News:

Thank you for your patience during the Forum downtime while we upgraded the software. Welcome back and see this thread for some new features and other changes to the forum.

Main Menu

If you "just miss" a green light, are you more likely to catch up later?

Started by kurumi, February 01, 2019, 11:52:16 AM

Previous topic - Next topic

jakeroot

Quote from: roadfro on February 09, 2019, 05:26:29 PM
Quote from: jakeroot on February 09, 2019, 02:29:13 PM
Quote from: roadfro on February 09, 2019, 12:56:11 PM
Quote from: jakeroot on February 06, 2019, 04:17:13 PM
Even in non-timed environments, where signals use detection, it helps to drive as part of a pack, as a line of cars will keep a light green until an unknown maximum green time is reached. One of my pet-peeves is when drivers fall behind, creating a gap just large enough for the signal to "time"-out.
The signal controller setting for this is the "Passage Gap", which is a countdown timer tied to the upstream vehicle detectors–after the initial green time, the passage gap timer starts counting down, but resets every time a vehicle is detected upstream. If the passage gap timer reaches zero prior to the maximum green time, the signal turns yellow early–this is often referred to by signal engineers as "gap out" or "gapping out".
Thanks for the explanation. Never knew how that worked. The timer seems to be set around 1.5 to 2 seconds at most lights in my area (with the initial green time at most signals around 3 seconds).
It varies depending on upstream detector placement and assumed roadway speed/speed limit, but I believe 2 seconds is a pretty common value for passage gap.

Which is odd to me, since I believe the WA driving manual suggests four-second following distances at 40 mph. Which is insane, of course, and ignored by everyone for being insane. But thank god for that, otherwise signals would be "gapping out" left and right.


roadfro



Quote from: jakeroot on February 09, 2019, 08:51:16 PM
Quote from: roadfro on February 09, 2019, 05:26:29 PM
Quote from: jakeroot on February 09, 2019, 02:29:13 PM
Thanks for the explanation. Never knew how that worked. The timer seems to be set around 1.5 to 2 seconds at most lights in my area (with the initial green time at most signals around 3 seconds).
It varies depending on upstream detector placement and assumed roadway speed/speed limit, but I believe 2 seconds is a pretty common value for passage gap.

Which is odd to me, since I believe the WA driving manual suggests four-second following distances at 40 mph. Which is insane, of course, and ignored by everyone for being insane. But thank god for that, otherwise signals would be "gapping out" left and right.

The upstream detectors and passage gap timer are set independently of any assumed vehicle following distance. But I do seem to recall (I'd have to check old textbooks/notes from undergrad) that detectors are supposed to be placed and passage gap set such that a car passing at the 85th percentile speed/speed limit would be able to clear the stop line before the light turns yellow.
Roadfro - AARoads Pacific Southwest moderator since 2010, Nevada roadgeek since 1983.

jakeroot

Quote from: roadfro on February 12, 2019, 09:36:53 PM
The upstream detectors and passage gap timer are set independently of any assumed vehicle following distance. But I do seem to recall (I'd have to check old textbooks/notes from undergrad) that detectors are supposed to be placed and passage gap set such that a car passing at the 85th percentile speed/speed limit would be able to clear the stop line before the light turns yellow.

I've always been bothered by those who insist on the exact posted limit. Now, I have a scientific reason for it to bother me.

roadfro

Quote from: jakeroot on February 13, 2019, 12:46:56 AM
Quote from: roadfro on February 12, 2019, 09:36:53 PM
The upstream detectors and passage gap timer are set independently of any assumed vehicle following distance. But I do seem to recall (I'd have to check old textbooks/notes from undergrad) that detectors are supposed to be placed and passage gap set such that a car passing at the 85th percentile speed/speed limit would be able to clear the stop line before the light turns yellow.

I've always been bothered by those who insist on the exact posted limit. Now, I have a scientific reason for it to bother me.

I was incorrect. I think I was confusing the passage gap calculation with the dilemma zone calculation...

The formula for setting passage time takes the maximum allowable headway (following distance) and subtracts another term derived from detector+vehicle lengths over average approach speed (usually posted speed limit is used). Passage gap considers headways between vehicles to determine how long to extend the phase, so the placement of detectors is actually not a critical factor–stop line detectors can be (but are not commonly) used for this–but the detector placement and design still needs to be considered.

BTW: Several factors in traffic signal timing rely on 85th percentile speed or average approach speeds–more often than not, posted speed limits are substituted for this term in calculations, since it is a known value that does not require a speed study to determine a precise value.
Roadfro - AARoads Pacific Southwest moderator since 2010, Nevada roadgeek since 1983.

MikieTimT

Other than signals that are managed from a single cabinet over a few too-closely spaced intersections in Arkansas, I've not noticed that signals are timed from one to another period anymore.  It used to be before the advent of cameras/loops that some effort was made to time lights on a given major thoroughfare, but it seems like those days are long gone.

jakeroot

Quote from: roadfro on February 13, 2019, 02:45:27 AM
BTW: Several factors in traffic signal timing rely on 85th percentile speed or average approach speeds–more often than not, posted speed limits are substituted for this term in calculations, since it is a known value that does not require a speed study to determine a precise value.

I see. I will partially retract my statement, though "keeping up" still seems like the easiest way to game the system. :D

Quote from: MikieTimT on February 14, 2019, 06:55:54 AM
Other than signals that are managed from a single cabinet over a few too-closely spaced intersections in Arkansas, I've not noticed that signals are timed from one to another period anymore.  It used to be before the advent of cameras/loops that some effort was made to time lights on a given major thoroughfare, but it seems like those days are long gone.

It seems to me that, especially in the 50s and 60s, before the wide use of things like protected left turns, signal programming was relatively simple. Each intersection likely had only two phases, and as there were no detection technology, every signal was timed. So they timed them appropriate for through progression, or as best they could assuming no other traffic. Now, with lots more traffic, signals for each movement, crosswalks (egad!), etc, more advanced technology has become necessary. What I see, now, in suburban areas, when lights are spaced around 10+ blocks apart, is for each to run independently, with the lights able to stay green when platoons are detected (using the aforementioned passage gap calculation). There's also lead/lag phasing, which can help TOD operation.

From what I've seen, these advanced signal technologies can handle heavy traffic really well, although they can be a bit maddening during lighter hours when they change, what seems like, every 7 seconds to account for each arriving car. Hell, they sometimes feel like four-way stops in the overnight hours.

MikieTimT

Quote from: jakeroot on February 15, 2019, 02:21:32 PM
Quote from: roadfro on February 13, 2019, 02:45:27 AM
BTW: Several factors in traffic signal timing rely on 85th percentile speed or average approach speeds–more often than not, posted speed limits are substituted for this term in calculations, since it is a known value that does not require a speed study to determine a precise value.

I see. I will partially retract my statement, though "keeping up" still seems like the easiest way to game the system. :D

Quote from: MikieTimT on February 14, 2019, 06:55:54 AM
Other than signals that are managed from a single cabinet over a few too-closely spaced intersections in Arkansas, I've not noticed that signals are timed from one to another period anymore.  It used to be before the advent of cameras/loops that some effort was made to time lights on a given major thoroughfare, but it seems like those days are long gone.

It seems to me that, especially in the 50s and 60s, before the wide use of things like protected left turns, signal programming was relatively simple. Each intersection likely had only two phases, and as there were no detection technology, every signal was timed. So they timed them appropriate for through progression, or as best they could assuming no other traffic. Now, with lots more traffic, signals for each movement, crosswalks (egad!), etc, more advanced technology has become necessary. What I see, now, in suburban areas, when lights are spaced around 10+ blocks apart, is for each to run independently, with the lights able to stay green when platoons are detected (using the aforementioned passage gap calculation). There's also lead/lag phasing, which can help TOD operation.

From what I've seen, these advanced signal technologies can handle heavy traffic really well, although they can be a bit maddening during lighter hours when they change, what seems like, every 7 seconds to account for each arriving car. Hell, they sometimes feel like four-way stops in the overnight hours.

I remember back when I was a kid and we lived out in the boonies 40 miles south of Ft. Smith, AR.  My dad used to take my brother and I to Irish-Made donuts about 6AM on a Saturday to get some freshly made donuts.  The traffic lights on Towson Ave. all used to be flashing yellow on Towson and flashing red on the intersecting streets during nights and weekends.  Made so much more sense back then than the way they do things now.  There is almost always one road that has more traffic than the other road at an intersection that it would make sense to change over to flashing yellow during the lighter times.  Flow through the intersections would be better to allow one road to always have priority in those off times and essentially a flashing red stop light for the intersecting road.  I guarantee that on average those on the intersecting roads would still wait at the intersection less than the sensors can switch their light to green.  Essentially the way things are now, stoplights really do act as stoplights for everyone in lighter traffic.

That same stretch of road was timed on stoplights during their regularly functioning times where driving the speed limit would only catch one of the stoplights, but allow you through the rest of the intersections on green after you'd been held up once.  I can't say that progress has really been made with the changes in traffic light control over the last 2 decades.



Opinions expressed here on belong solely to the poster and do not represent or reflect the opinions or beliefs of AARoads, its creators and/or associates.