October 2015 ABAC Meeting

[dasgeh]

On back on topic: The light for crossing Wakefield along Columbia Pike certainly needs to be rest in walk.

The lights crossing both George Mason and Glebe at the ramp from EB 50 (so south of 50) are technically along the Arlington Boulevard Trail, and should be rest in walk.

Any others? I’d like to wrap this up and email the County.

[elbows]

To add more intersections:

Yesterday, while riding past the marathon traffic parking lot, I noticed that the ped button at (I think) Arl. Blvd. Trail and N. Fairfax Drive does not seem to work at all.
Also, I know we’ve talked about this before but the ped button at Irving and 50 does work BUT, and I could be wrong, I think it must be pressed for a cyclist to get the green which is both not communicated to cyclists and a little awkward – or maybe that’s in part because I’m always carrying groceries there. Plus, while I’m pretty slow, even I don’t need 30 seconds to cross 50 – at places like that, maybe it would make sense to have a bicycle cycle (mostly cyclists use it and many drivers must wait – 20 seconds once the inter section clears X 20 drivers does add up).
Also, the ped button at Columbia Pike and S. Scott can be a ridiculously long wait sometimes. Like 4 or more minutes, it seems, even when the traffic is low. I guess I could verify this.

[dasgeh]

@elbows 126735 wrote:

Yesterday, while riding past the marathon traffic parking lot, I noticed that the ped button at (I think) Arl. Blvd. Trail and N. Fairfax Drive does not seem to work at all.

Can you show me where on a map?

@elbows 126735 wrote:

Also, I know we’ve talked about this before but the ped button at Irving and 50 does work BUT, and I could be wrong, I think it must be pressed for a cyclist to get the green which is both not communicated to cyclists and a little awkward – or maybe that’s in part because I’m always carrying groceries there. Plus, while I’m pretty slow, even I don’t need 30 seconds to cross 50 – at places like that, maybe it would make sense to have a bicycle cycle (mostly cyclists use it and many drivers must wait – 20 seconds once the inter section clears X 20 drivers does add up).

Irving and 50 is a horrible intersection that is in the process of being redesigned. I ask about it at most BAC meetings, and I think we’re getting an update on it next week. I believe we’ve suggested a bike light there, and I don’t remember what they said. I know we’ve been promised bike detection, so we won’t have to push no stinkin’ button.

@elbows 126735 wrote:

Also, the ped button at Columbia Pike and S. Scott can be a ridiculously long wait sometimes. Like 4 or more minutes, it seems, even when the traffic is low. I guess I could verify this.

I’ll add to the list, but if you could confirm, that would be great. Thanks

[elbows]

@dasgeh 126738 wrote:

Can you show me where on a map?

If I got it right, it is this one.

Thanks so much, dasgeh.

[scoot]

My understanding of “rest-in-walk”, after a few minutes of googling:

Where a major road and a minor intersect at a sensor-actuated signal, it is only used for walk signals parallel to the major road. The reason: green phases for the minor roads actuated by vehicle detectors are often too short to accommodate pedestrians trying to cross the large highway. So for crossings that are used by more vehicles than peds, triggering the walk cycle automatically would increase wait times and delays on the arterial. A beg button is used so that pedestrians can request the extra time when they are present.

Irving and 50 is a good example of such an intersection. I have been automatically detected with a CaBi over the sensor, but it is sporadic at best. So if there are no other vehicles waiting, I just push the button. It’s inefficient though: Since I can ride across the intersection in a few seconds, Arlington Blvd traffic is then stuck watching an empty intersection for another 25 seconds. Rest-in-walk is already used for pedestrians parallel to 50 crossing Irving, but would not work for Irving crossing 50.

Some of the other signals discussed in this thread, while problematic for various reasons, are also not candidates for rest-in-walk behavior. The Arlington Boulevard trail crossings at Glebe and George Mason are a special case, but they too might be a tough sell. Because those roads don’t intersect 50 at-grade, they become the major roads where they intersect the trail and off-ramps/service roads. IIRC, most of the intersections in Crystal City are on automatic cycles, not sensor-actuated, so those could incorporate rest-in-walk (and should, since it’s an urban area with lots of peds).

[dasgeh]

@scoot 126747 wrote:

My understanding of “rest-in-walk”, after a few minutes of googling:

You’re description is what was described to us at the BAC meeting. However, County staff also stated that
– not all signals are correctly calibrated, so that some pedestrian signals should be rest in walk but aren’t. E.g. crossing Wakefield (small street) along Columbia Pike – that should be rest-in-walk but just isn’t. County staff wants to know so they can fix it.

– trails should always get rest-in-walk, even if it delays the arterial, because they are trails (given, the example we were discussing when this statement was made was the W&OD, not the ABT, but the statement was generalized, so I’m holding them to it!). The ABT crossing of (I call it the on-ramp from Courthouse, elbows called it N Fairfax) certainly should be rest-in-walk with the mainline traffic continuing onto 50 (because it crosses the traffic coming from Courthouse road, getting on to 50).

– yes, automatically activating the pedestrian signal can cause some delay to motorists, but how much depends on the width of the road, and how many cars go through from the side streets. You picked the intersection with probably the highest possible delay in Arlington — there aren’t many cars that cross 50 at Irving, and 50 is VERY wide, so the difference in time that 50 is normally stopped (i.e. the time for one or two cars to cross on Irving) and the time that 50 would have to be stopped for a pedestrian to cross is huge. Think about a normal two lane road, and/or a road where typically, when the side street is actuated, there are 4 cars. The difference in stopping time for the arterial when there is one car and when there is one pedestrian is minimal (may 10 seconds?). The difference in stopping time for the arterial when there are four cars and when there is one pedestrian probably nonexistent. Add to that situation lots of pedestrians, who otherwise might just ignore the pedestrian signal and chance it, possibly getting “stranded” in the middle of the street when the signal for the arterial changes.

So in those examples, in order to encourage people to walk and bike and to keep people who do choose to walk and bike, the County should program to signals to rest-in-walk, and accept the potential of small delays for motorists on the arterials.

–

[mstone]

I’d be more willing to accept the “not fair for drivers to have to wait for empty crosswalk” argument if there were a quid pro quo involving peds/bikes not having to wait extremely long times as a baseline expectation.

[scoot]

To be clear, my post was intended to summarize the present engineering standards regarding “rest-in-walk”, not to advocate for or against them.

For biking, I usually choose between the Irving and Fillmore crossings based on traffic levels and direction. Most of my northbound crossings are during evening rush hour, and I prefer Irving. Fillmore can be uncomfortably busy during these times. There are usually a few vehicles at Irving in rush hour, so I don’t need to use the beg button. Most of my southbound crossings are later in the evening, and then I prefer Fillmore. It’s more direct to my destination, more reliable, and not totally devoid of activity. Plus no ride is complete without at least one little hill climb before getting home!

[elbows]

As for the Pike and S Scott and my claim that it was 4 minutes once, I went back there tonight at about 9:15 pm and it was only 65 seconds after I hit the call button until the walk sign came on. Still a little long especially given that traffic was low, but not 4 minutes.

[CaseyKane50]

@dasgeh 126493 wrote:

Here’s one part of the signals discussion:

At this month’s BAC meeting, Josh Nicholas from the Signals team in DES came to present. One issue we discussed was pedestrian signals that do not automatically turn to “walk” when the corresponding traffic light turns green, or that turn to “walk” then count down and change to “don’t walk” automatically after a certain amount of time, even if the corresponding traffic signal remains green. Josh asked us to send him a list of such signals that should be set to “rest in walk” but are not.

For example, along the north/west side Arlington Boulevard Trail, where the trail crosses the traffic entering from Courthouse, the pedestrian signal should be rest in walk, automatically switching to walk with the green for cars going along westbound 50. Another example is North Monroe at Lee Highway.

If you have other examples (like the intersections along Columbia Pike near 27 referenced in the WashCycle writeup), please post here (or otherwise send them to me) and I’ll compile a list.

Thanks

I think the intersection of South Meade and South Glebe might be a candidate to have the walk signal activated when the light turns green for traffic on South Meade.

South Meade has a bike lane that ends about 100 feet before the intersection.

[ATTACH=CONFIG]10051[/ATTACH]

As you approach the intersection, you can choose to get in the left turn lane or the right turn lane. Most bicyclists actually want to go straight across Glebe to the sidewalk.

[ATTACH=CONFIG]10052[/ATTACH]

My experience on a bike is that neither lane detects a bike and doesn’t activate the light. However, a car will activate the light, but the walk signal is only activated if the button is pushed. When the walk light is not activated, cars or bikes have a green light for about 15 seconds.

[Author]

Posts