help us count traffic!

[aroraborealis.livejournal.com]

We need a few volunteers to help us evaluate a new pedestrian safety option for crosswalks in Somerville called a pedestrian-head-start. With a pedestrian-head-start a person crossing a street will get the walk-signal 3 to 5 seconds before a light turns green for cars: giving pedestrians a chance to enter the intersection before turning cars. Cars see the pedestrians and yield to them as they are already in the crosswalk before they get the green light.

Currently in Somerville pedestrians must push a button and wait for an “all-walk” phase where all the traffic is stopped and pedestrians can cross without any conflicts. A disadvantage to the all-walk is that few pedestrians are willing to wait for the signal and end up crossing against the light. Often a walker will push the walk button and proceed before getting the walk signal. Then when the signal finally changes, all the cars end up waiting for a walk signal that serves no-one. Even when walkers do wait for the lights, both cars and pedestrians have to wait longer for the signal to proceed.

Cambridge has been using the Pedestrian-head-start for many years now. We need eight volunteers to help us do a comparative study of two intersections in Cambridge and Somerville. The data we gather will help us to measure safety factors between the two crossing-light methods.

We’re flexible about when the counts take place-- either Tuesday, Wednesday or Thursday of next week-- sometime near the evening rush hour on one of these days. We expect one count will take between one and two hours to conduct. If you’re interested in helping, please email mark.e.chase@gmail.com

For a video on the Pedestrian Head Start and how it works, please see this video: http://www.streetfilms.org/lpi-leading-pedestrian-interval/

Comments

[benndragon.livejournal.com]

I'm wondering how they will resolve the conflict between pedestrian head-start and initial green arrows (peds can't get across fast enough to not completely negate the arrow in a lot of intersections), particularly on busy roads like Mass Ave.

[ron_newman]

Seems like the answer to that is to have lagging instead of leading green arrows.

[aroraborealis.livejournal.com]

You don't usually put a green arrow in conflict with a WALK. The Leading Pedestrian Interval (LPI) is for more straightforward intersections where most of the vehicle traffic is moving parallel to the ped crossing. In this case, the LPI gives peds enough of a head start to "own" the intersection and be visible to those cars that are turning.

[dial-zero.livejournal.com]

Sounds great!

[mattdm.livejournal.com]

Don't think I can help, but it sounds great.

[in-parentheses.livejournal.com]

Sounds great! Just curious, though -- who's "we"? Is this something the city is working on, or is there a citizens' group trying to convince the city to look at it, or...?

[miss-chance.livejournal.com]

I'm wondering how they will resolve the conflict between pedestrian head-start and initial green arrows

Cambridge has several intersection with exactly that conflict, most notably at Mass Ave. and Prospect Street in Central Square. It's horrible for everyone.

[jadelennox]

I've personally, as a pedestrian, found Cambridge's streets much less safe for me than they used to be back when Cambridge had 4-way stop+walk light. But then, I press the button and wait in Somerville.

[jadelennox]

+1

Cambridge made the IMHO poor decision to switch citywide, regardless of traffic patterns.

[aroraborealis.livejournal.com]

In this case, "we" are the LivableStreets Alliance and WalkBoston, working in cooperation with the City of Somerville.

[aroraborealis.livejournal.com]

Yeah, that's a bad place for concurrent WALKs, because there's SO much traffic, with a lot of turning movements. That would be an ideal intersection, imo, for an all-ways walk, allowing people enough time to cross kitty-corner should they want to.

[miss-chance.livejournal.com]

Is it a problem that Mass Ave and th Westn Ave side of the cross street are so wide there? I'd worry about having elderly people getting kind of stranded in the middle of the road... there are a lot of people with cognitive difficulties in the area, and I picture the light changing and a driver who doesn't have a clear view of the whole intersection starts to make a turn, others see them moving and start going, and then next thing you know, there's some very scared and confused elderly person in bad trouble. Is there a max size that's recommended for an all-walk intersection?

[moria923.livejournal.com]

Could someone please describe this "green arrow" thing? I'm a blind person, and I don't know what they are or how they're affecting traffic. And I use that intersection *a lot*.

[moria923.livejournal.com]

My husband thorbol and I would be interested in knowing more about those organizations.

[aroraborealis.livejournal.com]

There are some pretty big intersections with all-walk phases, but the longer the crossing, the more time you need to allow for people to get across. I think the industry standard for calculating crossing time is 3.5 feet per second, but elderly people tend to need more time (4'/s), and that discrepancy becomes more problematic at wider crossings.

There are a number of alternatives, and obviously Cambridge has chosen to go with concurrent signals, and I presume they put a lot of thought into that choice -- it's a high traffic and (heh) centrally located intersection, after all. There are no left turns allowed from any direction there, and no green arrows for people turning right, which means that pedestrians SHOULD have the right of way, but it's such a high traffic intersection for people on foot, bike and car, that it's pretty tough to accommodate everyone with that.

[xlerb.livejournal.com]

It's a phase when traffic is allowed to turn left while the oncoming traffic that's going straight has a red light (and so does the cross traffic); that way the left-turning traffic doesn't get stuck waiting forever for a chance to turn.

[aroraborealis.livejournal.com]

There are actually no green arrows at that intersection, BUT:

At some intersections, cars that are turning may be given a protected turn, either during the regular green phase or before or after it. This protected turn is signaled by a green arrow in the direction they're allowed to turn (most often a left, which indicates that oncoming traffic has a red). In a well-signalled intersection, there should never be a green arrow conflicting with a WALK phase, because each of those should indicate that the user is being signaled for a low- or no-conflict movement.

[aroraborealis.livejournal.com]

LiveableStreets is at: http://www.livablestreets.info/ -- they advocate for a more multi-modal approach to transportation planning and spending

WalkBoston is at: http://www.walkboston.org/ -- we advocate for safe and welcoming infrastructure for people getting around on foot

I'm happy to answer more questions about WalkBoston or point you to someone who can help with more info about LiveableStreets, or you can check out either web page, too, at your convenience!

[rosiewoodboat.livejournal.com]

I agree. It IS kind of hard to understand why one would get rid of the "protected walk", especially considering that walkers trying to cross an intersection include the young and the elderly.

[qwrrty.livejournal.com]

Maybe that's why they're doing a study to evaluate the relative safety factors? That's just a guess, now.

[rosiewoodboat.livejournal.com]

That does sound like a guess, considering I doubt Somerville's study is going to advise what Cambridge has already done.

[aroraborealis.livejournal.com]

Right. The all-red, protected walk phase is theoretically safer for pedestrians, but only if people actually wait for the WALK. Because all-way stop phasing takes longer, people have to wait that much longer for their turn, and studies show that pedestrians won't wait. My anecdotal observation in Somerville supports this.

So! Time for real data: We're going to compare two intersections with exclusive phasing and two with concurrent phasing. We'll be looking for pedestrian compliance vs. jaywalking, and conflicts with cars (that is, sudden stops on the part of either car or ped to avoid an accident) to see how people actually move through the urban world.

[nathanjw]

Right, lots of people won't wait for the full cycle, so the overall safety benefit of all-walk is limited (It's definitely an annoying public-health kind of problem; the choice that provides the best overall results makes the situation worse for those who are following the directions correctly).

A version of LPI that I saw used elsewhere - Montreal, I think - coupled the leading pedestrian phase with a green-forward-arrow phase, and then switched to green-ball at the end of that lead time. It permits more traffic flow, but only if you can trust drivers not to take the right turn when there's a green arrow forward.

[aroraborealis.livejournal.com]

Oh, interesting! That would take some getting used to for drivers, but I could see it being great overall once people knew what to expect.

[rosiewoodboat.livejournal.com]

What about a protected walk as opposed to all-way stop? When a walk sign for a given cross-walk is lit, no car will be permitted to legally cross that walk. As an example, when walk light is lit, traffic would be prevented (by a red arrow) from making a right across the walkway.

That way at least the kind of universal understanding that a lit walk light means "I can cross without being run over" as opposed to the situation "This means I can cross and certain hazards are reduced, but there might still be people driving across the walk."

[oakenguy.livejournal.com]

Ohhh yeah. That's *exactly* the intersection I immediately thought of when I read this post.

[nathanjw]

Of course, the green-forward LPI has the disadvantage that you need 4-ish more signal heads per intersection to implement this, whereas Cambridge-style LPI "just" needs reprogramming of the signal controllers (not necessarily a trivial expense - it's possible that some of them aren't flexible enough to implement LPI and would need to be upgraded or replaced outright).

More generally, this is a good blog post on LPI vs. scramble crossings (and a good blog generally), with links to several others, good comment discussions, and an interesting analogy with transit trip frequency vs. transfers.

[littlecitynames.livejournal.com]

The one thing I would like most in the Boston area is what they do in NYC-- the pedestrian lights should be timed to stop blinking BEFORE the corresponding (parallel) green light turns yellow. As it is now, they stop blinking only as the yellow light turns red. It bothers me to no end because people from out-of-town (including me, when I initially moved here from NYC) never seem to realize that there is almost no delay between the solid hand and the green light for oncoming traffic.

[jesseh.livejournal.com]

Ah, but that varies widely, at least in Boston -- in my neighborhood, some do it one way, some the other, which is even worse.

[boblothrope.livejournal.com]

I can't stand all-way walks, as a pedestrian or as a driver.

I like concurrent walk phases, but I have a mild dislike for leading pedestrian intervals. I hardly ever see any pedestrians (besides myself) who pay attention closely enough to start crossing during those 3 seconds. So they end up being a small waste of time.

[boblothrope.livejournal.com]

The problem is you need a right-turn-only lane, so cars waiting to turn right don't block straight-ahead traffic. This either takes away a straight-ahead lane, or requires widening the road which makes things much worse for pedestrians.

See Arlington Center for an example of this setup causing huge traffic jams.

[ron_newman]

It would be better to get rid of the blinking hand entirely and have only the countdown timer. That way, each pedestrian can use her own judgment whether she can get across the intersection in the remaining time displayed.