Friday, July 24, 2015

ODOT Responds to new Dead Red Law with Detection Archeology

Getting a Green Light
ODOT pamphlet
Though we didn't dwell on it here, BikePortland readers may recall that ODOT formally testified in opposition to Senate Bill 533, which now that it's been signed into law, will permit a person on bike to proceed at a light if they wait one full cycle and the controller is broken or doesn't recognize them :
Interestingly, the Oregon Department of Transportation opposed the bill, as did the Governor’s Transportation Safety Advisory Committee. They fear the law will lead to deadly consequences for inexperienced riders who may not make good judgments on when it’s safe to proceed. Instead of a law, they would like to see a complaint-driven [process] where people could call ODOT and have signals fixed.
Now that this will be permitted by law, ODOT may be cranking up the PR machine to try to obtain the "complaint driven" process they wanted in the first place and to discourage people from using the law.

Folks from ODOT are sharing a pamphlet "Getting a Green Light" and asking folks to call ODOT to complain about apparently non-responsive lights on the ODOT system.

Which. You know. Fine. There's nothing really wrong with that.

At the same time from the standpoint of a person on bike: Too many intersections are complicated and confusing!

Here's something like the process of reading an intersection that ODOT would like to see:
  1. Is there a rectangular induction loop? If so, position you and your bike over it...
  2. if not, Is there a circular induction loop? If so, position you and your bike over it...
  3. if not, Is there a video controller? If so, locate it and position yourself in the bike lane or behind the stop bar...
  4. If two or three of these are present...Wait, what? (Yeah, many intersections have multiple layers of "detection," and it's not always possible to determine which is the active detection system. Sometimes cuts for circular loops directly overlap the cuts for rectangular loops, and it's a puzzle where to position you and your bike. In the face of sign clutter, it is not always easy to find the video detection. There are many intersections in Salem with all three: Rectangular cuts, circular cuts, and video cameras. ODOT's desired process means each intersection has to be practically interrogated for a detailed analysis by a person on bike.)
  5. If you have waited for more than two minutes, consider moving over to allow a car to move into a detection zone or go to the sidewalk and press the pedestrian push button.
  6. (Whatever you do, don't use the law now passed for your benefit!)
And here's a real life example. Southbound on Winter at Bellvue/Pringle Parkway. (I may go back and take a photo specifically to illustrate this, but one I have from the bridge demolition detour will do for the moment.)

Winter St. Southbound at Pringle Parkway (Bellvue)
It has all three: Rectangular loops, circular loops, cameras
Enlarged detail with some of the loops outlined
(I think there are others!)
How is an ordinary person on bike supposed to read this intersection (and others like it) for the first time and position themselves, confident that they'll trigger a green?

I think the actual detection system has been converted to cameras here - but if you don't follow traffic signal upgrades, how would you know this? As intersections are upgraded, there isn't an "erasure" or "cancellation" process that marks or eliminates now-obsolete and non-functional detection system layers. And Salem, anyway, hasn't adopted a bike detection alignment guide paint scheme that routinely shows the latest detection system layer. Also, sometimes loops are also in bike lanes and sometimes not. (I can't remember if there is a loop in this bike lane here.)

I think the ODOT process is a little ridiculous. At the very least it is unnecessarily complicated and confusing and relies too much on specialized, professional knowledge.

It is also another instance of prioritizing auto traffic flow at the cost of inconveniencing people on bike in ways we would not inconvenience other road users in cars. Imagine saying to car drivers that they needed to fiddle with the position of their car at an intersection and maybe have to get out to activate a pedestrian signal button on the sidewalk.


And it's all wrapped up in the mantle of "safety."

In the pamphlet ODOT says
Bicycles are considered vehicles in the Oregon Vehicle Code, so riders must wait for a green light before proceeding through a traffic signal.
It's true that bikes are considered vehicles - but it should also be true that getting a green light is as simple for people on bike as it is for people in cars. If it's significantly more complicated for people on bike, that is a sign that maybe bikes aren't really considered vehicles and that the "Level of Service" for people on bike probably merits a low grade.

The solution shouldn't be for ODOT to require people on bike to read an archeology of layers in roadway design and engineering - though of course being able to do this is always a help in advocacy and developing more cycling skills - the solution should be for ODOT to make it easy for people to bike and easy for people on bike to navigate traffic control.

If we want to encourage people to bike, we should simplify, not complicate. ODOT's response to SB 533 in the form of "secret adept detection archeology" is an autoist move in the wrong direction.

I'm glad there's this law. Especially at night, I don't want to fiddle around, stationary and vulnerable, on a stale red at an intersection on a deserted street. ODOT maybe fears the use of the law at rush hour, but I bet it's mostly welcomed and used during off-peak hours.

Remember this propaganda? It's also about getting the green...

Give Yourself the Green Light via the Atlantic and Prelinger Archives


Anonymous said...

Here's BikePortland's piece -

Closer maybe to a straight news story, but still a little dubious.

They also cite a study -

So a real question is whether this is because of ignorant riders and we just need better education, or whether this is a sign of pervasive design flaws and poor user interface.

Laurie Dougherty said...

How do the street loops work anyway? I once heard that they were metal detectors and wouldn't work with carbon - not that that's my problem. But I've got both steel frame and aluminum bikes and never can get those loops to turn a light green

By the way, do you know what all the painted marks on the street are at 17th & Mill? I'm hoping for a pedestrian signal. I go through that intersection by bike or on foot a dozen times a week or more and it can be very hard to get across at times. I think they just showed up today; I didn't notice them yesterday.

Salem Breakfast on Bikes said...

As I understand it, a pedestrian median and crosswalk is going it at 17th and Mill - but not a signal or flashing beacon.

Regarding the loops, I don't know. I have one intersection I use regularly that is tuned great. Without needing to be too careful I position both steel and aluminum frames over the outer third of the circle, and it triggers things promptly.

There are locations where I feel like a rat in a random reward trial: Sometimes I think I trigger the light, other times I don't think I do, and I can't discern a pattern.

(Thanks for the BP link, anon.) The commentary at BikePortland seems to focus on two distinct improvements in the user interface: Road stencils to show the active area for best positioning, and a blue indicator light or other feedback mechanism to show you that the signal is reading you and your bike.

Even with very imperfect detection technology, an indicator that the signal cycle has been triggered would be a great boon.