« Thursday Morning News Dump | Main | Failed President Somehow Clings to 43% Support, With 49% Opposing »

Red light cameras - for safety or revenue?

Some random clicking led me to this NPR piece on red light cameras which was interesting.

What's the point of a red-light camera — to make intersections safer or to generate revenue? That's the question prompted by researchers at the University of Tennessee, who say the cameras are sometimes used in ways that are more likely to make money than to improve safety.

Real shocker that something sold to the public for "safety" would be perverted into a revenue maker that might actually compromise safety, right?

OK, you're done rolling your eyes now I suppose...moving on...

In their research, Han and his colleagues explored what they call "dual, conflicting purposes" in the system: The more effective a camera is at making an intersection safe, the less profitable — and self-supporting — it becomes.

Naturally, when revenues decline, action is called for.

Han says some city and county governments have gone through "gyrations" to keep their red-light cameras operating.

"Some increased fines from $50 to $100 per offense," he says. "Some play with traffic signals to get more incidences of red-light running. Some added speed enforcement to the red-light camera program over time."

When people obeying the law and driving safely proves inconvenient, just make sure more people break the law - revenue restored!

For instance, cutting the duration of a yellow light by one second — from four seconds to three, for example — results in a higher frequency of drivers running a red light — as high as 110 percent, according to the study.

And if jacking the Yellow times down isn't producing enough violators, they have some other tricks up their sleeve as well.

And if the speed limit around an intersection goes up, drivers are more likely to be caught in the "dilemma zone" — where they must quickly consider their options, and have less room to stop comfortably. According to one report in the study, raising the limit by 10 mph makes it 45 percent more likely that drivers will run the light.

And if goosing the speed limits up ain't good enough, there's even more "safety enhancement" things they'll do to help make sure you violate.

Another strategy that was reported in the Tennessee paper but not closely studied is the practice of municipalities removing signs warning drivers of an upcoming signal.

So what's a reasonable yellow light duration in practice out on the streets under real life conditions?

Assume a typical recognition/react time of about 1sec. You see the green switch to yellow, decide you won't make it through on yellow, and then start flailing various appendages needed to start the braking process. 1sec for this process start to finish is about right...

...but you still gotta actually stop the vehicle. That 1sec "latency" just got your foot off the gas pedal and onto the brake pedal, you haven't decelerated yet.

There's a table here that will allow you to calculate stopping TIME.

the average driver can safely decelerate an automobile or light truck with reasonably good tires at the rate of about 15 feet per second (fps). That is, a driver can slow down at this rate without anticipated probability that control of the vehicle will be lost in the process.

If we use that "typical" 15 feet per second figure, and assume a speed of say 50mph (a typical speed limit on US 441 in my south FL area), the table says you're moving at 73.3 feet/sec at 50mph.

Using the 15' per second metric, that means you can bring the vehicle to a complete stop in 73.3/15 seconds, which is ~4.9sec once you are ON THE BRAKES. Add to that 4.9sec a standard 1sec latency and you're looking at a recognize/react/stop time of just under 6sec. If your ride is a 5,000lb 1970's gas guzzler hooptie bondo-buggy covered in Wisconsin chrome (i.e dust tape) with bad brakes, bald tires and the road is kinda wet, you may want to extend that 6sec just a tad.

How many yellow lights do you see on 50mph roads, with stop light cameras, with a 6 sec Yellow timing? Yea, not many. Damn few. Vanishingly close to zero. As a practical matter, you got a better chance of spotting Elvis in a 7-11.

Here's what I do -- I look ahead for upcoming stop lights and ease off the gas a bit when approaching them. If you're being tailgated, which is often, this may prompt the tailgater to pass you and zoom ahead into the waiting red light trap. That removes your risk of being rear-ended if the light starts changing.

Having a car or two in front of you at a red light isn't a bad thing either; it means THEY, not you, will likely be the ones to get T-boned when someone cruises a red light coming the other way. I always like it when I can have someone else, test the waters before I jump in.

In many areas, during off-peak hours lights are synchronized so that a vehicle traveling at some particular speed almost never gets one. ex. I can go the 10mi from home to PBI on surface streets at 40mph in the middle of the night passing through dozens of lights and catch maybe only one or two. Go the posted speed limits and I'd be catching 50% or more of them. That's something you can only learn by experience in an area and trial and error.

Here's a link to the research paper NPR mentions. There's a number of interesting graphs there.