The loneliness of a middle distance pedestrian

← Back to Forums


Benzo
Participant
#

How should it deal with peds? It should detect their location and wait until the intersection is clear before allowing traffic flow to restart in the directions peds are crossing.


rgrasmus
Member
#

They are working on it http://www.post-gazette.com/local/city/2013/11/25/CMU-helps-East-Liberty-run-smoother-pedestrians-next-CMU-s-E-End-traffic-effort-turns-to-pedestrian-safety/stories/201311250114

Also, Greg Barlow seems very responsive to concerns of pedestrians and cyclists. He seemed to like my idea of prioritizing pedestrian wait times during inclement weather. Also, check out one of their more recent papers http://www.cs.cmu.edu/~xfxie/paper/TRB14UTC.pdf


gjb
Member
#

We have always wanted to get feedback on the operation of Surtrac, particularly where there are specific problems we can address. It is most helpful when we can get that feedback directly so we can have a conversation to gather more information, rather than getting complaints third hand or on social media. We usually get to see 311 reports, but it sometimes takes awhile for them to get to us, so if you’ve got an issue you think we should know about, we’re going to be able to address it fastest if you send it to us at the same time you send it to 311.

One of the things I’m excited about with the expansion along Baum and Centre is that the radar detectors that will be installed do a much better job of detecting bicycles than the cameras in East Liberty (which were installed before our project started). If you find that the cameras aren’t doing a good job detecting bikes in specific places, please let me know, I can probably do something to improve it. We have been interested for a long time in doing a test deployment of sensors designed specifically to detect bicycles and pedestrians, but we haven’t had much interest from BikePGH or the city. Our signal optimization already includes pedestrians, but suffers from a lack of sensing, since we only have the buttons to go on.

Feel free to get in touch with me. I’m happy to setup a call or meet in person as well.

Greg Barlow, Robotics Institute, Carnegie Mellon University


jonawebb
Participant
#

You know, that’s the kind of response that makes a man reconsider his complaint.


Benzo
Participant
#

Cool. I have 311 on speed dial (Actually, I have a bookmark to the web form which I tend to use frequently). I’ll be happy to report any weirdness with respect to bike/ped stuff through this channel.

Greg, thanks for reading and commenting.


xfxie
Member
#

My apologies for a late reply, but I did not notice the new discussions in this thread until today when I was trying to collect more feedback and suggestions for improving the experiences of pedestrians and cyclists. As the creator of the core control algorithm of SURTRAC, I am sorry for the frustrations caused by our system on pedestrians and cyclists in this neighborhood.

At the time of starting the control algorithm, its main objective was to reduce drivers’ travel time, fuel consumption, and vehicle emissions to our environment, which were all from a vehicle-centric view. Started from around one year ago, I have been working on improving the broader mobility of other modes of traffic that are central to sustainable urban living, particularly pedestrians. As noticed by rgrasmus, the control algorithm has incorporated pedestrian pushbutton information, to improve pedestrian experience at some intersections, such as Penn Ave and Eastside. Several months ago, the control algorithm has been further generalized to optimize the delay tradoffs between pedestrians and vehicles. This new version has shown its capability of significantly reducing the pedestrian wait time.

I found many helpful discussions in this thread. Next step, Greg Barlow and I will try our best to solve some issues that were raised here. For the following months, we might test different versions to improve pedestrian experience, and we are expecting timely feedback and suggestions from you. One possible communication mode is to use this mailing list: surtrac-user@lists.andrew.cmu.edu, for receiving announcements of new changes and for sending your experience and feedback. In the meantime, you are of course welcome to drop us emails.

Xiao-Feng Xie, Carnegie Mellon University


Steven
Participant
#

FWIW, I was a pedestrian at Penn and Centre on Saturday. I eventually crossed three sides of the intersection, and waits were short each time, with the longest maybe 45 seconds or so.

We’ve been unwitting beta testers for a system that has a lot of promise, and naturally that experience has gone about as well as any other beta testing. :-)

It’s a bit worrying that the city’s talking of expanding the system before every issue has been worked out, but given the lead times for construction I guess it makes sense. Hopefully it’ll be issue-free for bikes and pedestrians before too long. Thanks, CMU folks, for posting here; that’s very encouraging.


edmonds59
Participant
#

“Reducing drivers travel time”, i.e., facilitating driving, and “reducing fuel consumption and emissions to the environment” are contradictory goals. You are, apparently unwittingly, being suckered into the same logic vortex that the highway builders of the ’60’s/’70’s pushed onto society – “More highways! Less congestion!” We are only now beginning to crawl our way out of that hole. Facilitating driving does not reduce traffic, it increases it. You are selling snake oil.
If you sincerely want to work toward the goal of reducing fuel consumption and emissions to the environment, you do not work on facilitating driving. You work on something that will make mass transit more functional and appealing to the American public.
Thanks!


jonawebb
Participant
#

I think Edmonds makes a very good point–even though I see that the CMU folks are fixing the bike/ped problems. If you effectively increase the capacity of the road by making the traffic control system smarter, what will happen is more people will drive. Wait times and emission will increase, not decline. This has been shown over and over. Cars just aren’t that efficient a way of moving people, compared to the alternatives.


StuInMcCandless
Participant
#

I am pleased to see that the CMU folks are following this discussion. To make analogy to naval warfare ~1800, they’re coming up with better cannonballs and powder, but someone else is still aiming the cannon and deciding what to aim at and when to fire.

I’m really happy that we have better cannonballs and powder, but I also really wish we could get the decision makers to pay better attention, too. Better traffic through-put is only OK *if* we can *first* not scare cyclists away from using the streets and pedestrians from crossing them.

Better signal technology does not translate into “leave the damn car home”.


rgrasmus
Member
#

Along these lines, does anyone know if this is still in the works:
“Stephen Patchan the bicycle coordinator for the city came to us to speak about creating a bike lane on Centre Avenue. This proposal would take away parking on one side of the street. This is in the early stages and he will come back as it develops further.”
This is from the 2012 Shadyside Action Coalition meeting minutes http://www.shadysideaction.org/index.php/meeting-info/2012-agendas-and-minutes/


Steven
Participant
#

Reducing transit time benefits buses as much as cars. And a smart traffic signal offers the possibility of detecting and prioritizing buses over cars. It’s possible the net effect of smarter traffic signals on a corridor like that one with frequent bus traffic is to make transit more appealing by making it faster.


jonawebb
Participant
#

@steven I think the issue is that the research appears to be vehicle-centric (even though it breaks out buses separately and could treat them as a special case). If you start from the point of view of moving vehicles more efficiently, as opposed to moving people more efficiently, you end up being very sensitive to changes in motorist behavior. And even if you try to give buses priority it’s not going to make much difference so long as buses and cars share the same lanes.
Suppose you started from the point of view of “what can we do to move more people in to work along the Penn/Center corridor”? I think the first thing you would try would be to get drivers out of their cars, which take up so much space, and into more efficient vehicles. You’d be trying different incentives to get them to change their behavior. You might be thinking about BRT, or dedicated bus lanes, or bike share, etc. You might even try to increase the inconvenience of driving, while offering a viable alternative.
I don’t see how you would come to the conclusion that reducing driver wait times at signals would help that much.


Steven
Participant
#

Well, I don’t think the goal is moving more people to work along that corridor, but moving the same people to work faster, as well as reducing pollution. (If we can’t agree on the goal, we’re not going to agree on how to get there.)

Even if we do a great job encouraging non-car modes, we’ll still have lots of people commuting by car. It will take decades to even make a dent in that, getting people to abandon the suburbs and move into the city while we build vastly better transit facilities. And with our hills, there will always be a very high percentage of people who just aren’t going to bike to work.

This says about 71% of Pittsburghers drive to work now. Suppose our goal is simply to reduce pollution, and we have the option to make all their trips 10% longer. That should decrease the 71% figure, if transit and biking are still just as fast, but by how much? If those 10% longer trips push an extra 1% of our population to use bikes or buses (to 70%), we’ve made a lot more pollution. If the longer car trips knock that 71% number down to 60%, we’ve made pollution better. (Well, probably: I’ve read that buses are so inefficient that moving people from cars to buses doesn’t do all that much to lower emissions. So it might depend on how many of those former car users switch to bikes or walking.)

Likewise, reducing drive times will directly reduce pollution and average commute times. Indirectly, it could get more people driving, but (I’m guessing) not enough to outweigh that.


Ahlir
Participant
#

Other trends are at work. For example, people are moving back into cities.
When they do they won’t need to drive as much.The sensible thing is to make this migration easier, by improving public transit and by making cities more accommodating of bicycles and other alternatives (inclines anyone?)

Transportation is expensive and people notice. Give them incentives to change their habits and they will do it. A gallon of regular gas is currently ~$8 in France, higher elsewhere. If our gas was priced more realistically there would be even more incentive for people to do the “right” thing.

We should all have access to a car when we need one. But we oughtn’t need a car just to get around in our daily life. That’s just bad transportation planning.

[grumpy postscript:]
Allegheny County should introduce a “pothole surcharge”, say $15 per car (less than 1/2 a tank!), more for heavier vehicles. How could anyone possibly object to attacking this cursed blight? (This is after having to deal with the non-solution of cold patch: are we all supposed to buy jumbo-fat bikes just to be able to make it down the street?)


jonawebb
Participant
#

I see the point of the research. It’s a good idea to try to make better use of existing road infrastructure.
I don’t see that in the particular case of Penn Ave, things are going to work out the way that’s being reported. Increasing throughput is going to lead more folks to drive, the way it has in the past.


Ahlir
Participant
#

Furthermore: City speed limit needs to 20mph, strictly enforced.

I will admit that I exceed this limit on occasion (on my bike), but I can live with slowing down, or suffering the consequences. For the public good.


czarofpittsburgh
Member
#

I’m not sure if I’ve seen these issues addressed but now that we have the designers here…

1) Is there a mechanism to time the lights such that the speed limit is enforced? Ideally, a car traveling well-above the speed limit between lights should be stopped by frequent red lights. It should only be possible to get from one intersection to the next with greens at both by going approximately the speed limit.

2) Will the large amount of data be analyzed? Many people here predict that if the travel times along this corridor are significantly reduced, then more people will start driving there and the congestion will go right back up and cancel out any benefits. This system is trying to turn a surface street into an alternative to the Parkway. If it works, more people will see it as an alternative and drive there. The data should bear that out, and I hope it will be monitored.


Steven
Participant
#

Allegheny County should introduce a “pothole surcharge”, say $15 per car (less than 1/2 a tank!), more for heavier vehicles. How could anyone possibly object to attacking this cursed blight?

Would the state let us? They seem to heavily restrict the types of taxes localities may use. And since the state keeps track of car registrations, not the county, I imagine they’d have to implement it too.

But we oughtn’t need a car just to get around in our daily life. That’s just bad transportation planning.

It goes beyond transportation planning, though. It’s very expensive to run mass transit to the suburbs. I’m not sure there’s a way to fix that without increasing population density. And most of those folks aren’t willing to bike 40 miles each way to work, even with ideal bike facilities.

Part of fixing this is getting people to move back to the cities (which of course they are now, a little, and we’ll have to see if that trend continues to the point where it has a real impact). But making that happen in a big way requires more than just transportation planning.


salty
Participant
#

Poor transportation planning is exactly what created the suburbs in the first place. People are moving back to cities because they’re starting to realize that there is no real solution – as you said, transit and cycling are impractical. Unfortunately too many people don’t shed their car dependence even when they move here, and of course we still have suburbanites by the gazillions coming and clogging up the roads as well (and complaining about potholes they help cause but don’t really help pay for, but…).

Back on Surtrac, I do appreciate Greg and Xiao-Feng posting here – I know I tend to rant at times but it is frustrating not to be heard. Like many people here, I also do not agree with the premise that moving more cars is a reasonable goal. Leaving the larger philosophical discussion aside for a moment, the practical issues are what frustrate me (and many others) on a daily basis – mostly the long signal times, with the short or nonexistent walk cycles being a close second. These issues are not technological, they are policy issues. I believe light cycles should be capped at 60 seconds, and should be much shorter during non-peak times. I also think when the green light is on, the walk signal should be on for the same amount of time, regardless of whether anyone pushed a button or not. I understand doing these things will lower traffic volume, which is why I think that is the absolute wrong metric.


edmonds59
Participant
#

Everything this Rube Goldberg system hopes to accomplish can already be done with a properly programmed system of pavement detectors, pedestrian pushbuttons, and timers, EXCEPT, ironically, cyclist detection. Which does not seem to have received any priority whatsoever, initially.
In urban conditions, pedestrian “friendliness”, safety, and functionality needs to be given priority one. This needs to be clearly communicated to your clients. Until that takes place, your project is misguided exercise in building a better mousetrap. Or, Monorail, as the case may be. But hey, good for you for finding a way to take home some cash from an already strapped city.

The initial failure of your project was to take the clients goal – “Move traffic!” – at face value, without fully evaluating all the all the associated parameters. Such as, oh, pedestrians. Like setting out to design a really energy efficient building and forgetting toilets. Or not considering that an O-ring might need to continue to function at sub-50 F temperatures. Poor engineering practice.


jonawebb
Participant
#

Ha, the quote button doesn’t work anymore for me on Chrome. Interesting.
@edmonds: EXCEPT, ironically, cyclist detection
Actually, the magnetic systems developed to detect cars can detect bicycles if you ride on the detector line (and you have metal rims). They are designed to do that. There are stencils that can be used to mark where to ride if the line isn’t visible as a cut on the road.
Experiments on real systems have shown that properly adjusted and maintained systems can detect bicycles. But sometimes the sensitivity is not adjusted correctly, and other times the system is just broken.


Ahlir
Participant
#

@Steven:
1) It’s not a tax, it’s a surcharge, or maybe a user-fee… mustn’t use words like “tax”.
2) The idea would not be to make transit and biking easier out to the suburbs, it’s to make it better in the cities so that people have reasons to migrate inwards (in addition to the amenities). This is where the good planning comes in.

On a separate note:
Many streets do seem to have sync’d lights: Liberty from the Bloomfield Bridge to the Strip (~30mph) and the drive between Homestead and Squirrel Hill (also ~30mph). Of course the signage doesn’t quite tell you this.


StuInMcCandless
Participant
#

…or not considering that an O-ring might need to continue to function at sub-50 F temperatures.

For those of us not getting the reference, see “Challenger disaster”.

Back to the present. Almost. An early form of speed control via traffic lights I saw in use one time was in Buffalo NY, circa 1979. It was a simple, standard traffic light, in the middle of a block, that cycled every 15 seconds, stayed red maybe one second, then back to green. Street posted 25. With over 200 yards of sight line, it caused people to slow down to 25 and stay that speed. I’m sure the technology has improved markedly in 35 years, but it’s useful to have seen this in practice at least once in my life.


gjb
Member
#

But hey, good for you for finding a way to take home some cash from an already strapped city.

Just to clear this up, since there may be some misperception, the city of Pittsburgh isn’t paying for any part of the Surtrac project. Quite the opposite, by the time the expansion along Baum and Centre is complete, we’ll have donated just shy of a million dollars worth of detection, communication, and other traffic signal control equipment to the city.

The Baum-Centre expansion of Surtrac is piggybacking on a signal retiming and limited equipment upgrade project that the city and the Southwestern Pennsylvania Commission (SPC) already had on the books. We put together the money from the foundations and UPMC to upgrade all the intersections that are part of the project with the communication and detection the city prefers.


jonawebb
Participant
#

@steven, BTW:
Well, I don’t think the goal is moving more people to work along that corridor, but moving the same people to work faster, as well as reducing pollution.
I was assuming what the end result will be. I predict traffic will shift from unoptimized routes onto this one, or motorists who have given up on driving to work will take it up now that the route is faster. So you end up moving more people on the same route, even if your original goal is to move the same people faster, making that the realistic goal.
This is an empirical question, which I hope the research will measure; but there’s lots of previous experience suggesting that is what will happen.
And, BTW, if your goal is to either to move more people or to move the same people faster, getting people out of their cars is the way to do it. If you could get a fraction of the people on Penn Ave to use more efficient modes that would mean less congestion for the rest, and faster travel times, even without optimizing the signaling.


StuInMcCandless
Participant
#

Let me take an auto-centric approach to addressing this problem. OK, given you want to both reduce pollution and increase throughput, but not attempt to reduce driving, how about this: Assume everyone is driving a manual shift vehicle. If you can get people from one end of your study area to the other without using the clutch once, you will achieve exactly the same goal, and also benefit what we cyclists want, too.

It’s the same thing as trying to get down the Parkway East at 7:15 a.m. without touching the clutch once (which I’ve done). Go at a very steady, even speed, even if a big space opens up in front of you. As if to say, “I don’t care how slow you go, but don’t slow down.” Be looking six to 10 cars ahead, and match that car’s speed, not the one in front of you.

Net effect, as applied to ELib, is nobody travels more than 15 to 20 mph, ever, but also very infrequently stops. Cyclists can easily ride with auto traffic, and pedestrians can get across the street at marked spots without unreasonable delay. Car pollution is minimized since there will be many fewer speed changes or stops.


Mick
Participant
#

StuInMcCandless wrote:

Net effect, as applied to ELib, is nobody travels more than 15 to 20 mph, ever, but also very infrequently stops. Cyclists can easily ride with auto traffic, and pedestrians can get across the street at marked spots without unreasonable delay. Car pollution is minimized since there will be many fewer speed changes or stops.

That is a great idea, and I like it a lot.

Trouble is, for almost all drivers, this mean “Go 35 in a 25 mph zone and stop at every light for a few seconds.”

Signage might make good driving strategies more clear, but I suspect clarity isn’t the issue. Example: what part of “Speed limit 25 mph” on those Greenfield Ave signs is unclear?


edmonds59
Participant
#

“We put together the money from the foundations and UPMC to upgrade all the intersections that are part of the project with the communication and detection the city prefers.” That’s too bad. So realistically you contribute very little to the overall vision.
I’ve got a better idea. Red lights for vehicles at all intersections, all directions, at all times, with detection devices of your preference that decides when a vehicle arrives to be let through, and pedestrian push-buttons with automatic override. Make it safer, more pleasant, and more convenient to walk a mile than to drive a mile, and inefficient, short-hop driving is reduced or even eliminated. Suburbanites using city streets as tertiary highways are discouraged. The entire sense of the urban environment changes.
Sadly, the situation described above is exactly what we have established for pedestrians. Our society has things exactly fucking assbackwards.


Ahlir
Participant
#

1) Surtrac is not a product, it’s a research project. The most valuable thing about it is the data it’s generating. and the experiments it can support. There will be enough knowledge in the end to support sophisticated modeling and the consequent ability to influence driver behavior according to reasonable policies.

2) Just to repeat: there’s nothing wrong with cars. It’s just that people tend to use them for all the wrong things (like everyday commuting). Offer alternatives and they will choose the right thing.


Benzo
Participant
#

Realistically, I own a car and ride a bike. I probably ride a bike more than I drive, since I use that for commuting and leave the car at home.

I’m just as interested in not getting stuck in traffic jams on both forms of transportation. If surtac can improve traffic flow, prioritize mass transit, and create safer conditions for pedestrians than current system I’m all for it.

It really amounts to what goals does the city and the designers of the system want to include and prioritize. The nice thing about this system, is that the software can be updated a bit more dynamically than the existing system (I assume), so the optimization can be adjusted to better suit different goals as the system grows and it’s capabilities are expanded.

I see this as a huge win if they can prioritize bus traffic at signals, especially if the system gets more widespread distribution.


jonawebb
Participant
#

@ahlir, SURTRAC is a spinoff, so a product and not just a research project. Nothing wrong with that, really, I guess, though I suppose if it turns out that optimizing traffic signals doesn’t reduce congestion some folks are going to be pretty disappointed. They’re not all going to be saying “that’s an interesting research result.”
Their web site http://www.surtrac.net/ has lots of info and links to articles. They might be overselling the system a bit. The article from Business Insider is titled “18 Brilliant Ways To End Gridlock And Save Billions.”


rgrasmus
Member
#

@benzo – I agree completely! I couldn’t have said that better.


StuInMcCandless
Participant
#

@benzo – It really amounts to what goals does the city and the designers of the system want to include and prioritize.

That’s what I was getting at when I said “someone else is still aiming the cannon and deciding what to aim at and when to fire.”

The technology may be great and wonderful, but it still has to be used by somebody for an intended purpose. My beef is with the purpose, not the technology. We still need to get people to
a) leave the car home,
b) to drive around not through this area if they’re going beyond it,
c) to expect to move at bicycle speed if they are going to go through here.

Lower the pollution counts and delays by a, b, and c, not by making it easier to drive through ELib.


WillB
Participant
#

(Quote function also no longer working for me in Chrome)

@stu “b) to drive around not through this area if they’re going beyond it,”

This is basically the logic behind Penn Circle (and many highways) and has proved to be deadly to vibrant business districts. One of the things that gives this corridor such economic potential is that so many people pass through it. I agree that they ought to be doing it at a slower top speed, but let’s not throw the baby out with the bathwater.


StuInMcCandless
Participant
#

I meant to not drive through ELib if all they’re doing is trying to get from Monroeville to downtown. There’s a perfectly good interstate highway a mile or two away. Use that, not blow through here as a supposed shortcut.


WillB
Participant
#

I know, and that’s what I was responding to. One of the things business districts need to succeed is for people to know about them. When people take highways instead of surface roads, that means that they don’t see (and occasionally stop at) the businesses. This is not to say that we want the whole of the Parkway East driving down Penn, just that we don’t want to get rid of too much traffic. If the only people who drove down that stretch of Penn were people going specifically to East Liberty, I think that would be very bad for East Liberty.


Benzo
Participant
#

We should just build the mon-fayette expressway. That will take congestion off these surface streets!

(not really, please don’t)


Marko82
Participant
#

This could be our chance to influence the rest of the country.

Intelligent Transportation Society to have 25th annual convention here next year

Read more: http://www.post-gazette.com/local/city/2014/06/04/Intelligent-Transportation-Society-to-have-25th-annual-convention-here-next-year/stories/201406040175#ixzz33xdKJZqZ


dfiler
Member
#

Is the east liberty portion of penn ave a penndot road? The reason I ask is that it seems PennDOT has zero concern for pedestrians, bikers, and the neighborhood through which roads travel. Their only concern appears to be maximum car traffic.

The most obvious examples of this are the new lights installed on Penn between braddock and 5th. Refrigerator sized signal control boxes were installed in the middle of the sidewalk. To make matters worse, the new poles are set back from the road, also blocking the sidewalk. Meanwhile, the old poles are still there, contributing to the obstructions. Pedestrians have to walk in the road or turn their shoulders sideways to get through the worst of these blocked sidewalks.

It is pretty comical to see a ADA compliant curb cut for wheelchairs and strollers… leading to an impassible sidewalk.

Put simply, the PennDOT employees responsible for overseeing that job should be fired.

← Back to Forums

You must be logged in to reply to this topic. Click here to login.

Supported by