Slashdot Mirror
Denver Must Prove Red-Light Cameras Improve Safety
An anonymous reader writes "An audit of accidents at Denver intersections where red light cameras were installed versus increasing the length of the yellow light shows little difference in the results. In a case of putting the public ahead of the corporation, the Denver auditor is recommending canceling the red light camera program unless the city can prove a public-safety benefit." I hope that private citizens offering analysis or recommendations are treated fairly.
...reading some years back that the Red Light camera companies had specific language in the contracts that restricted the length of yellow lights.
A cynical person might think they wanted people running red lights. But I'm not...oh, fuck it. I am cynical.
When Fascism comes to America, it will call itself Anti-Fascism, and tell you to give up your guns.
I know here in Canada and in all the places I've been in the US yellows are plenty long.
The issue is assholes entering the intersection to turn left when it isn't clear, people refusing to stop when the light does turn yellow, etc.
I'd actually want to see a very clear causal link between longer yellows and safety increases, because my gut tells me longer yellows would make people ignore them even more.
I used to think the intersection camera were a good idea. However, I changed my mind once a I listened to a local police chief explain that in his city traffic accidents had actually risen at the intersections where the cameras were in use. Folks would brake suddenly when they saw the camera causing the vehicle behind them to rear-end them. Once he said that I knew he was right. People would do that.
The cameras are a good idea in theory, but the real-world unintended consequences are too costly.
Long yellows to give everyone a chance to stop, and red light cameras to catch the bastards who don't take that chance.
Give me Classic Slashdot or give me death!
specifies that the duration of the yellow change interval should be between 3 and 6 seconds. And people have won court cases over red light tickets over the yellow time being too short.
http://www.ite.org/decade/pubs/IR-117-E.pdf
http://www.ite.org/safety/issuebriefs/Traffic%20Signals%20Issue%20Brief.pdf
http://www.ite.org/annualmeeting/compendium10/pdf/AB10H2601.pdf
Some science has already been done on this subject, and it suggests red light cameras actually increase the rate of accidents. If i remember correctly it was even covered previously on slashdot.
http://www.sciencedaily.com/releases/2008/03/080311151159.htm
Guess the person(s) / corporation who sold this idea to the decisionmakers were not so keen at looking at what had already been established.
Also, I posted the full link as I don't know how to "linkify" a word, and could not find a guide anywhere. I'm a med student and not a programmer. Please, have mercy.
Start putting timers on the yellow and green lights. I've been saying this ever since I starting to see cities put timers for crosswalks. Timers on traffic lights will help people know when that sucker is going to turn red. I run yellows all the time because some seem to last forever, while others flash for a brief second then its red. If I'm coming up on a light, with only 2 seconds left on a yellow, I'm more likely to slow down and stop for the red.
How about just slow things down a bit and increase the illusion of danger instead of the illusion of safety?*
* http://www.wired.com/wired/archive/12.12/traffic.html
As a EU citizen I understand americans hate regulations. But would this not be a thing that should be covered by law? I mean ... what the fuck? In your country a city can decide how long the traffic light is yellwo, that sounds pretty retarded to me.
In germany the duration of yellow depends on the speed limit of the affected road.
Cost free eBook I read (by iBook/Kobo/Amazon/ObookO/Gutenberg etc.): "The Green Odyssey" by Philip Jose Farmer.
My reading of the stats in the TFA is that the rate front to side impacts have decreased 5 times for read light cameras compared with a rate decrease of less than 2 for yellow light extension. Being T-boned at an intersection by a red light runner is far more dangerous than being rear ended by someone not stopping soon enough because they didn't see the light change. So I'd hardly call the change in accident rates a "little difference". Sure injury reduction has been about the same and front to rear is slightly better for the yellow light extension, but I'd hardly call that conclusive.
It astounds me that in the US red light cameras are so reviled. I am continually scared when facing a green light at an intersection and then having some one drive through the red light from my left to right. These people are trying to kill me. So supporting a system that lets them get away with it is nonsensical.
I am Slashdot. Are you Slashdot as well?
According to the law*, a yellow light is to be treated as a red light *if* the vehicle can safely stop. Only if you can't safely stop at a yellow are you to proceed.
Naturally, if folks are driving the posted speed limit, it's far easier to stop at a yellow, because stopping distance increases quite a bit when your speed goes from 30 mph to 35 to 40 to 45. We can bicker about speed limits on the interstate all day long, but local road speed limits are much more important to get right, because you've got pedestrians, cyclists, autos pulling in and out of driveways, right on red at intersections, etc. Stopping distance is really important. Do a better job enforcing local speed limits, and you'll find that folks are less likely to drive through a yellow (or "orange") light, improving safety for everyone.
The other part is this. Plenty of folks treat a yellow as green. Always. Lengthen the yellow, and folks get a feel for the longer length... and will continue to just plough through it as if it were green. Once folks re-calibrate, you've got a worse situation, because people will see a yellow and be even more inclined to accelerate.
There's no need to lengthen the yellow. We need to enforce local speed limit laws.
* all vary state to state, but this is generally speaking the case
Support a few technologists in Washington.
From the article it appears that the number of injuries at the intersection have actually declined since the introduction of the red-light camera. Front-to-side collisions are down and these are caused by the driver running the red light. These collisions are more dangerous than the front-to-rear collision since the vehicle directly enters the passenger area at a potentially higher speed.
Rear-to-front collisions are caused by the driver tailgating and these in general are due to him not being able to stop in time and the collision are at a much lower speed and do not directly enter the passenger compartment. The data provided in the article reenforces this hypothesis since there were 53 injuries prior to the cameras installation and only 18 afterwards. This is despite the gain of 1 front-to-rear accident.
These comments are my own and do not necessarily reflect the views or opinions of my employer or colleagues...
How about increasing the delay between red in your direction to green for the cross traffic, so if someone does run the red there will be a couple extra seconds before cross traffic starts to flow.
While we're at it let's remove what I call "Stupid stoplights", that do nothing but waste gas. How many times have you sat at a red light with NO cross traffic for 30 seconds or more.
If I didn't have absolutely NOTHING to do, I wouldn't be here.
Inevitably these are the words that will issue from some Superior Officer's mouth each morning so they can "prove" that red light camera improve safety even around the areas they're installed where there are no cameras.
And what follows is destroyed and distorted paperwork, reclassification of incidents, motorists NOT being issued tickets on certain roads, people being "let go" and individuals involved in accidents being encouraged to "work it out between yourselves so it doesn't go on your record".
We KNOW what happens when police are under pressure to produce downward statistics in crime each year, or in this case downward statistics for accidents. Policing becomes less professional and more third-worldy, even criminal.
Some examples: http://www.youtube.com/watch?v=m3mmuZsHmv8
http://www.huffingtonpost.com/2011/10/13/ex-nypd-cop-we-planted-ev_n_1009754.html/
It's not what the cops want to do, it's what well-intentioned people who think policing should be subject to the same kinds of productivity and performance metrics that other industries are subject to inadvertently cause.
Telling cops they need to produce such and such numbers for this and that reason is a stupid idea who time has never existed in the first place. Telling them they need to prove by stats that the camera improve intersection safety is a big mistake.
The way to work this is to let them do what from their experience they feel will work and have the insurance companies by law turn over their statistics to the government or the universities who then data mines it on an ongoing basis to see what works for traffic safety and what doesn't and what's trending and what isn't.
Don't make the source of the data also the beneficiary of the data when it leans a certain way. Also don't punish them when it leans some other way.
The police don't cause crime so it's not theirs to reduce year over year. Society causes crime, the economy causes crime, bad parenting and poor family environment causes crime, lousy neighborhoods cause crime. Not policing.
The vast majority of police forces do what they can in the best way they've learned how and results are really pretty good in most areas. But the lions share of the credit or blame goes to the population who either is or is not inclined to follow the law in the first place.
Squeezing departments to produce numbers is a sure fire way to have them enact a quota system which is a sure fire path to corruption which is a sure fire path to contempt for cop on the part of the citizenry which is a sure fire way to increase crime as the years go by.
We need to do everything we can to produce and maintain a justice system that honorable and equitable and run like hell from anything that tends to corrupt that system.
In some cases, though, this is not pre-set.
Particularly here in the UK, a lot of traffic lights have (possibly capacitive) sensors buried in the road which detect the traffic passing over them. Lights will cycle early if no traffic is detected passing through a green, and/or if traffic is detected waiting at a red, especially at night when there is less traffic. That makes it hard to know in advance when they're going to change.
Another 2c from me: yellow times should depend on the speed limit.
All the new fancy crosswalks with countdown timers for the pedestrians do exactly that. It allows me and others (i assume) to see how long before the light turns yellow. Allowing me to make better judgments.
We had ice-packed roads in Denver last night again. It is not possible to break in three seconds without skidding, especially in a vehicle without fancy electronic brakes. You either have to drive rather slowly- 25 mph or less. Or go through the red light. I do some of both.
What helps a lot is 80% of the light have pedestrian countdowns, which at zero go to yellow. (some states go to red at zero) I can decide to start braking if the countdown is in single digits.
One of the Denver TV stations (FOX) collected these statistics. The city council has commissioned a study. the increased rear ends are from more sudden-braking.
either or both?
I listen to both RIAA and non-RIAA stuff if I like the music, tangential business/politics nonwithstanding.
The city of Redmond, WA did a one year pilot study installing red light cameras at a few key intersections. The full study of how well they reduced traffic accidents is worth a read, but in a nutshell there was essentially no impact to the number of traffic collisions. 89% of the citations issued were for turning right on red without coming to a complete stop. The only place the cameras were useful was in the school zone.
Based on the above study the city decided to cancel the contract for the cameras.
Neil
I seem to remember something in driver's ed about Speed Limits being advisory -- they were primarily supposed to inform you of the maximum safe speed for the road which was also the maximum legal speed. But, it wasn't a right, as there was a "basic speed law" that said that the punishable speed limit could actually be lower, depending on driving conditions.
My problem with speed enforcement is that it's not generally automated -- the police setup speed "traps" where people are known to exceed the speed limit. But in my experience, these traps are really just that -- traps -- places where the road conditions are such that the posted speed limit is too slow for driving conditions (visibility, road conditions, traffic levels, limited access disruptions) which subtly encourage drivers to speed AND the officers have a secluded place (blind spot) from which to "catch" speeders.
Ironically, the places where speeding is most dangerous are the places where its most difficult to have speed traps because of traffic, road conditions (small shoulders, limited visibility, etc).
And I've never heard of the police using accident statistics to justify their placement of speed traps. I'm also told by those in law enforcement that speed enforcement in many metro areas has nothing to do with road safety but is considered a crime deterrent (criminals apparently avoid areas with police presence) and field intelligence tool as it allows officers to "interview" motorists and possibly find other, more substantial violations or criminal behavior; a thinly veiled checkpoint.