Since November, Paul Reynolds has captured the speed of nearly 30,000 vehicles that have traveled past his Locust Avenue home to see how many have exceeded the 25 mph limit.
“You’ll find that almost 85 percent of the cars going by are violators,” Reynolds said.
Reynolds, who spent many years on the faculty of the computer science department at the University of Virginia, installed a camera on his roof and began writing speed-monitoring software after a 12-year-old pedestrian was injured by a car last October.
“We can’t continue to tolerate this speeding every day,” he said.
Reynolds’ camera captures between six and eight hours of traffic each day. Recordings are then processed by software that Reynolds wrote.
“I first did tracking for the National Oceanographic and Atmospheric Administration 40 years ago to track the movement of the eyes of hurricanes to predict their course better,” Reynolds said. “I know what I’m doing and this is not a toy.”
Reynolds codes in a language called OpenCV that he said provides powerful routines for video processing. The software translates each vehicle into a shape and then measures its movement between two fixed points.
“How many frames did it take for them to move through this field, how many frames should it have taken them? A car going 25 mph, I see it 40 times between the goalposts,” Reynolds said.
“If they go through in half the frames, they’re going 50 miles per hour,” he said.
The software also calculates the general area of each vehicle, which tells Reynolds if it is something larger than a car, such as a truck or a bus.
At 4:27 p.m. last Wednesday, the system caught a Route 11 city bus doing 34 mph.
“At that moment, our bus was traveling just over 30 mph,” Jones said. “From the intersection of Calhoun to the intersection of East High, the bus averaged 26.9 mph.”
“My drivers need to slow down,” Jones said.
“They’re not the worst offenders but these guys are heavy and they are the people who should be setting the pace for us,” Reynolds said.
Charlottesville police stepped up enforcement last year, according to data provided by Lt. Stephen J. Upman. Officers issued 91 tickets on Locust Avenue in 2015, up from nine in 2010.
After the child was hit last fall, the city’s Department of Neighborhood Development Services installed radar-equipped signs that tell vehicles how fast they’re going. Reynolds collected his own data over the same period and said the data initially showed lower speeds but not for long.
“There was approximately 3 percent growth per day in infractions of the speed laws,” Reynolds said. “By the end of the third week, people were totally ignoring the signs that said slow down and use extreme caution.”
Reynolds wants city councilors to see the technology in action. So far, only Bob Fenwick has taken him up on the offer.
“I was very impressed,” Fenwick said. “The technology is there. The application is clean and neat. The effort is admirable.”
Reynolds and neighbor Bruce Odell will make a presentation on the software to the council Monday night.
“We in the Martha Jefferson community are very hopeful that the city shortly will be able to install sensible traffic-calming improvements along Locust Avenue, and the neighborhood association is working closely with them on an array of possible measures,” Odell said.
Locust is classified as a collector road, which means the city prefers not to install stop signs or speed humps. The city is working on alternatives that could include painting new parking spaces, installing planters or narrowing lanes.
Last Tuesday, Reynolds calculated that 37.6 percent of the vehicles he captured on camera were not speeding.
“Sixty-three percent of the cars were speeding, and that was on a day in which there was still snow five feet out into the street,” Reynolds said. The high amount of speeders, Reynolds said, makes him believe that narrowing the lanes, one proposed solution, would not be effective by itself.