Traffic hypothesis

[Steve81]

MIT did a study a while back regarding tailgaiting and traffic jams. I can’t seem to find the actual report, but it appears looking at the graphic that the analysis was confined to a single lane.

My hypothesis is that this effect is compounded in a multi-lane scenario, based on my observations driving in DC traffic. An additional source of trouble here may be people that are obligated to change lanes, but have to wait for a gap in traffic, thus slowing down their lane. The tendency of people to try and switch lanes to the faster moving lane may also exacerbate the issue.

I would also expect that a sudden unsafe lane change (cutting someone off) could exhibit similar effects, as the driver being cut off is forced to brake, sometimes significantly so.

[OldChemE]

I read your post-- but couldn't find any mention of the hypothesis.  What are  you trying to say?  My own observation from many years of commuting in an area subject to slow traffic is that the fastest route is usually the lane that is obstructed or being eliminated (as in three lanes going to 2).  There are always a certain percentage of drivers who are early adopters and change to the open lane early, leaving the others to wait and (politely) merge at the end.  Fewer vehicles in the lane that will be ending leads to faster progress in that lane.

 

There is also the matter of distance between cars.  If you calculate the number of cars passing a particular point at, say 20 miles per hour, the rate at which cars pass the point is inversely proportional to the gaps between the vehicles.  This argues in favor of keeping as close to the vehicle in front as is safe.

[studiot]

4 hours ago, OldChemE said:

I read your post-- but couldn't find any mention of the hypothesis

I agree. +1

More rigour please, Steve.

[exchemist]

On 8/13/2023 at 4:23 AM, Steve81 said:

MIT did a study a while back regarding tailgaiting and traffic jams. I can’t seem to find the actual report, but it appears looking at the graphic that the analysis was confined to a single lane.

My hypothesis is that this effect is compounded in a multi-lane scenario, based on my observations driving in DC traffic. An additional source of trouble here may be people that are obligated to change lanes, but have to wait for a gap in traffic, thus slowing down their lane. The tendency of people to try and switch lanes to the faster moving lane may also exacerbate the issue.

I would also expect that a sudden unsafe lane change (cutting someone off) could exhibit similar effects, as the driver being cut off is forced to brake, sometimes significantly so.

This sounds a bit like the kinetic theory explanation for why the viscosity of gases increases with temperature.🙂

[Steve81]

1 hour ago, studiot said:

I agree. +1

More rigour please, Steve.

Sorry, I’m still obviously new at this. If you could point me in the direction of a hypothesis you feel is particularly well written / in the format you’re looking for, I can aim to use that as a template.
 

Edited August 14, 2023 by Steve81

[studiot]

29 minutes ago, Steve81 said:

Sorry, I’m still obviously new at this. If you could point me in the direction of a hypothesis you feel is particularly well written / in the format you’re looking for, I can aim to use that as a template.
 

This is the point where I invite @iNow  to step in with their famous cartoon.

[exchemist]

54 minutes ago, Steve81 said:

Sorry, I’m still obviously new at this. If you could point me in the direction of a hypothesis you feel is particularly well written / in the format you’re looking for, I can aim to use that as a template.
 

Well you refer to an "effect"  but without describing it. Just describe what the effect is on which you are building your hypothesis. (One of the forum rules is people should be able to follow a discussion without being sent off-site to other links. So a précis of what the MIT link says should do the trick.)  

[Sensei]

The pipe has a limit on the amount of liquid that can flow through it in a given period of time. It's the same with a road or highway.

You can take into account the size of the car, multiply by the average number of passengers, average speed, etc. and get the theoretical maximum number of people who can travel on a given road.

I always say that a single bus is (or can be) like "half a kilometer of road full of cars". It can have 50 up to 300 passengers (equivalent of 300 cars with just driver, or 150 cars with driver and passenger).

Waiting for Uberbus app for smartphones..

The traditional single-ticket system forces people to get on a bus that goes to their destination, when they could get on any device that takes them closer to that destination. As a result, buses (or other transport devices) are half empty (i.e. not optimally used).

[Phi for All]

I couldn't find it quickly this morning, but a few years back there was a meta-study on traffic which concluded that virtually ALL modern traffic was caused by brake lights. You see people brake ahead of you and back off the gas, or step on your own brakes, continuing the signal back down the line to remove energy from the system. And most of the brake lights could be avoided if folks backed off and stopped tailgating, which would allow for more merging to smooth the flow.

MY hypothesis is that if more people practiced cooperative driving rather than competitive driving, we'd all make it to work/home/wherever more consistently.

[iNow]

2 hours ago, Steve81 said:

Sorry, I’m still obviously new at this. If you could point me in the direction of a hypothesis you feel is particularly well written / in the format you’re looking for, I can aim to use that as a template.

1 hour ago, studiot said:

This is the point where I invite @iNow  to step in with their famous cartoon.

Happy to oblige. Feel free to use the Bat-signal next time you need me 

[Steve81]

2 minutes ago, iNow said:

Happy to oblige. Feel free to use the Bat-signal next time you need me 

LOL

[exchemist]

23 minutes ago, Phi for All said:

I couldn't find it quickly this morning, but a few years back there was a meta-study on traffic which concluded that virtually ALL modern traffic was caused by brake lights. You see people brake ahead of you and back off the gas, or step on your own brakes, continuing the signal back down the line to remove energy from the system. And most of the brake lights could be avoided if folks backed off and stopped tailgating, which would allow for more merging to smooth the flow.

MY hypothesis is that if more people practiced cooperative driving rather than competitive driving, we'd all make it to work/home/wherever more consistently.

I’ve certainly read that if a certain distance is not maintained, one vehicle braking will cause the one behind to brake more sharply, due to human reaction time, and several vehicles back you have them resorting to a full emergency stop in order to prevent collision. So the braking wave is progressively compressed, eventually into what amounts to a shock wave. But I’d like to see a summary of the effect referred to in the OP, as I’m still not clear what we are discussing.

[Steve81]

1 hour ago, Phi for All said:

I couldn't find it quickly this morning, but a few years back there was a meta-study on traffic which concluded that virtually ALL modern traffic was caused by brake lights. You see people brake ahead of you and back off the gas, or step on your own brakes, continuing the signal back down the line to remove energy from the system. And most of the brake lights could be avoided if folks backed off and stopped tailgating, which would allow for more merging to smooth the flow.

MY hypothesis is that if more people practiced cooperative driving rather than competitive driving, we'd all make it to work/home/wherever more consistently.

That's exactly it. Thank you for better explaining what I was trying to say. Also noted for what you're looking for with a hypothesis.

Edited August 14, 2023 by Steve81

[Phi for All]

26 minutes ago, Steve81 said:

That's exactly it. Thank you for better explaining what I was trying to say. Also noted for what you're looking for with a hypothesis.

Can I make it a little stronger and more focused? The MIT study you cited spotlights tailgating, so that should be called out in any hypothesis about it.

How about: The use of cooperative driving techniques combined with anti-tailgating technology will improve traffic and result in fewer accidents and delays.

The study talks about the sensor technology available even now. I recently rented a big vehicle equipped with cameras all around, and alarms linked to proximity sensors. The car warned me of vehicles approaching on the sides, and showed a little green "car" when I was the proper distance from the vehicle in front of me. A couple of times the car thought I came up on another car too quickly, and lots of lights and alarms flashed. I don't know if the car had the ability to initiate the brakes on its own, something I'm still on the fence about. It's anecdotal, but I think this is evidence in support of how following too closely causes more traffic, and that changing our behavior and also adopting new technology can improve the situation.

[Steve81]

4 minutes ago, Phi for All said:

Can I make it a little stronger and more focused? The MIT study you cited spotlights tailgating, so that should be called out in any hypothesis about it.

How about: The use of cooperative driving techniques combined with anti-tailgating technology will improve traffic and result in fewer accidents and delays.

The study talks about the sensor technology available even now. I recently rented a big vehicle equipped with cameras all around, and alarms linked to proximity sensors. The car warned me of vehicles approaching on the sides, and showed a little green "car" when I was the proper distance from the vehicle in front of me. A couple of times the car thought I came up on another car too quickly, and lots of lights and alarms flashed. I don't know if the car had the ability to initiate the brakes on its own, something I'm still on the fence about. It's anecdotal, but I think this is evidence in support of how following too closely causes more traffic, and that changing our behavior and also adopting new technology can improve the situation.

Surely. I'm unopposed to such improvements.

[Phi for All]

On 8/12/2023 at 9:23 PM, Steve81 said:

I would also expect that a sudden unsafe lane change (cutting someone off) could exhibit similar effects, as the driver being cut off is forced to brake, sometimes significantly so.

Most of the behavior that needs to change hinges on simple compassion, like letting others merge, but I think we need to train new drivers about what really goes on in traffic. We all know people who consider themselves to be stellar drivers, yet they make aggressive lane changes, tailgate, and constantly move to the lane they feel is moving the fastest. They believe they're driving better than everyone else, and that's going to be hard to change. How do you convince someone who's a Fast & Furious franchise addict that their bold driving styles are causing the very traffic they're trying to overcome?

Another part of educating new drivers is getting them to understand about defensive driving. I cringe every day looking at cars following each other so closely. Those folks believe their reflexes are fast enough to match the braking of the car in front of them, and they might be if that car is only slowing down. If the car in front hits the brakes seriously, there's nothing to be done. Physics is a bitch, and I think most people forget that.

[Steve81]

20 minutes ago, Phi for All said:

Most of the behavior that needs to change hinges on simple compassion, like letting others merge, but I think we need to train new drivers about what really goes on in traffic. We all know people who consider themselves to be stellar drivers, yet they make aggressive lane changes, tailgate, and constantly move to the lane they feel is moving the fastest. They believe they're driving better than everyone else, and that's going to be hard to change. How do you convince someone who's a Fast & Furious franchise addict that their bold driving styles are causing the very traffic they're trying to overcome?

Another part of educating new drivers is getting them to understand about defensive driving. I cringe every day looking at cars following each other so closely. Those folks believe their reflexes are fast enough to match the braking of the car in front of them, and they might be if that car is only slowing down. If the car in front hits the brakes seriously, there's nothing to be done. Physics is a bitch, and I think most people forget that.

I concur 100% with your thoughts. From my observations, changing human behavior is a bitch as well, as you allude to. The only fix I can think of is education, and regular law enforcement crackdowns on aggressive drivers. A technological solution would be autonomous vehicles, or a modification of current crash-detection sensors to automatically brake if you're following too closely. These would take time to implement on a large scale of course.

Edited August 14, 2023 by Steve81

[TheVat]

It is hard to change, partly owing to driving really being a group activity - goal:  getting everyone safely to their destination - while many drivers want to play a zero-sum game of some kind where they "win" something.  Driving should be ballet, not battle.  One thing that I find mildly annoying is the way a long line of cars responds to a green light.  If we were all really focused on cooperation, we would all start rolling forward when the light turns green.  But, as we know, that is not what happens.  Instead, each driver waits for the car in front of them to start moving before they do.  (and given how tightly drivers pack themselves together at a red light, the optimal All Roll Together method would be difficult to achieve even if drivers were all attentive)  IOW, most drivers don't respond to the larger situation.  We're not fish, we don't coordinate our movements as a school of fish does.  

My overall answer, which is somewhat OT, is usually "mass transit," since I don't see most cities as really designed for huge numbers of private vehicles.  We try to do that in the US, but the results are pretty awful.  

[swansont]

1 hour ago, Phi for All said:

Most of the behavior that needs to change hinges on simple compassion, like letting others merge, but I think we need to train new drivers about what really goes on in traffic. We all know people who consider themselves to be stellar drivers, yet they make aggressive lane changes, tailgate, and constantly move to the lane they feel is moving the fastest. They believe they're driving better than everyone else, and that's going to be hard to change. How do you convince someone who's a Fast & Furious franchise addict that their bold driving styles are causing the very traffic they're trying to overcome?

The traffic I’ve dealt with for the last 25+ years (and deal with a lot less, recently) highlight these “A” personality drivers, who act like the rules don’t apply to them. Ignoring the dotted lines dividing the lanes, crossing a lane or two to turn or exit. And forcing themselves into a lane instead of getting in line. I’ve wondered if it’s due to the higher density of lawyers and their ilk (be they practicing attorneys, lobbyists or whatever). There’s also the problem of the folks with diplomatic immunity, possibly not caring too much about following the traffic laws. 

(Also the incompetent - coming to a stop on the onramp!)

[Phi for All]

17 minutes ago, TheVat said:

It is hard to change, partly owing to driving really being a group activity - goal:  getting everyone safely to their destination - while many drivers want to play a zero-sum game of some kind where they "win" something.  Driving should be ballet, not battle.

That's the bumper sticker version (and I love it!). It really should be equated with one of those fancy ball dances where there are no divas, and the whole idea is that everybody works together to make sure all do the same dance well and nobody collides. 

17 minutes ago, TheVat said:

One thing that I find mildly annoying is the way a long line of cars responds to a green light.  If we were all really focused on cooperation, we would all start rolling forward when the light turns green.  But, as we know, that is not what happens.  Instead, each driver waits for the car in front of them to start moving before they do.  (and given how tightly drivers pack themselves together at a red light, the optimal All Roll Together method would be difficult to achieve even if drivers were all attentive)  IOW, most drivers don't respond to the larger situation.  We're not fish, we don't coordinate our movements as a school of fish does.

Have they ever tried this with driverless cars? If they were networked, this could be part of their programming. I wonder if we aren't headed in that direction.

I used to commute back home on a 3 lane, one-way street with lots of traffic lights, and learned quickly that they were timed for about 32 mph, meaning you'd hit every light green after the first one if you kept it at 32. Unfortunately, so many people wanted to race to be first, and would inevitably get stopped by the red light. Only a few times in about 3 years did I find myself surrounded by others who'd figured it out too. We all cruised along at 32, making all the lights, with no annoying leadfoots stopped at a light that's about to turn green. Good times.

11 minutes ago, swansont said:

The traffic I’ve dealt with for the last 25+ years (and deal with a lot less, recently) highlight these “A” personality drivers, who act like the rules don’t apply to them. Ignoring the dotted lines dividing the lanes, crossing a lane or two to turn or exit. I’ve wondered if it’s due to the higher density of lawyers and their ilk (be they practicing attorneys, lobbyists or whatever). There’s also the problem of the folks with diplomatic immunity, possibly not caring too much about following the traffic laws. 

I've heard traffic in DC can be a nightmare. All those people making laws, then get in their cars and break those laws. A cop friend once told me that lawyers were the worst people to pull over for traffic violations, for various reasons.

The movies haven't helped. Toxic driving = cool stunts. Following the laws and speed limits have become a meme for "boring, basic, brainless". The idea seems to be that, if you're moving fast enough, they won't see your turn signals until it's too late anyway, so why use them? Making others eat your dust seems to appeal to those with a more competitive than compassionate outlook, and it's the plot of too many of our stories.

[CharonY]

There are a lot of papers out there on phantom traffic jams (i.e. jams by folks breaking which then perpetuates down the line)- a cursory search showed over 1k publications (at the very least). There are also many papers promoting suggestions, which mostly seem to focus on a collaborative system and/or technical devices that can simulate or promote such behaviour.

However, the most interesting paper on traffic jam avoidance to me is this one here, for perhaps obvious reasons:

https://elifesciences.org/articles/48945

 

 

[swansont]

53 minutes ago, Phi for All said:

I've heard traffic in DC can be a nightmare. All those people making laws, then get in their cars and break those laws. A cop friend once told me that lawyers were the worst people to pull over for traffic violations, for various reasons.

Nightmare, yes, but actual lawmakers are relatively few in number, and most don’t drive themselves around. 

[studiot]

3 hours ago, Phi for All said:

Most of the behavior that needs to change hinges on simple compassion, like letting others merge,

Yes I agree

I learned my driving in London and I used to say that it was a greater pleasure to drive there than any other city ssince the drivers had long learned the value of cooperation in the pohilosophy of

"Let one go"

There were so many interconnecting side streets in London that drivers would spread out and know that one car at a junction would automatically be allowed into the stream, because the other drivers would not have very far to go before they would also need to make the selfsame manouver.
Sadly many connections have been blocked and I note a corresponding deterioration in driving manners and competence.

2 hours ago, TheVat said:

One thing that I find mildly annoying is the way a long line of cars responds to a green light.  If we were all really focused on cooperation, we would all start rolling forward when the light turns green.  But, as we know, that is not what happens.  Instead, each driver waits for the car in front of them to start moving before they do.  (and given how tightly drivers pack themselves together at a red light, the optimal All Roll Together method would be difficult to achieve even if drivers were all attentive) 

I don't know what the regulations are in America but in the UK that strategy would be technically be illegal and a cause of driving test failure.

In the UK a green light does not mean 'go'.
It means @You may proceed for caution if your way is clear.'

Drivers have become so bad at this that for some years we have had yellow boxes hatched onto the road.
It is illegal to enter such a box 'unless your  exit is clear' Stopping in such a box is alo illegal and many have been prosecuted for this.
The problem is that too many drivers at junctions enter such zones in the middle of the junction when there is clearly already a jam at their exit.
The result is that no one else can use the junctio  and gridlock ensues.

 

I don't know if you have such regulations in America ?