Going postal in the security line

[kallend 2,113]

Interesting read.

www.avweb.com/blogs/insider/AvWebInsider_FatCaptain_201987-1.html

...

The only sure way to survive a canopy collision is not to have one.

[JohnRich 4]

What's interesting about it?

[rhaig 0]

looks like someone complaining (whining) about air travelers and crew not thinking of other people first when traveling.

wah!
a business traveler knows all these things that were whined about. get on your flight earlier, you'll have overhead space. get to the airport a bit earlier, you'll not care who's taking an extra 3min in security line.

I was hoping for a news story about someone going postal in the security line. Not a whiny avweb article.

--
Rob

[mnealtx 0]

Quote

looks like someone complaining (whining) about air travelers and crew not thinking of other people first when traveling.

wah!
a business traveler knows all these things that were whined about. get on your flight earlier, you'll have overhead space. get to the airport a bit earlier, you'll not care who's taking an extra 3min in security line.

I was hoping for a news story about someone going postal in the security line. Not a whiny avweb article.

Flight crew should damn well know those things too, don't you think?

Mike
I love you, Shannon and Jim.
POPS 9708 , SCR 14706

[Andrewwhyte 1]

Quote

And last, the merge thing. You serve the freedom of the commons more efficiently by merging as late as possible, because even though you whiz by the early mergers, you're using all the available pavement for longer, so everyone moves faster.

Damm I wish people would learn that! Never merge before the person in front of you. Merge exactly one vehicle behind the person in front of you.

[rehmwa 2]

Quote

Quote

And last, the merge thing. You serve the freedom of the commons more efficiently by merging as late as possible, because even though you whiz by the early mergers, you're using all the available pavement for longer, so everyone moves faster.

Damm I wish people would learn that! Never merge before the person in front of you. Merge exactly one vehicle behind the person in front of you.

agreed - early mergers 'think' they are being courteous so I'm not torqued at them, just frustrated they don't see the logic. But it would also be nice to see people use all the pavement and at the merge point at the end courteously allow alternating cars in (courteous try to get into the alternating sequence)


you're using all the available pavement for longer, so everyone moves faster

conservation of mass, professor - you're just moving the same choke point downstream an insignificant distance, so everyone moves the same. The benefit is getting people away from the on ramp and into a place where it's easier to merge and let's the ramp have a bit more margin to transient loading.

...
Driving is a one dimensional activity - a monkey can do it - being proud of your driving abilities is like being proud of being able to put on pants

[billvon 3,085]

>I was hoping for a news story about someone going postal in the
>security line. Not a whiny avweb article.

Thus the old saying - "if it bleeds it leads." Had this story had a little blood in it, heck, USA Today would already have picked it up.

[billvon 3,085]

>And last, the merge thing. You serve the freedom of the commons more
>efficiently by merging as late as possible, because even though you whiz by
>the early mergers, you're using all the available pavement for longer, so
>everyone moves faster.

No, the mergers move faster, the people who are being merged in front of move slower. Average stays the same, because the average is set by traffic in _front_ of the entrance ramp. However, if filling up that merging lane helps get some traffic away from the nearest traffic light, then it might help traffic on the surface streets leading to the freeway.

(The above is not valid when traffic is moving, of course; the acceleration lane then becomes just that rather than a "place to put more cars.")

[SkyDekker 1,465]

I think it is pretty clear that traffic merges faster if it follows the pattern of a zipper, specially in heavy traffic.

Stay in the lane until it ends, move over.

[billvon 3,085]

>I think it is pretty clear that traffic merges faster if it follows the pattern of a
>zipper, specially in heavy traffic.

Again, in heavy traffic, what sets the speed is the traffic _ahead_ of the merge. Whether the merge happens at the beginning of the acceleration lane or at the end doesn't matter.

It's a simple problem. There are two inputs (highway traffic and entering traffic) and one output (the traffic past that point.) When traffic is very heavy (i.e. bumper to bumper, often stopping) the limit is set by the output, and no trick you can do will change that. When traffic is more moderate (i.e. moving slowly, at say 20mph, but still tight) then the best you can do is slow both inputs down by some ratio to maintain the speed past that point.

If traffic is moving well (i.e. moving constantly at some speed above 20mph) then the ideal is to have a patterned merge based on the traffic. The "zipper" thing only works if the highway lane is carrying as much traffic as the merging lane, and that's rare. Most of the time the merging lane is carrying less traffic, so an ideal pattern will be something like 3 cars highway - 1 car merge lane - 3 cars highway etc. (Traffic flow-control lights help set such ratios to help maintain highway flow.) It doesn't really matter where this happens in the lane as long as it happens before the end of the lane. All "filling up the lane" does is get more cars on open pavement; it doesn't increase the capacity of the highway or intersection.

If traffic is completely free (i.e. moving at highway speeds) then most of the above doesn't matter, and the lane is used to accelerate.

[kelpdiver 2]

Quote

I think it is pretty clear that traffic merges faster if it follows the pattern of a zipper, specially in heavy traffic.

Stay in the lane until it ends, move over.

Nothing clear at all about it. But it sounds like good rationalization for the assholes who want to use the dying lane to pass.

I-5 between NorCal and LA is a good example of a higher speed traffic flow with frequent drops to one lane (due to slower moving trucks). Because of the passers/late mergers, the junctions become compression points where everyone has to brake and readjust their speeds to give a semblance of following distances. Traffic flow would be more even if people merged sooner, not later. (and worse, people don't do proper following distances, so accidents occur)

In slow moving traffic, the late mergers cause more trouble as their action forces the entire moving lane to stop and go frequently, the most costly action as drivers are not coordinated - must wait for the car in front to move before they can initiate.

[rehmwa 2]

Quote

>No, the mergers move faster, the people who are being merged in front of move slower. Average stays the same, because the average is set by traffic in _front_ of the entrance ramp.

not "in front of" but "past the final merge opportunity" - that's where the limiting pathway is.

but I agree with you - it's a mass flow calculation

...
Driving is a one dimensional activity - a monkey can do it - being proud of your driving abilities is like being proud of being able to put on pants

[rehmwa 2]

Quote

it sounds like good rationalization for the assholes who want to use the dying lane to pass

nonsense, if we filled up that dying lane and everybody merged at the merge point instead of earlier - then the assholes CAN'T pass - they'd all have to wait their turn

...
Driving is a one dimensional activity - a monkey can do it - being proud of your driving abilities is like being proud of being able to put on pants

[SkyDekker 1,465]

Quote

Again, in heavy traffic, what sets the speed is the traffic _ahead_ of the merge. Whether the merge happens at the beginning of the acceleration lane or at the end doesn't matter.

Absolutely.

What I meant was that traffic would move faster if all would drive to the end of the ending lane (keep in mind this does not have to be an on ramp) and then be allowed to merge.

If both lanes are heavy, the zipper (one for one) works fastest, and in my world that scenario isn't that rare.

What holds up traffic is the idiots that want to merge prior to the ending of the lane, now stopping traffic behind them etc.

[SkyDekker 1,465]

Quote

In slow moving traffic, the late mergers cause more trouble as their action forces the entire moving lane to stop and go frequently, the most costly action as drivers are not coordinated - must wait for the car in front to move before they can initiate.

Nonsense. The stopping occurs because people don't leave room for other's to merge. Riding some body's bumper to make sure they cannot get in front of you. It is that behaviour that actually slows traffic.

Filling up the dying lane with traffic actually speeds traffic up. Assuming people can be polite enough to allow people to merge.

[Tokter 0]

I never understood the early mergers, an invisible line must exist for them. If you merge after that line you're an asshole if you merge before the line you're courteous.
If you have a couple of early mergers in front of you then that lines keeps moving further and further away from the actual merge point.
Until an asshole resets the line then whole game starts over...

[rhaig 0]

the guys that tick me off are the ones that race up past everyone else (sometimes on some shoulder) and expect me to let them in.

I let people in who merge before the line goes away. After the line goes away, I'll let them in if they have their signal on.

Never let the hot chick in front of you though. Because very rarely will she ever pull over and fuck you after.

--
Rob

[billvon 3,085]

>What I meant was that traffic would move faster if all would drive to the
> end of the ending lane (keep in mind this does not have to be an on
>ramp) and then be allowed to merge.

No. On average it would not. Only the people in the merge lane would move faster; the people in the lane next to them would have to move slower as a result. Thus the average flow stays the same.

>What holds up traffic is the idiots that want to merge prior to the ending
>of the lane, now stopping traffic behind them etc.

Right - but that's because they're idiots. If they get to the end of the lane and then stop traffic by trying to force their way in - same problem. The location that they enter the traffic flow isn't the issue, the ability to merge is.

[kelpdiver 2]

Quote

I never understood the early mergers, an invisible line must exist for them. If you merge after that line you're an asshole if you merge before the line you're courteous.

sounds like a variant of the George Carlin line about speed.

[DanG 1]

You're right as an academic exercise in a perfect world. In real traffic, the problem with early mergers is that they create multiple merge points. If there is only one merge point, then it obviously doesn't matter where it occurs. If there are multiple merges all happening into the same lane, it makes the local flow in that lane less efficient.

If you're looking at only having one merge point, then the end of the lane is the logical spot.

Don't get me started about mergers who don't give a polite wave after you let them in (not the one-fingered kind). There is a special circle of hell reserved for them.

- Dan G

[kelpdiver 2]

Quote

You're right as an academic exercise in a perfect world. In real traffic, the problem with early mergers is that they create multiple merge points. If there is only one merge point, then it obviously doesn't matter where it occurs. If there are multiple merges all happening into the same lane, it makes the local flow in that lane less efficient.

Not following this.

The advantage of multiple merge points is that they can be selected when advantageous - in particular, there is a lot of space and speeds can be matched.

when you have only one merge point - the one at the very end, there is no choice of when to do it and it's often not an ideal moment to do it. You frequently have to radically change your speed or cut off traffic to complete it.

This would be the fun sort of modeling on a computer, where you can define different personality types and put several hundred drivers into action and see how it plays out.

[billvon 3,085]

>If there is only one merge point, then it obviously doesn't matter where it
>occurs. If there are multiple merges all happening into the same lane, it
>makes the local flow in that lane less efficient.

Are you talking fuel efficiency or total traffic flow? The ideal for fuel efficiency is as early as possible, so average speeds are kept lower for as long as possible.

For total traffic flow it doesn't matter when the highway is at capacity. Again, you could be super efficient and have everyone merge at exactly the same point at the very end of the lane - and it wouldn't affect overall traffic flow one bit, because you cannot get more people into that lane than the lane is accepting no matter how "efficient" you are.

If the highway is below capacity (i.e. heavy traffic but moving near maximum speeds) then the most 'efficient' strategy is to merge where there's room; that minimizes braking and allows speeds to remain higher on average. As an example:

Two cars come down the entrance ramp towards the traffic flow, one right after the other. They begin to accelerate to traffic speed on the ramp. They are right next to a group of four cars that are as close together as is safe. There is room before and after the group. The best strategy for the mergers is for the front merger to speed up to enter ahead of the group and the rear merger to slow down to enter behind the group, and then to get to speed as quickly as possible. This is also the most fuel efficient solution; you minimize acceleration/braking of a larger number of cars (the four cars in the lane) and require acceleration/braking of two instead.

An alternative would be to have the cars in the lane start spacing themselves to allow the two cars in. This requires the second car to slow a bit, the third car to slow even more, and the rearmost of the four cars to slow down significantly. This in turn will slow the traffic behind them, and reduce lane efficiency overall. This can then be "made up" by repacking the cars and accelerating them all back up to speed once the new cars have merged, but isn't as efficient traffic-wise or fuel-wise.

[rehmwa 2]

Quote

>Only the people in the merge lane would move faster;

not really - only for the brief moment as all the merge moves to the new steady state point - then the mass flow would return to the previous state

the only real benefit to overall flow might be if there is a 'standard' merge process (zipper or whatever) that would reduce accidents due to more predictable behavior noticed between operators....

...
Driving is a one dimensional activity - a monkey can do it - being proud of your driving abilities is like being proud of being able to put on pants

[rehmwa 2]

Quote

>
For total traffic flow it doesn't matter when the highway is at capacity. Again, you could be super efficient and have everyone merge at exactly the same point at the very end of the lane - and it wouldn't affect overall traffic flow one bit, because you cannot get more people into that lane than the lane is accepting no matter how "efficient" you are.

ok, I'll stop berating the same point now

...
Driving is a one dimensional activity - a monkey can do it - being proud of your driving abilities is like being proud of being able to put on pants

[billvon 3,085]

>not really - only for the brief moment as all the merge moves to the
>new steady state point . . .

Right. And even for that brief moment the average speed doesn't change, just the instantaneous speed of the people in the merge lane.

>the only real benefit to overall flow might be if there is a 'standard'
>merge process (zipper or whatever) that would reduce accidents due to
>more predictable behavior noticed between operators.

Agreed there. Anything that reduces accidents in the long run will help both with traffic flow (accidents slow everything tremendously) and safety.