How can I get my city to improve traffic signal timings?
October 28, 2012 4:20 PM Subscribe
I live in San Diego, and I hate the traffic lights along a particular road that I always drive on (Genesee Avenue). I'm regularly astonished at how soul-crushingly stupid they can be, and I'm constantly yelling at them from my car like an idiot. What can I do about this? My dream would be to talk to the traffic engineers in my city, really pick their brains and offer my observations. How does one go about making this happen?
I've tried calling the traffic division to report a "problem", but nothing ever gets done. Now, I realize that the timings are not an easy or straightforward thing to work out. The timings at an intersection have to work with all the other intersections. Or whatever. Maybe they don't... at least it seems they don't sometimes. I can understand tweaking the timing at an intersection so that the next block doesn't overflow with cars all stopped at a light, and things like that. But there are instances where a light definitely does not need to cycle to red (i.e., there are no cars or pedestrians, and the light following this one has just turned green); and there are instances where lights don't need to stay red for so long (i.e., a left-turn signal that stays red for about 2 minutes on a timer, even when there are no cars anywhere and it's just me and maybe 1 other car behind me). Then there are the left-turn lights that stay green for no one (thereby blocking the traffic that wants to go straight from the opposite direction), and then inevitably turn red just as a car is approaching to make the left turn. I'm also aware that these things are easier to notice because I hate them and they're stupid, and that I might not be noticing when the lights DO work. I feel like I've accounted for that. The lights that are smart ARE good, and I love them. But the stupid things I notice are repeatable, and happen consistently.
My favorite is a stretch along the UTC block where, if you start from a red light at Nobel and head towards La Jolla Village Drive, there will be NO CARS in front of you, and the light at La Jolla Village Drive will stay green (again for no one but you), but once you get to the light, it turns red. I always have to go like 65mph (limit is 45) in order to make the light. I swear it is taunting me.
OK, ranting aside... I would gladly volunteer to do simulations and work out scenarios for different amounts of traffic and etc etc, but I am at a loss for how I can offer this help. What would be awesome is if the city were like, "OK, here's how the intersections are laid out, here's where the sensors are, blah blah blah. Who can come up with the best timings?" and have people submit what they've figured out.. cuz whatever they're doing now can probably be improved. Do they gather real-time data? Do they measure how much waiting each person does? What are all the factors they take into consideration?
I've tried calling the traffic division to report a "problem", but nothing ever gets done. Now, I realize that the timings are not an easy or straightforward thing to work out. The timings at an intersection have to work with all the other intersections. Or whatever. Maybe they don't... at least it seems they don't sometimes. I can understand tweaking the timing at an intersection so that the next block doesn't overflow with cars all stopped at a light, and things like that. But there are instances where a light definitely does not need to cycle to red (i.e., there are no cars or pedestrians, and the light following this one has just turned green); and there are instances where lights don't need to stay red for so long (i.e., a left-turn signal that stays red for about 2 minutes on a timer, even when there are no cars anywhere and it's just me and maybe 1 other car behind me). Then there are the left-turn lights that stay green for no one (thereby blocking the traffic that wants to go straight from the opposite direction), and then inevitably turn red just as a car is approaching to make the left turn. I'm also aware that these things are easier to notice because I hate them and they're stupid, and that I might not be noticing when the lights DO work. I feel like I've accounted for that. The lights that are smart ARE good, and I love them. But the stupid things I notice are repeatable, and happen consistently.
My favorite is a stretch along the UTC block where, if you start from a red light at Nobel and head towards La Jolla Village Drive, there will be NO CARS in front of you, and the light at La Jolla Village Drive will stay green (again for no one but you), but once you get to the light, it turns red. I always have to go like 65mph (limit is 45) in order to make the light. I swear it is taunting me.
OK, ranting aside... I would gladly volunteer to do simulations and work out scenarios for different amounts of traffic and etc etc, but I am at a loss for how I can offer this help. What would be awesome is if the city were like, "OK, here's how the intersections are laid out, here's where the sensors are, blah blah blah. Who can come up with the best timings?" and have people submit what they've figured out.. cuz whatever they're doing now can probably be improved. Do they gather real-time data? Do they measure how much waiting each person does? What are all the factors they take into consideration?
One thing: I'm not sure San Diego has surface road sensors. I think it only has freeway sensors. I could be wrong and my info could be outdated, but when I worked for an industry leader in traffic data aggregation, San Diego was not a fully sensored market.
posted by batmonkey at 4:56 PM on October 28, 2012 [1 favorite]
posted by batmonkey at 4:56 PM on October 28, 2012 [1 favorite]
I don't do traffic engineering, but half my office does nothing but that. Signal timing design is done on specialty software (that I know next to nothing about) but I do know that it is a complicated thing, especially since changes can have a large ripple effect. Options for how to time signals are limited by what type the signal is - is it actuated (uses sensors to detect traffic), does it have the ability to have different timing for different times of day, are the signals along the corridor interconnected? Also, the signals are probably designed for the peak hour traffic (probably either AM or PM rush hour) so if you are traveling at a low volume hour the lights are not timed for you and timing is optimized for the heaviest volume movements, normally these will be the thru lanes, so turn lanes may have more waiting time.
I agree with grouse to email and not to rant. Also, please don't suggest that the general public would be better at doing an engineers job than the engineers who have spent years doing this work.
posted by Sabby at 6:03 PM on October 28, 2012 [3 favorites]
I agree with grouse to email and not to rant. Also, please don't suggest that the general public would be better at doing an engineers job than the engineers who have spent years doing this work.
posted by Sabby at 6:03 PM on October 28, 2012 [3 favorites]
Best answer: The effect of traffic signals can be counterintuitive.
For example it is intuitive to reason that "If there are zero cars that need me to stop, but the light was red for me anyway, then that was a suboptimal signal". But ramp metering is a simple counterexample: there, a red light stops you from entering the freeway, and there is no cross-street at all, let alone any cross traffic. It just works out better overall for everyone if you stop you for a while.
In the case of ramp metering, it's because of the "fundamental diagram" that relates the density of cars to how fast they flow down the road. In a nutshell, the more cars (vehicles per mile) you put on the road, the more cars (vehicles per hour) drive down it ... up to a point ("critical density"). Beyond that point, the more cars (vehicles per mile) you put on the road, the fewer cars (vehicles per hour) flow down it (i.e., congestion happens).
So, the effect of traffic signals can be counterintuitive, but you could still be right. Different regions use different methods (of different quality) to work these things out... You could ask what methods were used to work out the timing for certain intersections, and that might get you in touch with an engineer quickly too.
Related: Gridlock Buster is a little flash game from the University of Minnesota about traffic signal timing.
posted by hAndrew at 6:14 PM on October 28, 2012 [5 favorites]
For example it is intuitive to reason that "If there are zero cars that need me to stop, but the light was red for me anyway, then that was a suboptimal signal". But ramp metering is a simple counterexample: there, a red light stops you from entering the freeway, and there is no cross-street at all, let alone any cross traffic. It just works out better overall for everyone if you stop you for a while.
In the case of ramp metering, it's because of the "fundamental diagram" that relates the density of cars to how fast they flow down the road. In a nutshell, the more cars (vehicles per mile) you put on the road, the more cars (vehicles per hour) drive down it ... up to a point ("critical density"). Beyond that point, the more cars (vehicles per mile) you put on the road, the fewer cars (vehicles per hour) flow down it (i.e., congestion happens).
So, the effect of traffic signals can be counterintuitive, but you could still be right. Different regions use different methods (of different quality) to work these things out... You could ask what methods were used to work out the timing for certain intersections, and that might get you in touch with an engineer quickly too.
Related: Gridlock Buster is a little flash game from the University of Minnesota about traffic signal timing.
posted by hAndrew at 6:14 PM on October 28, 2012 [5 favorites]
Response by poster: It appears my humorous (to me, at least) ranting about traffic lights has escaped a couple here :) Rest assured that any actual communication with traffic-related authorities will be civilized.
Would the general public be better at a traffic engineer's job than a traffic engineer? In general, no. And it shouldn't be.
Also, considering how many people drive the streets and experience the timings first-hand, it seems like there would be relevant real-life data that could be provided by those many people, vs what is determined via computer simulation. Data from Waze, for example, could be useful.
Thanks for the replies so far!
posted by wuMeFi at 6:38 PM on October 28, 2012
Would the general public be better at a traffic engineer's job than a traffic engineer? In general, no. And it shouldn't be.
Also, considering how many people drive the streets and experience the timings first-hand, it seems like there would be relevant real-life data that could be provided by those many people, vs what is determined via computer simulation. Data from Waze, for example, could be useful.
Thanks for the replies so far!
posted by wuMeFi at 6:38 PM on October 28, 2012
Response by poster: hAndrew, that umn link is awesome, thanks!
posted by wuMeFi at 6:42 PM on October 28, 2012
posted by wuMeFi at 6:42 PM on October 28, 2012
Best answer: OK, ranting aside... I would gladly volunteer to do simulations and work out scenarios for different amounts of traffic and etc etc, but I am at a loss for how I can offer this help. What would be awesome is if the city were like, "OK, here's how the intersections are laid out, here's where the sensors are, blah blah blah. Who can come up with the best timings?" and have people submit what they've figured out.. cuz whatever they're doing now can probably be improved. Do they gather real-time data? Do they measure how much waiting each person does? What are all the factors they take into consideration?
I think we've established the best course of action is a polite e-mail to the appropriate agency. Mentioning the Nobel / La Jolla Village Drive (LJVD from here on out) timing problem is concrete and analyzable, for instance.
I took a quick Streetview look at the area (god, how great the Internet is), and there's not a lot of sensors around there. The most common sensor in an established area is an inductive detection loop that senses when there's a big metallic thing parked on top of it. Newer signals (past 5 years or so) can use cameras to detect cars, but anything from 1950-2005 is a loop. They are visible as the round cuts in the pavement here in the left turns from Genesee onto Nobel. (Dark round circles where the wire has been buried into the pavement; they've painted white dots in the middle.) I didn't notice them anywhere else in the couple of adjacent intersections on Genessee.
So they may not have a lot of data out there, and the lights don't know that there's no one who wants to turn left. Further, the area you're talking about is a complex mix of uses; I see shopping areas in particular, which tend to be tougher to time properly. (Some timing plans are based on the rush hour demand, which will not necessarily serve evenings and weekends with a lot of shoppers.) It's also possible that they're planning on putting more sensors out there, but the pavement is in pretty rough shape so they've decided to hold off a couple of years and do the work when the road is repaved or something like that.
The government I used to work for would do timing plans for a corridor on an as-needed basis; if there were a bunch of complaints, or if it was a long time since the corridor was done. The entire corridor pretty well needs to be done at the same time, and that's a lot of work, especially since my guess is that they don't have a lot of data. But asking politely is the first step.
posted by Homeboy Trouble at 7:35 PM on October 28, 2012 [4 favorites]
I think we've established the best course of action is a polite e-mail to the appropriate agency. Mentioning the Nobel / La Jolla Village Drive (LJVD from here on out) timing problem is concrete and analyzable, for instance.
I took a quick Streetview look at the area (god, how great the Internet is), and there's not a lot of sensors around there. The most common sensor in an established area is an inductive detection loop that senses when there's a big metallic thing parked on top of it. Newer signals (past 5 years or so) can use cameras to detect cars, but anything from 1950-2005 is a loop. They are visible as the round cuts in the pavement here in the left turns from Genesee onto Nobel. (Dark round circles where the wire has been buried into the pavement; they've painted white dots in the middle.) I didn't notice them anywhere else in the couple of adjacent intersections on Genessee.
So they may not have a lot of data out there, and the lights don't know that there's no one who wants to turn left. Further, the area you're talking about is a complex mix of uses; I see shopping areas in particular, which tend to be tougher to time properly. (Some timing plans are based on the rush hour demand, which will not necessarily serve evenings and weekends with a lot of shoppers.) It's also possible that they're planning on putting more sensors out there, but the pavement is in pretty rough shape so they've decided to hold off a couple of years and do the work when the road is repaved or something like that.
The government I used to work for would do timing plans for a corridor on an as-needed basis; if there were a bunch of complaints, or if it was a long time since the corridor was done. The entire corridor pretty well needs to be done at the same time, and that's a lot of work, especially since my guess is that they don't have a lot of data. But asking politely is the first step.
posted by Homeboy Trouble at 7:35 PM on October 28, 2012 [4 favorites]
Best answer: So, a friend of ours questioned the timing of our signal lights after his son got a ticket from one of the red-light cameras. Being an engineer, he went out and timed the lights, and found they did not meet state standards. He got them changed, and he is trying to get the city to refund tickets given before the city changed the timing on the lights.
There's an email at the linked article you can use to reach him, to ask him how he went about it. It might help you with your own traffic signal issues. Good luck!
posted by misha at 8:45 PM on October 28, 2012 [2 favorites]
There's an email at the linked article you can use to reach him, to ask him how he went about it. It might help you with your own traffic signal issues. Good luck!
posted by misha at 8:45 PM on October 28, 2012 [2 favorites]
Aaaahahaha. I used to live on Genesee. HELL.
Sorry, nothing to add, but I can corroborate your claim that spending any nontrivial time driving on Genesee is utterly crazymaking.
posted by town of cats at 10:58 PM on October 28, 2012 [1 favorite]
Sorry, nothing to add, but I can corroborate your claim that spending any nontrivial time driving on Genesee is utterly crazymaking.
posted by town of cats at 10:58 PM on October 28, 2012 [1 favorite]
I don't know San Diego, but:
My first suggestion would be consistantly driving the route at exactly the speed limit --- I know you mentioned speeding up is the only way to make one particular light, but have you tried staying at the precise limit the whole way?
And while I agree the stop-and-go stuff is crazy-making, are you traveling WITH or ACROSS the main traffic flow? I mean, if you're going east/west, but the major traffic goes north/south, that could be the problem: Traffic Control may have the lights timed to make sure the heavier traffic doesn't back up, on the theory that it's better if five cars get delayed than fifty.
posted by easily confused at 2:38 AM on October 29, 2012
My first suggestion would be consistantly driving the route at exactly the speed limit --- I know you mentioned speeding up is the only way to make one particular light, but have you tried staying at the precise limit the whole way?
And while I agree the stop-and-go stuff is crazy-making, are you traveling WITH or ACROSS the main traffic flow? I mean, if you're going east/west, but the major traffic goes north/south, that could be the problem: Traffic Control may have the lights timed to make sure the heavier traffic doesn't back up, on the theory that it's better if five cars get delayed than fifty.
posted by easily confused at 2:38 AM on October 29, 2012
Best answer: From personal experience in the big city, changes like you request need allies. All you will get from writing to the Traffic Department is "This is why we do it."
So before you write anything else to the traffic engineers, write to your city council member voicing your concerns. Your council member will then (hopefully) write to the Traffic Department requested official comment. If a change is needed, your City Council member will bird dog the issue as part of their constituent services.
I assure you that the Traffic Department is more apt to looking into, commenting on, and addressing your concerns if they are routed through the San Diego City Council.
Be aware, it may take years for change.
posted by lstanley at 7:41 AM on October 29, 2012
So before you write anything else to the traffic engineers, write to your city council member voicing your concerns. Your council member will then (hopefully) write to the Traffic Department requested official comment. If a change is needed, your City Council member will bird dog the issue as part of their constituent services.
I assure you that the Traffic Department is more apt to looking into, commenting on, and addressing your concerns if they are routed through the San Diego City Council.
Be aware, it may take years for change.
posted by lstanley at 7:41 AM on October 29, 2012
@ easily confused, the main traffic route is (or at least should be) Genesee, up to the intersection the OP points out, but! That also seems to depend on the time of day so... grar. The probelm is likely that in high traffic times you can't get from Nobel to LVD before the light turns anyway so it isn't likely to matter. It is a very busy intersection during rush hours. (one of those wait for the green light more than once sorts of busy)
Agreed that this is a very annoying road. There is this great spot further down on Genesee (at the 52) where the route I take to work is clearly less travelled and the green light in my direction goes once while the left turn from the opposite direction goes twice (in the cycle). Completely wigged me out the first time I watched that happen, but from the traffic pattern it is understandable.
posted by Feantari at 11:47 AM on October 29, 2012 [1 favorite]
Agreed that this is a very annoying road. There is this great spot further down on Genesee (at the 52) where the route I take to work is clearly less travelled and the green light in my direction goes once while the left turn from the opposite direction goes twice (in the cycle). Completely wigged me out the first time I watched that happen, but from the traffic pattern it is understandable.
posted by Feantari at 11:47 AM on October 29, 2012 [1 favorite]
« Older Does a director on the level of Paul Thomas... | what do you call 2 girls in prom dresses and... Newer »
This thread is closed to new comments.
If you want something to change, make no phone calls and most of all engage in NO RANTING. That's a quick way to get ignored. Find the appropriate department (probably the Street Division, not the Traffic Division) and email them a succinct, factual description of the issue (not like what you have posted here), your suggested solution, and request a response.
posted by grouse at 4:31 PM on October 28, 2012 [6 favorites]