NextBurgh, PA          @VannevarB      About      Pittsburgh Murals
May 02, 2014

Faster Isn't Smarter; Cars Aren't All The Traffic. Did We Think About This? Were We Smarter?

From the Post-Gazette and the excellent Jon Schmitz:
City expands use of high-tech traffic signals
Officials, CMU students team up again to reduce red-light wait times
"Imagine if traffic signals had the ability to see the traffic that was there. Then imagine the traffic signals had the ability to talk to each other, second by second," Mayor Bill Peduto said at a news conference. "It's not something from George Jetson. It's something that's happening right here in Pittsburgh today.

The computer formulates a plan based on traffic observations and communicates it to neighboring intersections. It can instantly adjust to accommodate surges caused by disabled vehicles, crashes, street closures or bursts of traffic after events.

CMU will work next on ways to improve pedestrian and public transit flow. Mr. Peduto said he expects further expansion of the signal network as well.

"We're in a part of the city that's growing very quickly," he said at the conference, held at Morrow Park in Bloomfield.

"We don't have the capacity to tear down houses and make wider roads" to speed traffic, Mr. Peduto said.

"We do have the ability to think about it and do it smarter."

Here's the problem:
  • traffic is not just cars
  • faster cars is not smarter traffic

The reduced wait-time for cars comes at the expense of reduced crossing time for people that aren't in cars - you know, people who are walking - some of whom are old, or in wheelchairs, or school children. You know: people - as opposed to cars. This system has sensors for cars and not for people, and this system prioritizes cars over people.

Cities around the world are realizing that the key problem in car-human interaction is: car speeds. You've probably heard of Twenty Is Plenty, which points out that human survivability of car-human collisions are highest at 20 mph - and increasing the speed of cars has the direct effect of increasing the deaths of the people that cars hit.

Look at this chart, and explain why any elected official is trying to speed up cars?

You may have also heard of cities implementing Vision Zero, (San Fran, New York City) which holds as its central tenet that there is no morally acceptable way to cost-justify human death in traffic fatalities - in other words, it is unacceptable to consider killing people a cost-justified expense of the normal way of life.

In a nutshell: Mr. Peduto's statement that we can think about this and be smart about this provides the answer to his own mistake: speeding up car traffic, and hindering pedestrian and bicycle traffic, is a dangerous change that shows the cultural dominance of car culture and our unwillingness to accept empirical safety data.

2 comments:

Bram Reichbaum said...

My first question is, whether or not these devices come with immutable factory settings which indeed can only accord motorists very high or maximal priority, or whether once a City purchases them it can tinker with the preferences in such a way that non-motorists are accorded mandatory (high) minimum consideration as well. In other words, if they can't sense pedestrians, they can at least "blindly" give them a nice extended Walk signal every cycle, and when the traffic lights do change do to the genius sensory gnomes, there should be a nice healthy All Points Red interval every time regardless.

I don't like to jump to the conclusion that street traffic (or anything else) is a zero-sum game. If "old-fashioned" traffic signals are based on arbitrary guesswork and nothing at all, I should think it's very possible that a smart enough machine could improve traffic flow for motorists, cyclists, and pedestrians all at the same time.

And I don't know whether the aim of decreasing motorist wait time is to get the cars up from 20 to 40, or rather to get them up from 0 to maybe 10 during rush hour. When signaling is particularly bad, cars tend to stack in a gridlock -- you can't move during a green light because the following light is red, and so on and so on. (And if there's enough 0 mph going on, you end up with a lot of angry motorists on "tilt" seizing slender opportunities to dart ahead, which is its own safety hazard.) So when we're talking about speeding vehicle traffic, I don't know what the goal is.

Basically there's too much I don't know about this yet to jump to any conclusions, but I wouldn't be comfortable with a blanket proposition that we should never do anything to improve automobile traffic flow. In terms of "traffic calming" I'd rather see things like speed bumps / humps / tables or new roundabouts, and in terms of discouraging driving in favor of other forms of transportation I'd rather see a congestion tax, emissions standards / taxes and dynamic parking prices, than simply making driving stink by failing to discourage gridlock.

tonycpsu said...

Several of my coworkers and I had a pretty lengthy discussion about this yesterday afternoon when we read this story, and I have to say I'm cautiously receptive to this project.

Let me start by stipulating that I'd happily support a 25mph or even 20mph speed limit on the entirety of Baum, Centre, and Penn along this corridor. Having said that, I think it's important to keep in mind is that during peak periods, nobody is coming close to those limits, so the goal (from a vehicular traffic and emissions point of view) is to minimize the number of people going at 0mph because they're at a red light. The article points out that this is about "wait times", not maximizing speed, and I think that's the right thing to focus on.

Nobody, not even pedestrians, are helped by an intersection that has a green with no cars going through it and a red with cars stacked up. The sensors are just there to manage these situations by detecting vehicles and adjusting the signal times. That's it.

The one place where I'll admit it gets a bit hairy is where the signals are giving drivers a green arrow to turn into crosswalks that have walk signals illuminated for pedestrians, like the intersection in front of the Target. In this situation, drivers are always supposed to yield to pedestrians in crosswalks (just as they must do when turning on green), but a lot of idiots who don't understand the law see the arrow as an unlimited right of way. Perhaps some better signage reminding drivers to yield to pedestrians could help make this less of a hassle, but if this continues to be a problem, I'd be okay with eliminating this situation by making the green arrow a red signal, giving peds exclusive access to the crosswalk there.

Anyway, I guess what I'm saying is that, yes, we don't need anyone in that area traveling faster than 20mph, but I think we do want vehicles traveling faster than zero, providing pedestrian safety can be protected.

(Full disclosure: I work at CMU, albeit in a department completely unrelated to this project.)

Post a Comment

Comments and Feedback? Love that stuff. Please leave your thoughts in the box below--