First rule of ant traffic: no overtaking(technologyreview.com)
technologyreview.com
First rule of ant traffic: no overtaking
http://www.technologyreview.com/blog/arxiv/23176/
14 comments
The real difference is that stopping distance for an ant is a fraction of an ant length, but for a car it's several car lengths. So for cars, most of the space is following distance For ants, most of the space is ants. Car density decreases as the square of speed, while ant density is constant. That's not because they have better algorithms but because they're in a different realm of physics.
I agree with you about a major difference between us on motorways and anits is the variance in our speed compared to that of an ants, which I presume all travel at roughly the same speed.
Interestingly, in the UK (and probably lots of other places) we have average speed cameras at road works which limit people to 50mph pretty effectively, this results in everyone travelling the same speed with no overtaking. Anecdotally I've found this to work quite well at keeping the traffic moving.
Interestingly, in the UK (and probably lots of other places) we have average speed cameras at road works which limit people to 50mph pretty effectively, this results in everyone travelling the same speed with no overtaking. Anecdotally I've found this to work quite well at keeping the traffic moving.
>Having sat in long, single lane stop and go traffic I can >anecdotally suggest it's irrelevant.
They're suggesting that no overtaking prevents the bottleneck in the first place, not that it gets you out of an existing one.
Seems like this would prevent drivers behind overtakers from having to slow down which propagates back causing a bottleneck.
How about multi-lane highways with tolls for each lane priced dynamically based on traffic conditions.
They're suggesting that no overtaking prevents the bottleneck in the first place, not that it gets you out of an existing one.
Seems like this would prevent drivers behind overtakers from having to slow down which propagates back causing a bottleneck.
How about multi-lane highways with tolls for each lane priced dynamically based on traffic conditions.
(Warning: improvised English, this has been very hard for me, sorry)
There is a annoying effect that I can see in my city in most of the jams that I suffer. It's about the highway exits. The circular highway is usually three lanes wide. It has exits, many of which are slow (entries to the city). The exits take the form of an additional lane to the right. When the "contact zone" of the rightmost lane with the additional lane is long and the traffic is middle or heavy, the jam is a sure thing.
Why? Even if the exit has a reasonable speed, there are drivers that wait until the last moment to take the additional lane. There are so many that most of the vehicles that reach the exit are of this kind. If you are kind enough to take the additional lane as soon as it is possible, you get caught in a trap for ten to twenty minutes.
It's a moral dilemma. Either I am a good citizen and agree to be victimized for a quarter of an hour daily or I do the same to others. Please, don't ask me what I do.
It also creates a jam for the cars that aren't taking the exit, because the late changes of lane affect them. Not only the right lane, but also the middle one are filled with "late-exit-takers". And the slow cars have to change lanes to avoid the stopped cars waiting to exit.
I've observed that there is a disposition of lanes that prevents the problem: separating two lanes to exit and two lanes to follow on the highway, making the exit lanes long enough (to have a "buffer" for the slowness of the exit) and the fork instant, not a long "contact zone". But this disposition is seldon used and I haven't really made the experiment to be sure :-)
About the article: I don't think the ants are similar to cars at all. Drivers have freedom to behave differently, some are in a real hurry and we have different speed of reaction (to use the holes in the exit to wait to the last moment).
There is a annoying effect that I can see in my city in most of the jams that I suffer. It's about the highway exits. The circular highway is usually three lanes wide. It has exits, many of which are slow (entries to the city). The exits take the form of an additional lane to the right. When the "contact zone" of the rightmost lane with the additional lane is long and the traffic is middle or heavy, the jam is a sure thing.
Why? Even if the exit has a reasonable speed, there are drivers that wait until the last moment to take the additional lane. There are so many that most of the vehicles that reach the exit are of this kind. If you are kind enough to take the additional lane as soon as it is possible, you get caught in a trap for ten to twenty minutes.
It's a moral dilemma. Either I am a good citizen and agree to be victimized for a quarter of an hour daily or I do the same to others. Please, don't ask me what I do.
It also creates a jam for the cars that aren't taking the exit, because the late changes of lane affect them. Not only the right lane, but also the middle one are filled with "late-exit-takers". And the slow cars have to change lanes to avoid the stopped cars waiting to exit.
I've observed that there is a disposition of lanes that prevents the problem: separating two lanes to exit and two lanes to follow on the highway, making the exit lanes long enough (to have a "buffer" for the slowness of the exit) and the fork instant, not a long "contact zone". But this disposition is seldon used and I haven't really made the experiment to be sure :-)
About the article: I don't think the ants are similar to cars at all. Drivers have freedom to behave differently, some are in a real hurry and we have different speed of reaction (to use the holes in the exit to wait to the last moment).
Ants are not trying to text their friends, watch their GPS, and change the radio station while they are in formation.
This article does make a good point. It has been my experience that highway traffic can move more quickly when people do not change lanes. I have seen this happen a couple times in very snowy conditions; the traffic density was such that it should have been stop & go traffic; however, due to the large amount of snow between each lane, everyone pretty much stayed in their lane. This made traffic go at a steady 20-25 mph. (Note that the set of circumstances in which snow can cause traffic to speed up is very rare).
The main problem that this article does not address is bottlenecks. Even with a perfect system, it is impossible to keep a constant speed if there is a large influx of traffic at a certain point or the road narrows to fewer lanes.
The main problem that this article does not address is bottlenecks. Even with a perfect system, it is impossible to keep a constant speed if there is a large influx of traffic at a certain point or the road narrows to fewer lanes.
The article makes several good points but I'd have to disagree with the no overtaking rule. Taking the human out of the equation, vis a vis 'IRobot' style, where road travel is managed by computer seems like a more viable option in the long term. There have been several advances in this technology, from MIT especially (Google 'Self driving car', do you really expect me to divert from this tab for you?)
I work in Highway Engineering and Management, the company i work for manages quite a few of the UK's motorways. I wonder if I could wing this point into a meeting. Another Gem HN, thanks.
I work in Highway Engineering and Management, the company i work for manages quite a few of the UK's motorways. I wonder if I could wing this point into a meeting. Another Gem HN, thanks.
The more I think about this the more I think the comparison is not relevant. Ants are all going from point A (the colony) to point B (the food). Additional ants are not joining the line midway, nor are any splitting off to some other destination, nor do different streams of ants cross one another's paths.
If you had a one-way one-lane road with no cross streets, merges, or exits, MAYBE the situation would be comparable; in the real world, I don't see a whole lot of relevance.
If you had a one-way one-lane road with no cross streets, merges, or exits, MAYBE the situation would be comparable; in the real world, I don't see a whole lot of relevance.
In "Critical Mass" by Philip Ball chapter 7 covers "The inexorable dynamics of traffic" which is highly worth a read if this subject matter interests you. There's also some very interesting sections on crowd dynamics and route finding.
More importantly, there is a big difference between personal locomotion and commanding a vehicle. For example, the difference between my walking pace and that of an Olympic champion sprinter is not THAT great. However, the difference between my freeway merge speed and that of my mom is an order of magnitude. Add in fixed and inflexible travel lanes, the ability to speed up and slow down consciously and unconsciously, traveling at speeds that will knowingly cause death in the event of a problem, etc. and you have a lot more variance with higher stakes to the participants.
I'm not saying it isn't a fascinating bit of trivia, but I think Tom Vanderbilt's book "Traffic" is apt to lead to more insight than this particular nugget of knowledge.