Another WWJD

> So, if I'm reading your proposal correctly, the 2nd car, which just a
> couple
> car lenghts behind the first car, would be expected to *instantly* stop as
> the phase would immediately change to red as soon as the 1st car passes.
> The queue of cars, each of which is likely within a few car lengths of
> each other, would then be required by physics to suddenly perform the same
> abrupt stop, or some quick evasive maneuvers at least...
>
> Gee, I wonder what could be wrong with your "solution"...
>
> brink
>

As I understand it, this is only a problem on ramps with 2-color signals.
That is, there is no yellow phase. Best solution would be to default to
green, then have any car passing the signal trigger a yellow phase, followed
by a red phase. If there were any rear-end action then, ticket the driver
in the rear and let the insurance companies fight it out. -Dave

Mike T. wrote:
> > So, if I'm reading your proposal correctly, the 2nd car, which just a
> > couple
> > car lenghts behind the first car, would be expected to *instantly* stop as
> > the phase would immediately change to red as soon as the 1st car passes.
> > The queue of cars, each of which is likely within a few car lengths of
> > each other, would then be required by physics to suddenly perform the same
> > abrupt stop, or some quick evasive maneuvers at least...
> >
> > Gee, I wonder what could be wrong with your "solution"...
> >
> > brink

Here (NY Tristate area), metering ramps where they exist tend to be at
the beginning of the on-ramp. Regardless, I am not understanding your
concern - if the group of vehicles is travelling at higher speeds
anyway, and the drivers are so completely oblivous, then the first
vehicles is gonna get rear ended when it stops for the metering ramp -
else the second car is gonna get creamed when it stops for the metering
ramp. Otherwise if the group is slow, or if the drivers are paying
attention, then nothing much happens either system.
I think what you are saying is that under the current system, the first
vehicles putzes along seeing the default red light, stopping and then
proceeding, and slowing up the conga line of vehicles following enough
that the next driver, paying no attention what-so-ever, doesn't
rear-end him...kinda lame, don't you think?
Of course, I highly support using the good ol' red+green = yellow phase
the old two color traffic lights that I saw growing up in Queens
used... (of course, since these phases can be as short 2 seconds
long, that means green->green/red->red->green in 4 seconds - hopefully
they don't use the metering ramp as a drag race christmas tree).

"Sir Ray" > wrote in message
ups.com...
>
> Mike T. wrote:
>> > So, if I'm reading your proposal correctly, the 2nd car, which just a
>> > couple
>> > car lenghts behind the first car, would be expected to *instantly* stop
>> > as
>> > the phase would immediately change to red as soon as the 1st car
>> > passes.
>> > The queue of cars, each of which is likely within a few car lengths of
>> > each other, would then be required by physics to suddenly perform the
>> > same
>> > abrupt stop, or some quick evasive maneuvers at least...
>> >
>> > Gee, I wonder what could be wrong with your "solution"...
>> >
>> > brink
>
> Here (NY Tristate area), metering ramps where they exist tend to be at
> the beginning of the on-ramp. Regardless, I am not understanding your
> concern - if the group of vehicles is travelling at higher speeds
> anyway, and the drivers are so completely oblivous, then the first
> vehicles is gonna get rear ended when it stops for the metering ramp -
> else the second car is gonna get creamed when it stops for the metering
> ramp.

It's a freeway on-ramp -- of course people drive fast on them! Whether they
should or not -- people see FREEWAY and they think MASH THE GAS -- as well
they should when they have a green light in front of them! Merging onto the
freeway requires speed!

You're proposing giving them a GREEN light -- why would drivers suspect they
need to "cautiously proceed" through your green light? They're merging onto
a freeway with a green light!

When the ramp is metered, they KNOW to slow down because they see a red
light ahead... you're proposing springing a surprise red light on them and
expecting them to be ready to stop. Why?

Otherwise if the group is slow, or if the drivers are paying
> attention, then nothing much happens either system.
> I think what you are saying is that under the current system, the first
> vehicles putzes along seeing the default red light, stopping and then
> proceeding, and slowing up the conga line of vehicles following enough
> that the next driver, paying no attention what-so-ever, doesn't
> rear-end him...kinda lame, don't you think?
> Of course, I highly support using the good ol' red+green = yellow phase
> the old two color traffic lights that I saw growing up in Queens
> used... (of course, since these phases can be as short 2 seconds
> long, that means green->green/red->red->green in 4 seconds - hopefully
> they don't use the metering ramp as a drag race christmas tree).

Even with a yellow phase, wouldn't you expect a line of 4-5 cars to sneak
through? And justifiably so? People cruise in a line up and down these
ramps... you'd want a good 3-second yellow phase to warn drivers of an
imminent stop.

It's an accident waiting to happen and it wouldn't meter out cars one by one
like it should.

brink

>
> Even with a yellow phase, wouldn't you expect a line of 4-5 cars to sneak
> through? And justifiably so? People cruise in a line up and down these
> ramps... you'd want a good 3-second yellow phase to warn drivers of an
> imminent stop.
>
> It's an accident waiting to happen and it wouldn't meter out cars one by
> one like it should.
>
> brink

I think most people would be willing to live with that compromise, though.
That is, default to green, add a yellow phase. If a few extra cars sneak
through on each yellow phase, what's the problem there? The big question
is, does metering work? It's odd that the only place this "metering" crap
is used is in areas that have the very worst traffic problems to begin with.
So if they work, you'd think that either the traffic wouldn't be as bad as
it is reported to be, OR there must be some statistics showing that the
metering ramps do improve traffic flow, in terms of increased number of
vehicles passing from one end of the highway to the other in X amount of
time. That is, take a 15-mile stretch of congested highway and count how
many vehicles can get from exit 1 to exit 15 (for example) in an hour. Then
turn the metering signals on so that the exits now are controlled by
signals. If your number of cars per hour travelling from exit 1 to exit 15
increases, that would be a clear indication that the metering works.

I suspect if anybody actually did the study though, they'd find that the
metering is totally ineffective. On a truly congested highway, it might
help the congestion on the highway. But then, the backup on the onramp is
likely to add just as much travel time as was saved on the highway. So I
suspect there is no fricking difference, except that it moves the physical
location where you spend some of your time idling and going nowhere. I know
some exits I'm familiar with, a meter allowing one car at a time to enter
the highway would have traffic backed up (approaching the highway) for
several miles by 7AM any weekday. -Dave

"Dave" > wrote in message
reenews.net...
> >
>> Even with a yellow phase, wouldn't you expect a line of 4-5 cars to sneak
>> through? And justifiably so? People cruise in a line up and down these
>> ramps... you'd want a good 3-second yellow phase to warn drivers of an
>> imminent stop.
>>
>> It's an accident waiting to happen and it wouldn't meter out cars one by
>> one like it should.
>>
>> brink
>
> I think most people would be willing to live with that compromise, though.
> That is, default to green, add a yellow phase. If a few extra cars sneak
> through on each yellow phase, what's the problem there? The big question
> is, does metering work?

In my observation, yes, it does. Again, the problem with ramps is that cars
tend to come in clumps when a phase changes at the bottom of the ramp. You
can have 20 cars in a queue driving up a ramp and with that many coming at
once, there's no way they'll be at freeway speed given most ramp lengths.

The resulting wave effect of sudden braking as the line of cars tries to
"zipper" into the flow of the right freeway lane, coupled with the wave
effect of braking as cars in the middle lane are dodging cars trying to
change lanes to avoid the line of cars, and then the wave effect of...
well, you get the picture.

The delay of 20 cars suddenly turns into the delay of hundreds if not
thousands of cars on a large freeway -- this is what "waves" create, not to
mention an increased risk of accidents because of sudden changes of freeway
speeds and lane changers.

It's odd that the only place this "metering" crap
> is used is in areas that have the very worst traffic problems to begin
> with.

Well, you can say the same thing about stoplights. You don't see stoplights
at low-volume intersections. I fail to see how this is an argument against
the implementation of meters.

> So if they work, you'd think that either the traffic wouldn't be as bad as
> it is reported to be, OR there must be some statistics showing that the
> metering ramps do improve traffic flow, in terms of increased number of
> vehicles passing from one end of the highway to the other in X amount of
> time. That is, take a 15-mile stretch of congested highway and count how
> many vehicles can get from exit 1 to exit 15 (for example) in an hour.
> Then turn the metering signals on so that the exits now are controlled by
> signals. If your number of cars per hour travelling from exit 1 to exit
> 15 increases, that would be a clear indication that the metering works.

Good point. I've crossposted over to misc.transport.road since there are
several traffic engineers who post there. I'm sure someone has hands-on
experience with freeway metering, perhaps they'll be willing to share the
info they may have.

brink

In article >,
brink > wrote:
>It's odd that the only place this "metering" crap
>> is used is in areas that have the very worst traffic problems to begin
>> with.
>
>Well, you can say the same thing about stoplights. You don't see stoplights
>at low-volume intersections. I fail to see how this is an argument against
>the implementation of meters.

In an area with traffic growth metering is a temporary patch.
Ramp meters slightly increase the volume that can be carried
before traffic stalls. They should be treated as interim
fixes while the urgently needed capacity improvements are
constructed. A region with permanent ramp meters has a
failed transportation policy.

Ordinary traffic signals work over a much wider range of
volume.

--
John Carr )

> brink wrote:
>
> > So if they work, you'd think that either the traffic wouldn't be as bad as
> > it is reported to be, OR there must be some statistics showing that the
> > metering ramps do improve traffic flow, in terms of increased number of
> > vehicles passing from one end of the highway to the other in X amount of
> > time. That is, take a 15-mile stretch of congested highway and count how
> > many vehicles can get from exit 1 to exit 15 (for example) in an hour.
> > Then turn the metering signals on so that the exits now are controlled by
> > signals. If your number of cars per hour travelling from exit 1 to exit
> > 15 increases, that would be a clear indication that the metering works.
>
> Good point. I've crossposted over to misc.transport.road since there are
> several traffic engineers who post there. I'm sure someone has hands-on
> experience with freeway metering, perhaps they'll be willing to share the
> info they may have.

Although the ramp meter didn't begin in Minnesota, MnDOT has been one
of the leading pioneers behind ramp meter usage. Every study MnDOT's
done on them (including one in 2000 where they were turned off for 6
weeks and the effects studied) has found that they increase freeway
speeds and capacity and decrease accidents.

Of course, the catch is that they tend to clog up the on-ramps, but
overall delay on the mainline is lower. They're not for every area,
though. The public tends to dislike them on freeway-to-freeway
junctions (especially those with heavy traffic volumes). They usually
work best on those urban freeways with close or relatively close
interchange spacing (less than 1 mile), as they'll discourage the short
freeway trips which tend to clog up the freeway...and in fact that was
one of the original goals of ramp meter usage.

Contrary to Mr. Carr's opinion, they're an important element of the
overall freeway management system.

Froggie

wrote:
> > brink wrote:
> >
> > > So if they work, you'd think that either the traffic wouldn't be as bad as
> > > it is reported to be, OR there must be some statistics showing that the
> > > metering ramps do improve traffic flow, in terms of increased number of
> > > vehicles passing from one end of the highway to the other in X amount of
> > > time. That is, take a 15-mile stretch of congested highway and count how
> > > many vehicles can get from exit 1 to exit 15 (for example) in an hour.
> > > Then turn the metering signals on so that the exits now are controlled by
> > > signals. If your number of cars per hour travelling from exit 1 to exit
> > > 15 increases, that would be a clear indication that the metering works.
> >
> > Good point. I've crossposted over to misc.transport.road since there are
> > several traffic engineers who post there. I'm sure someone has hands-on
> > experience with freeway metering, perhaps they'll be willing to share the
> > info they may have.
>
> Although the ramp meter didn't begin in Minnesota, MnDOT has been one
> of the leading pioneers behind ramp meter usage. Every study MnDOT's
> done on them (including one in 2000 where they were turned off for 6
> weeks and the effects studied) has found that they increase freeway
> speeds and capacity and decrease accidents.
>
> Of course, the catch is that they tend to clog up the on-ramps, but
> overall delay on the mainline is lower. They're not for every area,
> though. The public tends to dislike them on freeway-to-freeway
> junctions (especially those with heavy traffic volumes). They usually
> work best on those urban freeways with close or relatively close
> interchange spacing (less than 1 mile), as they'll discourage the short
> freeway trips which tend to clog up the freeway...and in fact that was
> one of the original goals of ramp meter usage.

There are two types of ramp metering IMO:

1) A fast cycling meter which will spread the platoon of arriving cars
out onto the freeway
2) a slow meter which serves to limit access to the freeway by holding
cars on the ramps for an excessive amount of time, while keeping an
empty mainline.

IMO, the first form is acceptable, the second is not. I've sort of
suspected that MN uses the second form.

>
> Contrary to Mr. Carr's opinion, they're an important element of the
> overall freeway management system.

I would tend to agree.

>
> Froggie

brink wrote:
> "Dave" > wrote in message
> reenews.net...
> > >
> >> Even with a yellow phase, wouldn't you expect a line of 4-5 cars to sneak
> >> through? And justifiably so? People cruise in a line up and down these
> >> ramps... you'd want a good 3-second yellow phase to warn drivers of an
> >> imminent stop.
> >>
> >> It's an accident waiting to happen and it wouldn't meter out cars one by
> >> one like it should.
> >>
> >> brink
> >
> > I think most people would be willing to live with that compromise, though.
> > That is, default to green, add a yellow phase. If a few extra cars sneak
> > through on each yellow phase, what's the problem there? The big question
> > is, does metering work?
>
> In my observation, yes, it does. Again, the problem with ramps is that cars
> tend to come in clumps when a phase changes at the bottom of the ramp. You
> can have 20 cars in a queue driving up a ramp and with that many coming at
> once, there's no way they'll be at freeway speed given most ramp lengths.
>
> The resulting wave effect of sudden braking as the line of cars tries to
> "zipper" into the flow of the right freeway lane, coupled with the wave
> effect of braking as cars in the middle lane are dodging cars trying to
> change lanes to avoid the line of cars, and then the wave effect of...
> well, you get the picture.
>
> The delay of 20 cars suddenly turns into the delay of hundreds if not
> thousands of cars on a large freeway -- this is what "waves" create, not to
> mention an increased risk of accidents because of sudden changes of freeway
> speeds and lane changers.
>
> It's odd that the only place this "metering" crap
> > is used is in areas that have the very worst traffic problems to begin
> > with.
>
> Well, you can say the same thing about stoplights. You don't see stoplights
> at low-volume intersections. I fail to see how this is an argument against
> the implementation of meters.
>
> > So if they work, you'd think that either the traffic wouldn't be as bad as
> > it is reported to be, OR there must be some statistics showing that the
> > metering ramps do improve traffic flow, in terms of increased number of
> > vehicles passing from one end of the highway to the other in X amount of
> > time. That is, take a 15-mile stretch of congested highway and count how
> > many vehicles can get from exit 1 to exit 15 (for example) in an hour.
> > Then turn the metering signals on so that the exits now are controlled by
> > signals. If your number of cars per hour travelling from exit 1 to exit
> > 15 increases, that would be a clear indication that the metering works.
>
> Good point. I've crossposted over to misc.transport.road since there are
> several traffic engineers who post there. I'm sure someone has hands-on
> experience with freeway metering, perhaps they'll be willing to share the
> info they may have.

When CDOT put in the ramp meters on north I-25, it kept records of
travel times and speeds and found that after the meters went
operational, they improved peak hour speeds and travel times. CDOT uses
meters that are timed according to the flow of traffic on the freeway,
as opposed to some jurisdictions that have a fixed time cycle to the
green. On Denver freeways, the meter signal timings are constantly
changing with the speed sensors on the roadway, allowing cars to enter
more frequently from the ramps as traffic volumes lower and speeds
increase on the mainline.

> Mike Tantillo wrote:
>
> There are two types of ramp metering IMO:
>
> 1) A fast cycling meter which will spread the platoon of arriving cars
> out onto the freeway
> 2) a slow meter which serves to limit access to the freeway by holding
> cars on the ramps for an excessive amount of time, while keeping an
> empty mainline.
>
> IMO, the first form is acceptable, the second is not. I've sort of
> suspected that MN uses the second form.

Both. Except for isolated stand-alone meters (which are few and far
between), all of the Twin Cities meters are centrally coordinated.
MnDOT uses a traffic algorithm based on traffic sensors in the pavement
(plus extensive use of cameras for visual monitoring) to determine
which level to set the meters at. Heavier congestion will, thusly,
result in a slow meter.

Froggie