyone was traveling 75 mph, while the impatient driver behind him/her wished everyone else would just speed up to 90+ mph or just get the hell out of the way. This is where the magic of the speed limit comes in. You don't have to rely on undocumented convention (such as "everyone just travels 5 mph over the speed limit"), which may not be known to all drivers. If everyone traveled the speed limit, to the best of their abilities, many (but not all) traffic problems would be solved, and major roads would be able to carry capacities closer to their theoretical limit. (Not to mention the fact that a lot of fuel would be saved, since fuel economy and speed do not have a linear relationship. As you drive faster, wind resistance actually gets stronger and reduces your fuel economy. Therefore, driving the speed limit has other advantages, not just for the individual driver, but for all of society.)
I sincerely wish you would re-write your website to address this issue first. Although you do an excellent job addressing anti-traffic techniques, you really need to address the other sources of the problem. Remember, an ounce of prevention is worth a pound of cure. If people would just learn to drive the speed limits and leave a safe distance between themselves and the car in front of them, we wouldn't have nearly as many traffic problems as we currently have.
Furthermore, in discussion of linear vs. non-linear dynamics, when you claim that a single car can make a difference and supports the non-linear view, consider that you also admit that you have to do the job of 10 drivers by leaving much larger spaces than you should normally have. Basically, if everyone left a single space in front of them, you hypothesize that many of these congestion situations wouldn't exist. That sounds very linear to me. Personally, I don't see traffic congestion as a linear problem, though exponential or logarithmic seem more likely to me. However, traffic definitely has its dynamic components as well, since each car in a defined segment of traffic will be moving at a variable velocity. But since vehicle speeds usually vary around a mean (or two or more peaks in a distribution, such as when one segment of traffic is mostly traveling around the speed limit and another faster wave traveling 15 mph over the speed limit passes them), I suspect a more of a middle-ground between dynamic and linear dynamics.
Also, without having known any professional truckers personally, I cannot speak to their knowledge. Like many of your critics, I've suspected braking distance was part of the explanation of their behavior. However, the other part of the explanation could also be fuel economy. To move a very large truck with a heavy payload, you burn a lot of energy. Common sense -- which is apparently lacking from people who travel more for convenience than for a living -- tells you that in a traffic jam, you'd be burning a lot of fuel just to quickly close a gap with the car in front of you, only to hit the brakes once you got there. It's kind of like my criticism of people who accelerate towards red lights. What's the point? So for truckers, fuel economy is critical for being able to make a living, and their large, heavy trucks greatly amplify the affects of poor driving habits on fuel economy. (If you don't think they're more sensitive to fuel economy than the average financially irresponsible American, consider all the recent trucking strikes that have been in the news lately.) Personally, I suggest stopping at a major truck stop and interviewing truckers to get a definitive answer as to the true nature of their behavior.
This is one problem I noticed with your spacing idea, assuming that it is adopted by everyone, as you wrote in the following:
"Do large spaces between cars cause traffic to fill highways? Only if people increase their spaces on average, but that's not what I'm discussing here. Suppose traffic is moving at 40mph. Those cars will have several car-lengths of space between them. If those drivers encounter a slowdown... what happens if they REFUSE TO PACK TOGETHER. This doesn't increase the space between drivers, since they already were widely spaced when they were moving at 40mph. Yet if they maintain their wide spaces as they slow down, then they will have little trouble in speeding back up again."
The problem here is a practical one, whereas the situation outlined in a nice theoretical one. Assuming you could condition all drivers to drive as you suggested, there would still be a lag time as drivers reacted to the change in speed of the drivers in front of them. In the ideal situation, where you can see far in front of you, you only have to worry about people being distracted or tuned-out from their monotonous commutes or long trips. In hilly situations, curving roads, or where some people drive huge SUVs, while other drivers drive smaller cars, not everyone can see what's going on in front of them. In those cases (and others like it), the distance between cars will change in proportion to the driving speed since you will be dealing with reaction speeds and driver distractions. Faster moving traffic will result in faster changes in inter-car distances, and the reserve is true for slowing moving traffic. Either way, the distances between vehicles will not remain constant.
"If EVERY driver was to constantly maintain a HUGE space regardless of speed, then it would probably cause problems. The capacity of major highways would be reduced. On-ramps would become choked as traffic backed up into them, and there would be slowdowns extending far out into the countryside. Yet if ALL drivers were to change their habits, then only small spaces would be required, yet waves and stoppages would be seriously squelched."
If the carrying capacity of every road and freeway were infinite this would be true. However, each road can only carry so many cars per minute. Despite all your points and mine (regarding homogenizing speed), the truth is that waves and stoppages will continue to occur as traffic overloads the carrying capacity of the roads intended to carry it. This is simply a growth problem that lies in the hands of urban developers, not traffic-busters like you and speed-limit-drivers like me.
FYI: Your charity example in your FAQ is short-sighted... for exactly the same reasons that most conservation is short-sighted. (Sorry.) You wrote, "Giving to charity improves the whole world in the long run even though you lose money in the short run. Reciprocal altruists give to charity (they aren't true altruists, instead they are paying out money and expecting long term results!)" The second sentence is correct. (I have a degree in evolutionary biology, so I know exactly where you're coming from.) Giving to charity won't improve "the whole world in the long run" because the money will be used within a capitalistic system that is unsustainable. (This is known as Jevon's Paradox.) You're adding fuel to the very fire that's consuming the world. (A good, biological example of this is giving food to starving nations, since giving them food raises their carrying capacity and, without the social structure to encourage small family sizes, ultimately increases population, creating even more hungry mouths to feed.) Even most conservation efforts that result in monetary savings don't help because the money saved is ultimately spent on our unsustainable economic system or put into savings, which then gives banks loaning power to fan the economic flames. Therefore, charities do not improve the world, but make it worse. We need to be living sustainably for them to truly make a difference. If you're looking for an good ethics lesson, dig a little deeper. ;-)
Alex Dunkel <visionholder*AT*gmail*DOT*com>
Newbury Park, CA USA - Wednesday, June 04, 2008 at 00:42:28 (PDT)
I think Gravel-pit etiquette has moved to http://www.skaggmo.com/gravelpitetiquette.htm
Your jam solution reminds me that years ago when my parents were living in Aruba, I found that it was actually the law that one had to maintain at least one car's length behind the car in front "to allow merging in emergency". Everyone drove wickedly fast, with lots of passing, on roads with one lane in each direction. But it all went very smoothly.
Smooth driving to you,
Ken Landaiche <Kenneth.LandaicheETl-3com.com>
Santa Rosa, CA USA - Tuesday, June 03, 2008 at 17:57:09 (PDT)
Yes! Yes! I thought I had come up with this idea and I even have a great business plan to boot that would make this sort of driving a money-making venture. One key idea to help you learn this technique, "Try to use your brakes as little as possible and try to make other people use their brakes as little as possible." Sounds scary, but it forces you to leave space in front of you and to plan ahead so you don't cut other people off (so they don't have to brake).
One place this technique can also help is when YOU are merging, say from a metered light. Instead of speeding up just to slow down again, create a space in front of you to eat up the slow-pokes who had to stop to merge, then ride the merge lane all the way to the end, gradually picking up speed and merging! It works great! I haven't noticed anybody be angry at me because I went to the end of the merge area and I suspect it is because I was able to merge without making them hit their brakes. They were none the wiser and we were now going 20 to 30mph from a complete stop. (Okay, maybe somebody cursed my name for being a jerk and waiting until the end, but it fixed the merging for minutes behind me.)
For those skeptics out there, just give it a try. You will be amazed! Also, like anything else, it takes practice to get it just right. But when you do, you can claim victory over traffic and feel great about helping those behind you. Who knows, maybe because you did this, you changed the whole mood and course of someone else's day!
Seattle, WA USA - Tuesday, June 03, 2008 at 10:26:36 (PDT)
Nice job on a very interesting and thought-provoking article; I have a long commute myself over here on the East Coast, and I appreciate the usefulness of fluid dynamics when it comes to analyzing traffic patterns and the causes of traffic jams.
But I often find myself wondering if anyone's given any thought to the impact on traffic flow of the ever-increasing size of the "molecules" themselves. Everybody knows tractor trailers, school buses and other big vehicles can't move nearly as fast as our zippy little Toyotas and Hondas, and many of us get extremely frustrated when we get stuck behind them. Simple physics is all you need to realize that the bigger and heavier a vehicle is, the higher its center of gravity and the longer it takes to get up to speed, slow down, turn and do all the other things you need to do in traffic. As a result, the more of them there are on the road, the slower traffic moves overall.
But what are we doing in America? We're buying bigger and bigger SUVs and minivans and more and more of them every day. I can't help but thiink this has to be a major contributor to traffic problems. To keep traffic moving, we're going to have to move slower and slower ...and slower...
Just something to think about.
King of Prussa, PA USA - Friday, May 16, 2008 at 13:21:19 (PDT)
I'm no expert in this, but I have a friend who did an extensive research project regarding general queue theory. There is a very precise mathematical model for traffic jams out there, you just have to be crazily into math to figure it out. I just thought I'd throw that out there for anyone interested enough to look into it.
Joseph <rather not say>
Richardson, TX USA - Monday, May 12, 2008 at 15:25:55 (PDT)
Ideally, if everyone drives at 100 mph and keep a large distance between them, and all the merging cars merge in at 100 mph and fit into the slots, and the slots are slowly expanded before and after the newly-merged cars, then disruptions to the throughput of the highway will be minimized. But of course that's not possible because everyone has to be smart, deft, young, non-Mexican, imprudent, ungreedy, and not in a hurry. I believe throughput can be maximized in traffic jams, but only before the jam happens. It takes only one idiot driver to quickly cause a congestion; and resolving a jam is difficult if not impossible. Your theory is a good theory, but make it more concrete with a simulation.
Milpitas, CA USA - Friday, May 09, 2008 at 09:10:18 (PDT)
Interesting reading, I have myself cancelled out waves travelling back down a line of traffic many times, but am a little ashamed to admit I'm not very good at allowing people to merge in front of me. Inspired to give it a try now though.
Interesting to read the comments about the M25 around London, UK. Within days of the variable speed limits being introduced and enforced commuters around the south and western sections of the M25 realised their journey is often quicker, far less tiring and stressful, as the variable speed limits restrict the flow of traffic into the congested areas. This tends to be worse around Heathrow airport Europe's busiest air port, lots of entering and exiting the motorway/freeway across several junctions and junctions with other major motorways. These days people are pretty good at observing these speed limits as the benefit is now so well known.
Adrian <adrian at camdaw.net>
Alton, Hampshire UK - Thursday, May 08, 2008 at 08:30:15 (PDT)
Wow, somebody else who gets it. I've been busting clots like this for years, but not for traffic reasons. I do it to save gas. Mine and everybody in the line behind me. Thanks for the "let them merge" tip. Kind of like "pay it forward". Thanks for all your effort.
J Blob <dontblobme@mail>
USA - Wednesday, May 07, 2008 at 12:43:52 (PDT)
I have always thought that with a merge coming up, if everyone merged to the center so that everyone drives over the center line, things would go faster. It makes both lanes merge and eliminates the lane where people can pass ahead and merge near the front.
MI USA - Wednesday, April 30, 2008 at 13:13:24 (PDT)
Fellow Road Warriors:
I am so AMAZED that Bill and others have discovered/studied what I have, over the course of living in many states, including Montana(my home now) and living in Seattle for years.
I just remembered what I call 'Autobahn rules' or 'hedgehopping': ALL traffic stays extreme RIGHT unless passing. This in its basic theory will have the slowest drivers in the right lane, with the next speed strata 'hedgehopping' them. Then the next speed strata will be passing the faster mid traffic, and merging back extreme right if possible.
Imagine that the right-lane right wall is viscous and the left-lane left wall is teflon. Theoretically the inevitable faster traffic will 'slip' over the doofusesXXXXXX, er, I mean slower drivers. Signalling your intentions is considerably important, and courteous.
I have long believed that speed laws exist for revenueing purposes alone. I remember when I-5(Interstate 5) was 55mph; Now it is 75mph..did something magically happen to make the highway safer? NOT! If they handed out tickets to incapable slower people who did not stay right, instead of capable faster people who leap-frogged, a huge amount of congestion would be relieved from the highways. Take Oregon I-5 for example: from Salem to Portland(100miles), you generally have 'freeway plugs' infesting EVERY LANE going enough the same speed, so it is very difficult to expedite your trip. Now if everyone 'stayed' right, the inevitable faster people could would clear the highway faster than the inevitable slower people, cause everyone does not drive the same speed.
I could go on for hours about the wisdom gained from YEARS and hundreds of thousands of miles spent driving the western United States, But I am glad Bill and others are voicing the words of reason-Remember: they don't drive bad because they are stupid, just ignorant..
Christian von Delius <email@example.com>
Kalispell, MT ` - Thursday, April 24, 2008 at 20:17:00 (PDT)
Thank you for putting together this layman-friendly website! I had started a Care2 group, Drive Aware for A Safer Traffic Environment as a resource for information, advice and community-action encouragement. I have now delightfully mentioned your site as a place to go for good information!
I have always had similar theories about traffic flow and congestion. I had even experienced the ire of another driver behind me because I was letting people merge ahead of me in a road construction area (about 2 blocks ahead of the CVS where this driver ended up going, how impatient is that?). Although it is difficult to sell someone on an attitude adjustment, the rational, pragmatic advice here can make sense to just about anybody. I wish I could leave flyers about this stuff on cars in parking lots everywhere I go...
Have you ever visited John Farlam's Smart Driving website? He's another gem of driving widsom: http://www.smartdriving.co.uk/Driving/DefensiveDriving/Space/Space.htm
Here is my group website: http://www.care2.com/c2c/group/drive_aware
There are dozens of sites like mine, but I don't believe there is such a thing as "too many resources" with something like this.
Thanks again, Mr. Beatty!
Malee Holland <firstname.lastname@example.org>
Orlando, FL USA - Sunday, April 06, 2008 at 07:04:28 (PDT)
I already new this by instinct. I have been driving this way from the start (since 1976). Courtesy always pays.
Lynne Calvin <email@example.com>
Harrisonville, MO USA - Wednesday, April 02, 2008 at 08:29:56 (PDT)
Fascinating stuff. I've observed similar dynamics in my travels, and have experimented a bit, though usually with the goal of improving my gas mileage.
I read the FAQ about a particular spot on a hill in Seattle-- sometimes I think it's just a hill itself that can create a wave or jam-- people slow down slightly without noticing as they go up a hill, creating bunching at the bottom. (An example that some readers in Boston might be familiar with is on 128 travelling North leaving Needham, before your get to route 9).
Another aspect that I don't see explored is the effect of external constraints on a highway jam. For example, a traffic light at the begining of an on ramp or end of an off ramp. Or, an off ramp that then merges with a street that has heavy uniform traffic with close spaced cars that the cars on the off ramp can't merge with. For example, on my current commute home a traffic light controls the entrance of traffic to an on ramp that then merges onto a small highway. So the on ramp releases a big clump of cars, that have insufficient gaps to merge onto the traffic already on the highway. This slows down the cars travelling up the on-ramp (which is already an uphill climb) as well as creating a jam right at the onramp (which is also uphill, which doesn't help).
Anyway, just two situations I'm familiar with. We could probably develop a systematic index of all these factors, and look at the results of their combinations on flow and what kinds of behavior changes might affect them...
Reed Hedges <firstname.lastname@example.org>
NH USA - Wednesday, March 26, 2008 at 13:28:52 (PDT)
What's best to do in this situation?
On a two lane divided highway, my lane is ending ahead. However, the cars in the continuing lane are bumper to bumper
moving very slowly, crawling. Should I slow down (almost stop) and merge immediately or should I continue forward as far as I can go then merge. I usually do the latter; however I find that people in the continuing lane get upset, I'm sure they feel that I am cheating to get to the front of the line.
Ottawa, ON Canada - Monday, March 24, 2008 at 08:52:00 (PDT)
How about adding a webpage to this site, organized by city, with suggested places to try the anti-traffic technique? You've mentioned a couple of spots in Seattle... I'd like to say that, in Austin, Northbound IH-35 at the 183 North exit is a great place to observe the effects of Big Empty Spaces. There's traffic merging left onto IH-35 from the ramp just north of 290/Koenig, and traffic trying to get right so they can exit to 183N. The result is usually a horrible jam in the morning. I'm pleased to say that by decreasing my speed a tiny fraction (foot off gas, nothing as drastic as braking) and letting people merge in front of me from both sides, I have singlehandedly broken up jams half-a-mile in length. Any other Austinites practicing BES? Seems like MoPac near downtown and Loop 360 would be good candidates. Oh, and southbound MoPac around 45th/35th, which is prone to those Mystery Jams, where you think there's an accident ahead but when you get there... there's absolutely nothing and people zoom off!
To email, add my first name to the addy given.
Kathy <grace at austin dot rr dot com>
Austin, TX USA - Monday, March 24, 2008 at 08:06:48 (PDT)
Great explanation and it matches my experiences also. I like your idea about merging and will adopt your suggestions and see how it goes.
Albuq, NM USA - Wednesday, March 05, 2008 at 08:02:36 (PST)
In Oklahoma they have passed a law that says that you have to merge as soon as you first see the signs that a narrowing will occur ahead of you. This is to prevent jerks from intentionally zooming along in a lane that they know will close then bulling their way into the open lane.
I'm curious if this will help create more zipper merges or if it will just move the standing wave further up the highway.
[I think the empty lane is the only thing that lets people cheat. With a law that creates empty lanes, you'd need a cop at every merge zone if you wanted to stop cheaters. Since that won't happen, I think the law is creating traffic jams. If instead it became illegal to block merging cars, and the law recommended wide spaces and "let one in ahead" merges, then Zippers would be created on purpose. -billb]
Stillwater, OK USA - Friday, February 29, 2008 at 20:40:00 (PST)
This is not a question ;)
I'm glad I found your article! I'm a wave-smoother myself, but i couldn't convince any of my friends/relatives of it's benefits. Everytime I miss a greenlight, the passanger says that it was because i was not rushing after the empty space ahead of me.
I became a wave-smoother after a fluid dynamics class. Considering some ideal conditions, you can easily show that the flux rate of a fluid is constant at all times - everything that enters must come out.
This helped me in the following aspect: if i'm driving in a 60km/h*2-lanes road, and one road gets blocked, the only way the flux stays the same is by driving at 120km/h in the available lane (which's not likely). What?s usually happens is that the flux is now half (60km/h*1-lane flux) so there?s no way but slowing down to 30km/h, before the block, (30km/h*2-lanes = 60km/h*1-lane).
Simple, but people find it hard to apply!
My 2 cents.
Brazil - Thursday, February 28, 2008 at 07:28:58 (PST)
Just want to tell how interesting are your articles-of-life:) It's 4:07AM, had to go sleep but found your article... and now reading them non-stop...
Thanks for such work.
Vahan A. <email@example.com>
Los Angeles, CA USA - Thursday, February 14, 2008 at 04:08:20 (PST)
I agree with (and practice) almost everything you suggest, with one exception -- the "cheaters" in merge situations DO need to be punished. Otherwise, you create an environment that will simply breed more cheaters.
[ True, but if we don't block cheaters, then soon everyone will start "cheating," and the empty lane will fill up. Does that mean *all* of them cheaters? Of course not, since both lanes are now going the same speed. Nobody can race ahead when the lane is filled! I conclude that "cheating" only arises: 1. when everyone merges early, 2. then they start driving bumper-to-bumper and refusing to let anyone merge early or late (or merge at all, ever.) At jammed exits or lanes which are ending, the traffic jams appear when people in the "through lanes" try to block merging cars. If instead I open up a huge space and let all the cheaters merge ahead of me, then no cheaters are racing to the far end and bulling their way in. And then there is no reason for anyone to drive slow. And that's what busts the traffic jam. Yes, that's right: you can shatter a merging-lanes traffic jam by letting lots of cheaters jump in ahead of you. I've done it many times, and it's stunning to watch the entire backup suddenly evaporate. There are two places in Seattle where this works amazingly well. -billb]
I always allow people to merge into my lane for exiting, until it gets to be "too close" (which is admittedly a matter of judgment) at which point I'll absolutely refuse to allow people to merge. I'm not thinking "just merge behind me" but rather "you waited too long, you have to skip this exit and loop back."
Having grown up in NJ (where this type of practice is quite common -- we HATE "cheaters" there) I know that this type of behavior eventually discourages cheating to the point that it diminishes. When I moved to the Bay Area I was astonished at how people would let others merge -- even across multiple lanes of traffic -- and was constantly irritated by the backups this would cause near every single exit, even when I was in the through lanes. There is NEVER a backup in the through lanes in NJ, no matter how much traffic there is at the exit, because everybody knows better than to try to cheat and cut across at the last minute.
I think this is probably why people driving through NJ consider drivers there to be complete jerks, but it's really just adaptive behavior for a region of the country that's had to deal with high levels of congestion for much longer than anywhere else.
Bill Clark <firstname.lastname@example.org>
Oakland, CA USA - Tuesday, January 08, 2008 at 08:57:40 (PST)
I take a weekly trip north on I-93 and for a while got stuck in traffic every week. Like you, it gave me plenty of time to watch the patterns of traffic around me. Where pinch points happen and where the waves of traffic seem to emanate. Picturing it in my head and creating "rules" allowed me to observe how my own actions can affect other drivers behind me. (And sometime around me too)
Eventually I began to practice the same scenarios that you describe. Slowing down gradually to create a gap in between my car and the car ahead of me that would be caught in the jamb first. I certainly noticed a difference after practicing a few weeks in a row.
I began to notice that drivers behind me would start to emulate my behavior. They would stop changing lanes often and remain in their spot. The biggest kick i got out of the whole experiment was watching aggressive drivers fly up to the jam weave through a few people and then get stuck as I coast by. (This was not always the case but in most instances I found that I would maintain a semi-permanent position in relation to aggressive drivers.) It was also a thrill to look to my right (I am usually in the far left lane) and see that people around me were doing the same thing. Within a 20-yard radius the cars were behaving in a similar manner, eating up the waves that we came up against.
I just want to add a few notes from my own experience.
1. I believe it is easier to "eat" traffic waves in a car with a standard transmission.
a)Breaking automatically alerts the person behind you that you are slowing down. In a standard you can slow down without breaking by down-shifting so that the person behind you must become conscious of their own speed and position within the group.
b) Shifting gears and attributing certain speeds to what gear you are in helps you identify the severity of the traffic jam and the period of each traffic wave
2. You mentioned that at first you thought it was a philosophical solution, ie maintaining a "cool-head" and driving with patience, but later abandoning that reasoning and adopting a more scientific view of upholding certain driving habits. I would counter, and say that it is both.
a) I don't think one is capable of doing one without the other. If I had not let go of my own ego and released myself from the recurring negative thoughts that I must be first, and that I must continue to pass people to make headway in traffic, i would not have been able to observe the laws that were governing the traffic jam. Thus, I would not have been able to improve my driving and (since conscious driving alleviates the traffic for those behind me) the drive for those other people with whom i share the road.
b) if you "project" your good driving behavior you cause other drivers to become aware of their own driving and, perhaps, improve themselves. Which would cascade into better drivers all around. (we all know Massachusetts needs more of those)
Lastly, I just wanted to say thanks for taking the time to put together your essays and post them publicly. This is a great resource for others. Peace.
Morgan Lawless <email@example.com>
Boston, MA USA - Wednesday, December 26, 2007 at 09:15:09 (PST)