MIT working to eliminate phantom traffic jams [w/VIDEO]
Phantom traffic jam - Click above to watch the video after the break
Seriously, how irritating is it when traffic inexplicably comes to a grinding halt for no apparent reason before starting right back up with no trace of a cause? Not only are these so-called phantom traffic jams the bane of many a commuter's existence, they also waste gas and therefore increase vehicle emissions. Mathematicians from MIT are reportedly studying these phantom traffic jams with an eye toward figuring out how to prevent them in the future.
Of course, there's little that can be done in a timely fashion about the sheer number of cars on the roads, but MIT researchers believe that computer modeling could help them come up with a solution. "We wanted to describe this using a mathematical model similar to that of fluid flow," says Aslan Kasimov, a lecturer in MIT's Department of Mathematics.
MIT's models may lead to the formation of highways with extra lanes where these phantom traffic jams are likely to form. We can only hope. Click past the break for two short videos explaining how these jams can form.
Gallery: Traffic Jams
[Source: WIRED]
Video:
Reader Comments (Page 1 of 5)
Rain 8:22PM (6/18/2009)
You don't need MIT to figure this out its called:
Chinese Fire Drill.
Reply
cyaml 12:01PM (6/21/2009)
No. It's called idiots at the front of the line.
Dan 1:42PM (6/21/2009)
Just like stop lights. when the light turns green. if everyone would movelike a train. you wouldnt be all of the sudden moving at a red light only.
John Rowell 9:05PM (6/18/2009)
If sensors are placed along the highway, and when these phantom traffic jams occur the sensor triggers a sign 3 or 4 miles "upstream" of the traffic jam to reduce the speed limit along that section of road - that should smooth out the speed of traffic and make the traffic jam dissipate within a few minutes, and then the speed restriction can be lifted again.
Reply
Matt 9:23AM (6/19/2009)
An electronic control valve of sorts may work, but it's so hard to get people to obey the speed limit. I took a few classes on traffic design and safety as a civil engineering student, and it usually correlates well with fluid dynamics. The problem is, when there are abrupt changes you get a water hammer effect (you may have heard old plumbing rumble when you turn on or shut off the water quickly, as an example). So, lets say a driver cuts somebody off and that person slams on the brakes in the middle of the interstate. What happens next is a chain reaction of drivers up stream slamming on their brakes until the gap between cars is big enough that heavy breaking is not necessary. The problem in this situation is not 100% the result of the initial bad driver. The big issue is that there is nowhere for the other cars to go; the road is at or even beyond its designed capacity. Generally speaking, merging lanes and sharp turns are the worst places for this kind of scenario, but also abrupt changes in elevations can have similar effects. It's all about "pressure" changes on the road, and it sounds like MIT is headed in the right direction (as usual).
Stew 9:27PM (6/18/2009)
You will never get rid of something like this without getting rid of the human element, which will never happen.
Stew
Reply
GoodCheer 9:19AM (6/19/2009)
Several new cars now come equipped with following distance cruise control (I think some M-Bs and the new Prius have it). I would think if that technology becomes common, it could go a long way towards removing the human element. Figuring out just how prevalent it would have to be to have a significant effect would be an interesting project for MIT.
janni 1:53PM (6/21/2009)
of course it will happen that human's do not drive cars; there are already such cars at airports in japan and on-going trials in the uk
Simon 12:03AM (6/22/2009)
I believe the human will eventually be reomoved from controlling the car eventually. It makes sense and I think will be more popular once people get used to the benefits. How to introduce the technology will be the major hurdle. However, once it is done I believe we will be dealing with vehicles that will be smaller, more efficient and totally automated. In fact, vehicle ownership may even be a thing of the past. Sure, it is unlikely to occur next year - but I see it as inevitable at some stage down the track.
usbseawolf2000 10:23PM (6/18/2009)
The solution is the electric motor. Precise control of torque and instant regenerative braking allow the driver to respond to the traffic better.
The key is to remove the delay from the driver. The fault includes torque converter power delivery lag and hydraulic brakes.
Reply
joegibes 8:00AM (6/19/2009)
The solution is the electric motor. Precise control of torque and instant regenerative braking allow the driver to respond to the traffic better.
The key is to remove the delay from the driver. The fault includes torque converter power delivery lag and hydraulic brakes.
"The key is to remove the delay from the driver."
The delay does not come from the powertrain -- I drive a 12 year old sedan and while its throttle response and brake pedal response could be better (it's no sports car), the biggest delay is my own. Human reaction time is far longer than the control delay in modern cars.
Also, I don't believe that faster responses would help -- reducing traffic jams requres smooth flow, not immediate and jerky motion. It's the go fast/slow down/go fast/slow down mentality that *creates* jams in the first place.
Serge 4:00PM (6/19/2009)
I'd rather see the "driver" removed, at least on the highway. I hope one day to get into my car, input a destination and promptly go take a nap in the back.
Luke 10:58PM (6/18/2009)
I doubt that there are many people who make bad decisions in fluid, so looking at fluid won't help much. People aren't raw elements that move from higher pressure to lower pressure, or simply follow gravity.
In traffic, when everyone is doing what everyone expects, no one applies the brakes. Therefore, we need to look at those who do the unexpected, as well as those who have unrealistic expectations. Or, we can take control of the vehicles and give it over to a central control system.
Reply
ian in hamburg 2:04AM (6/19/2009)
Unless you can somehow rid the human psyche of the fear of dying in a huge ball of flame, you will always have this happening. One foot too close and one mile an hour too fast, people will hit the brakes, and so it goes down the line.
Reply
Alan 4:26AM (6/19/2009)
Exactly, the problem is caused by people driving too close to the car in front, then over braking, which then causes a ripple effect. If people made more effort to maintain an equal distance between cars and reduced speed gently by taking their foot off the gas pedal these problems would be very much reduced.
Reply
Nick 5:10AM (6/19/2009)
To all the naysayers out there, who say this can't be solved, what a bunch of fools.
This CAN be solved with technology. Automatic distance regulators, warning signals, car-interconnectivity etc.. can make a difference.
Reply
Roxanna 10:30AM (6/21/2009)
Yes, let's take driving completely out of the control of the driver! I don't want to live in a world like that. My car already does too many things on its own like lock the doors when I start driving.
Some of us enjoy the driving experience. To those of you who want it all automated ("passengers") - get off the road and take the train. More room for us "drivers"!
Jean 7:08AM (6/19/2009)
There is a simple solution that consist in slowing down a bit more than the slowing traffic ahead of you. It requires drivers' education though.
Reply
wisi 8:57AM (6/19/2009)
I call Bullshit on that one: the one behind you does the same and that gives you the oscillation that causes the Jam...
I don't know what MIT wants to find out. The Problem and the cause has been well known for over 20 years:
More cars require more throughput. That means closer spacing or HIGHER speed. Humans can only handle a certain combination of speed and spacing thus the traffic throughput is limited. When you go over that limit traffic starts oscillating and there you have your phantom jam.
We could up that limit with ACC Systems, but they are anyway not allowed to have closer spacing then we normally use.
The only thing you can do is accept the lower throughput, drive the resulting lower average speed and not induce any disturbances by braking.
GoodCheer 8:26AM (6/19/2009)
But slowing down a bit more than the traffic ahead of you is exactly what causes this to occur. If car A slows from 65 to 64 (even if just for a second), then car B behind it slows from 65 to 63, car C from 65 to 62 etc .... 64 cars back you have to come to a complete stop.
To avoid this, you'd have to slow down a bit _less_ than the traffic ahead of you. This can only be done if you have extra following distance that you can 'take up' for a few seconds when the car ahead of you briefly slows down.