May 17, 2007
Amended Carona bill passes out of the House
Back in April, a bill passed out of the Senate that would explicitly give cities the right to operate red light cameras, while simultaneously putting restrictions on their use. Now an amended version of that bill has passed out of the House, which would insert a sunset provision for the cameras.
The sunset provision was added to a Senate bill that would give cities formal legislative authority to enforce red lights with cameras. The differences in the House and Senate versions will have to be worked out in a conference committee.
Rep. Carl Isett, R-Lubbock, a longtime opponent of red-light cameras, amended the bill to prohibit the use of cameras after Sept. 1, 2009, unless they get an affirmative vote from the Legislature.
"Today is the first opportunity we've had to talk about how they will be implemented," Isett said. "We've always had constitutional problems with these. We're frustrated having this conversation over and over again. We're frustrated with the way this became law."
Senate Bill 1119 gives cities the legislative authority to operate red-light cameras. It prohibits contracts between cities and vendors that base compensation on the number of citations issued. And it requires cities to study intersections' traffic volume, collision history and frequency of red-light violations before installing cameras.
Rep. Jim Murphy, R-Houston, the House sponsor of the bill, attempted to table Isett's amendment, saying it reduces local control, would create an unfunded mandate and is a potential breach of contract. Murphy's motion failed on a close vote.
Murphy accepted 11 other amendments to the bill. Among them, cities would have to monitor and report the number and type of wrecks at each intersection to determine whether collisions declined after cameras were installed.
As always, the next step is a conference committee, at which time anything can happen. In theory at least, I'm okay with Isett's amendment. I've said all along that I want to see evidence that the cameras actually have a palliative effect on collisions and injuries at the intersections where the cameras are installed. I just want to know what criteria will be used by the Lege to judge their effectiveness. Will every camera-enabled intersection have to show a reduction in collisions, or will a reduction at the intersections as a whole suffice? A reduction in total collisions, or in the per capita collision rate? Just collisions, or collisions with injuries? All I'm saying is that the metric should be defined and agreed upon before the new rules are implemented, or else we'll just be making the same "does not! does too!" arguments all over again.
UPDATE: The Walker Report gives a progress report on the cameras in Balcones Heights.
Posted by Charles Kuffner on May 17, 2007 to Planes, Trains, and Automobiles
The stats from Balcones Heights are not indicative of a red light running problem in my opinion.
According to the Walker Report
At the Balcones Heights Council meeting Monday, May 14th, Police Chief Bill Stannard presented figures that indicated in a two month period, over 1,800 citiations were issued to violators.
They have 4 intersections with cameras. Now - lets assume that the intersections are only monitored on one approach as is typical. This means four cameras total.
So in two months (60 days for the sake of simplicity) they have issued 1,800 citations. This is 30 citations a day - or about 7.5 per intersection per day.
Now the typical suburban or urban non-CBD signal cycle length will be between 90-130 seconds or so. So lets assume a cycle length of 120 seconds (2 minutes). That would be 720 cycles a day if a signal ran the two minute cycle all day long.
Recognizing that signals run shorter cycles during off peak hours, thus creating more opportunity for red light exposure, and that they often run fully actuated in the deep night hours, lets look only at 12 hours of the day - 7A to 7P - so we have 30 cycles an hour for 12 hours or 360 cycles.
Thus our 7.5 citations per intersection per day are equivalent to one red light violation per 48 cycles.
I have no knowledge of the traffic volumes on these roadways so I can't take this the step further and calculate approximate RLR violations per 10,000 vehicle-cycles which is the recommended metric for violations by TTI. However, we can just make an assumption for argument's sake.
Lets assume that the subject approach is a very low volume signalized roadway in the range of only 250 vehicles per hour. 250 veh/hr * 12 hours = 3000 (Of note, many major urban roadways experience volumes in excess of 1,200-2,000 vehicles per hour in one direction during the peak one hour). Now 3,000 veh * 360 cycles = 1,080,000 vehicle-cycles. 1,080,000 divided by 10,000 = 108. 7.5 violations / 108 = 0.069 violations per 10,000 vehicle-cycles.
TTI recommends using the threshold of about 1.0 violation per 10,000 vehicle-cycles to determine if an intersection has a potential RLR problem and is in need of engineering countermeasures.
As can be seen, even with a very low vehicular volume assumption, the 7.5 violations does not equate to a problem. When you increase the vehicular volume that 0.069 violations per vehicle/cycle would decrease even further and get further away from the threshold value of 1.0.
I'd love for any city to publish their violation data in the units recommended by TTI.
I have seen first hand the way Cities are implementing these things in my daily duties as a consulting traffic engineer. They are first and foremost about money. A secondary benefit is the safety benefits. Period.