Smarter traffic lights, calmer commuters

Traffic in Los Angeles, March 25, 2013. Artificial intelligence and other advances in traffic systems hold promise to ease commuters’ headaches. The New York Times
How often have you cursed out a traffic light that took forever to change? Or the lights on a long stretch of road made you stop at every cross street, just when you felt you might actually make it to work on time? Take heart: Less painful commutes may be ahead, with advanced electronic control systems that apply artificial intelligence to the task of keeping traffic moving.

The benefits promised are more than just tamping down some road rage. Less time stuck in traffic, multiplied by many thousands of commuters, can lead to less fuel burned and less carbon emitted.

The first known attempt to control traffic with signal devices, over 150 years ago, did not end well. To manage the city’s growing volume of horse and buggy traffic, the London police used manually operated, gas-powered signal lights. That experiment concluded with a gas lamp explosion, injuring the policeman operating it.

About 40 years later, with Ford’s Model T killing residents at record rates, Detroit installed stop signs and traffic signals and formed a police squad charged with controlling traffic. Soon, Cleveland upped the ante and installed the first electric traffic light.

While traffic lights have made streets safer, they can be an annoyance. Early attempts to manage intersections relied on stored signal-timing plans developed by manually counting traffic and matching schemes to a day and a time. The efficiency of this system is limited by its “after the fact” response and the inflexibility of fixed plans.

More advanced systems respond to traffic conditions in real time as determined by sensors. Introduced in the 1980s and updated as the technology advanced, the Sydney Coordinated Adaptive Traffic System has been widely adopted to manage traffic signals across Australia. Phillip Jordan, managing director at Road Safety International in Melbourne, said, “A city like Melbourne, with almost 5 million population, simply would not function without our area-wide linked-signal system.”

This system, known as SCATS, uses sensors at each intersection — primarily wire loops embedded in the pavement — to measure traffic. It can distinguish some types of vehicles and recognises pedestrians at crosswalks if they push a button. Signal-timing decisions at an intersection are made via a control box. Decisions affecting the entire system are generated by a remote computer. Changes in signal operation come in real time and in small increments so as not to temporarily impede flow.

The technology is in use in several US locales, including Atlanta and several California municipalities.

Oakland County, Michigan, decided to use the system after sending staff to Australia in the late 1980s to evaluate it, said Craig Bryson, senior communications manager for the county road commission. The system, which has been refined since, controls signals at 819 intersections.

To ensure this technology remains a good choice, the county has regularly evaluated other traffic-signal control systems.

“We have a test intersection where we run multiple systems,” Bryson said. “Each has its pluses and minuses. SCATS offers advantages that work for us.”

Siemens Mobility offers a similar system, Split Cycle Offset Optimisation Technique, which has been refined many times since it was introduced in Britain in 1980. SCOOT, like SCATS, measures traffic at intersections and optimises signal timing in real time, using small incremental changes. The system relies on wire loops in the pavement, magnetometers and video for vehicle detection. Connected vehicle communication is being adopted as more vehicles become capable of high-level communication.

Where necessary and where capable detection devices are installed, the Siemens system can distinguish different types of vehicles and detect pedestrians. It uses a central processor to apply algorithms to collected data and determine the best signal operation.

Siemens Mobility, in partnership with Ann Arbor, Michigan, studied the system on a busy road. Travel times decreased 12% on weekdays and 21% on weekends, the study found. The benefits were more pronounced during nonpeak hours, since signal timing has less impact when a road is congested.

A similar study by Siemens in Seattle calculated that 36.4 hours of reduced travel time per vehicle annually saved 32 gallons of fuel and reduced carbon dioxide emissions by 621 pounds.

NoTraffic, an Israeli technology company, has developed a system that is used in Phoenix and several California cities. Billed as the world’s first autonomous traffic management platform, it allows municipalities to enact policies they have developed and control traffic flow both at the grid level and for each intersection where the technology has been installed. It also assists with preemption for emergency vehicles and signal priorities for public transportation and pedestrians.

NoTraffic measures approaching car, pedestrian, bike and public transit traffic using radar and video. The company claims nearly perfect detection of traffic in all weather and precise classification of type, including bicycles and pedestrians. An Optimisation Engine — essentially a computer processor and supporting electronics — at each intersection gathers the data, applies algorithms and calculates hundreds of simulations per second to determine the best signal operation. The system can communicate with connected vehicle systems to warn drivers of potential problems, such as a car that is about to run a red light or an accident blocking the road.

The technology communicates with a central processor that gathers data from many other installations and applies algorithms. Thus, it learns how to deal with a vast number of situations.

In Maricopa County, Arizona, NoTraffic claims to have achieved an average delay improvement of 54%. If the technology were put in place countywide, NoTraffic projects it would cumulatively eliminate 93,106 months of delays, reduce emissions by 531,929 tons and provide over $1.1 billion in economic benefits in a year.

Those are NoTraffic’s own numbers. Tom Cooper, a NoTraffic vice president, said the Maricopa County Department of Transportation had reviewed and approved the data.

Vladimir Livshits, director of transportation technologies for the Maricopa Association of Governments, which includes Phoenix, couldn’t confirm the data. He said a consortium of three local universities was evaluating the system and that he couldn’t provide results at this time. He did say that, overall, NoTraffic was highly promising technology.

While adaptive signal control systems, like SCATS, SCOOT and NoTraffic, are gaining acceptance in the United States, they are far from universally deployed. Part of the reason may be cost. According to the Intelligent Transportation Systems website of the US Department of Transportation, the cost of SCATS and SCOOT technology is upward of $20,000 per intersection, with SCOOT being somewhat more expensive. However, those figures can vary widely if infrastructure changes, such as the development of a fiber-optic communication system, are necessary as well.

Cooper said he couldn’t pinpoint a price for NoTraffic installations but indicated that they could cost up to 70% less than other systems. He added that real-world numbers for all systems varied widely.

Adaptive control systems are expensive, but the fixed-plan systems they are replacing are not cost-free. Cooper noted that the cost of re-timing older systems could be $5,000 per intersection or more.

Traffic control is a much studied subject, and as new technologies evolve and vehicles become smarter, intelligent adaptive control systems are likely to become the norm. That should bring some comfort to those who have been stuck at a red light for what may have seemed like an eternity.

c.2020 The New York Times Company