Why do pedestrian crossing signals count down but the traffic light stays green?

[DaveC426913]

It is possible there's one point of ambiguity left.

The pedestrian signals I'm referring to are not pedestrian signals for crossing the main road. They are pedestrian signals for crossing the secondary road, walking parallel with the main road.

 

[DaveC426913]

Then what is wrong with the simple hypothesis that some civil servant messed up?

Does this really need a three page thread?

As I have said about three times now, all the lights behave this way (in more than one city I might add) and it has been this way for years.

A mistake is the least plausible explanation.

 

[DaveC426913]

Does this really need a three page thread?

No! If the participants had read the thread before responding, I would not have had to repeat the scenario multiple times for Pete's sake.

 

[Pythagorean]

No! If the participants had read the thread before responding, I would not have had to repeat the scenario multiple times for Pete's sake.

I think it was a matter of awkward/ambiguous language, but we worked it out (even if it took three pages).

I don't see how somebody could legally or ethically allow themselves to program people to get hit by cars... so I agree with DH, it has to be an error, either in programming the traffic lights or a policy error that forces the traffic engineers to program it this way.

 

[Pythagorean]

wait a minute! I just saw your diagram. You're talking about pedestrian lights that are PARALLEL to the car lights (i was assuming perpendicular.)

This is not an error. The point here is to allow turners to turn. You have to stop pedestrian traffic, even if there's no cross-traffic, so that cars can turn right.

 

[DaveC426913]

wait a minute! I just saw your diagram. You're talking about pedestrian lights that are PARALLEL to the car lights (i was assuming perpendicular.)

This is not an error. The point here is to allow turners to turn. You have to stop pedestrian traffic, even if there's no cross-traffic, so that cars can turn right.

But this is on a long stretch of road with little pedestrian traffic. Why would you need to stop pedestrians? Like every other turn in the world, if there are pedestrians, the driver would simply wait until they've cleared the intersection, then make their turn.

 

[Pythagorean]

But this is on a long stretch of road with little pedestrian traffic. Why would you need to stop pedestrians? Like every other turn in the world, if there are pedestrians, the driver would simply wait until they've cleared the intersection, then make their turn.

hrm, good point. Maybe it's a programming error.

 

[DaveC426913]

hrm, good point. Maybe it's a programming error.

 

[edward]

The problem we have here is that drivers do not wait until the pedestrians have cleared the intersections because the intersections are so congested the drivers don't see the pedestrians. This happens with both right and left vehicle turns.

Here is one of our many intersections with 6 lanes on one street and 4 lanes on the cross street.

On one side of the three lanes in each direction street there are two left turn lanes, three through lanes and a right turn lane.

The streets with two lanes in each direction have one left turn lane two through lanes, one right turn lane and a partridge in a pear tree.

We have no count down lights for pedestrians. They often are stranded on a narrow concrete median in the middle of the street. (if they are lucky)

We call it the pedestrian lottery.

 

[Pythagorean]

That more what I was imagining. Maybe a traffic engineer from a busier area is working on DaveCs block or somebody just used default software, or maybe the pedestrian right of way makes it mandatory policy.

 

[DaveC426913]

a traffic engineer from a busier area is working on DaveCs block

Oh my god don;t make me use the 30pt font.

This is all over the place, both here in my town and all across neighboring Mississauga - city of a half million people.

 

[DaveC426913]

My buddy proposed an explanation.

He suggests that the traffic signal cycle is standard and fixed. So, every two minutes or so, the 'traffic lights are going to go red' sequence begins, and the crosswalk starts counting down to zero - but that the actual test to see if there's any cross traffic does not occur until just before it's about to change. When it does not detect any cross traffic, it aborts the traffic light change.

 

[Containment]

That seems fairly unlikely as dirt covering the sensor that checks for traffic would force a repair person out to fix the lights about every other day. I would bet it's programmed wrong as I believe a few people have said.

 

[Moonbear]

Oh my god don;t make me use the 30pt font.

This is all over the place, both here in my town and all across neighboring Mississauga - city of a half million people.

I think that's the point, that it's a default setting that will work in populated or unpopulated areas. Is there a button to push to cross in the other direction that will change the green light for cars to red? It could be that the pedestrian light in the one direction just keeps running on its timer, and if a vehicle tripped a sensor for the other direction, or someone pushed the button to cross the other way, that's the timing the light would use for cross traffic.

I'm really inclined toward this just being the default setting on the pedestrian lights and nobody bothered changing them because it doesn't matter as long as traffic and pedestians are light, but if there was an increase, they could set them differently later.

There's really no rational explanation for it.

 

[DaveC426913]

That seems fairly unlikely as dirt covering the sensor

The sensors are large metal detecting coils embedded in the road.

 

[Borg]

My buddy proposed an explanation.

He suggests that the traffic signal cycle is standard and fixed. So, every two minutes or so, the 'traffic lights are going to go red' sequence begins, and the crosswalk starts counting down to zero - but that the actual test to see if there's any cross traffic does not occur until just before it's about to change. When it does not detect any cross traffic, it aborts the traffic light change.

Sounds logical to me. We have traffic lights in my area with different settings but usually the timers rely on the pedestrian pushing the button first - if it doesn't get pushed, there is no countdown. So, the problem here is that if they don't push it, they can be in the middle of the street when the light changes. At least this way, your traffic light appears to give an indication that it could change.

The next step of course is to drive onto the coil on the crossing side to see if it actually waits until after the countdown to change. Bonus points for making the main road traffic stop unnecessarily and then backing up from the coil after the light changes.

BTW, I am amazed at how often I will see people stopped way off the coils wondering why they don't get a green light.

 

[Moonbear]

BTW, I am amazed at how often I will see people stopped way off the coils wondering why they don't get a green light.

I'm amazed at how often I get stuck BEHIND those people at that same light. I have to keep inching up on their bumper to nudge them forward onto the sensors. I always wonder how long they'd sit there waiting for the light to turn if nobody came up behind them and got too close for their comfort making them edge forward onto the sensor?

 

[russ_watters]

Here the sign changes from Walk, to Walk with a countdown, to Don't Walk.

Your version sound like the person who programmed the light played one round too many of "beer hunter" the niguht before.

Didn't read the whole thread, but I have a slightly different take: it is common in industrial controls for the manufacturer to provide pre-programmed sequence of operation modules to make life easier for the integrator. The downside is the integrator has limited flexibility to make custom sequences. There is an intersection near me with a similar issue:

A straight soutbound road with another forking left at a 30 degree angle and no turning between the forks and two lanes going south. The southbound right lane could stay green while the left cycles between green and green with left arrow - with the north angled fork between green and red. But it doesn't: the southbound lights both go red for 1 second (including a normal yellow) before switching back to green. This strikes me as a simple programming limitation: the light cycle must include a full cycle of all colors with a minimum adjustable red time of 1 sec.

Dave's sounds like the opposite problem: a maximum cycle length for the "walk" signal.

 

[MarcoD]

Is there a button to push to cross in the other direction that will change the green light for cars to red?

This was my best bet too. If it never changes, it is broken, or it never changes because there is no traffic which signals the system that it should change state (to red or green, depending where you're at.)

 

[TheStatutoryApe]

My buddy proposed an explanation.

He suggests that the traffic signal cycle is standard and fixed. So, every two minutes or so, the 'traffic lights are going to go red' sequence begins, and the crosswalk starts counting down to zero - but that the actual test to see if there's any cross traffic does not occur until just before it's about to change. When it does not detect any cross traffic, it aborts the traffic light change.

This is sort of what I was figuring. The lights stay green because there is no cross traffic. The pedestrian signals go through their cycles and return to green because there is no cross traffic. When I living in Long Beach they had automatic pedestrian signals, no buttons, so that people did not have to rush to push the button. If they always stayed green though then any approaching cross traffic would automatically have to stop and wait for the pedestrian cycle before the light could change and the pedestrian cycle would have to be connected to the ground sensor as well. So the pedestrian lights cycle letting pedestrians know that if a vehicle approaches then the light will be changing when the cycle ends. There was almost always traffic on that main road though, while I think I remember seeing the lights there do this I do not recall clearly as it would not have been occurring for any prolonged period of time.

Edit: Oh, and I think your theory about it being a control on the continuous traffic may be a secondary reason to keep it this way. I believe civic engineers actually consider such ideas. There are some freeways around here that they built curvy intentionally to keep drivers from suffering road hypnosis.