Showing results for 
Search instead for 
Do you mean 

Big data offers big insight into big traffic

by Siemens Dreamer Siemens Dreamer on ‎03-20-2017 12:43 PM

By: Robert Behre, columnist for the Post & Courier

 

COLUMBIA — South Carolina’s highway planners have more insight than ever into how congested our roads are, thanks to the phones, navigation devices and other personal electronics that we carry with us in our cars each day.

 

And this data confirms what many Lowcountry commuters know too well: The Charleston metro area has many of the state’s most clogged roads.

 

In fact, almost half of the 20 most congested interstate and non-interstate segments in the state are found in Berkeley, Charleston and Dorchester counties. U.S. Highway 17 between Interstate 26 and downtown Charleston and the Mark Clark Expressway at Interstate 26 sit atop the list.

 

And the data — mined from the same sources that are used to create the real-time television and smartphone maps with red-, yellow- and green-colored roads — also is providing insight into the success of road projects as well as other solutions to ensure that traffic continues to flow as well as possible.

 

And that will become an increasing challenge as South Carolina’s economy continues to grow, said Dipak Patel, who directs the state Department of Transportation’s system performance management.

 

“Can we beat congestion? No, we can’t beat congestion,” he said. “There will never be enough funding to do that.”

 

Measuring congestion

Transportation planners didn’t get serious about measuring congestion until the 1960s, said Tony Voigt, a research engineer with the Texas Transportation Institute in Houston.

 

That’s when California’s traffic planners began installing inductive loops into many of that state’s roads — metal strips that sense a change in the magnetic field when a car or truck passes over them. These loops — commonly found today in the roadway at many traffic lights — provide a traffic count and can even measure speed if they’re spread about 40 feet apart.

 

They also work like pneumatic tubes, which use air instead of electricity, but those tubes are used mostly for temporary measurements because the tubes’ rubber will decay over time — especially when pounded by thousands of car and truck tires.

 

Voigt said the technology of measuring traffic has evolved over the last 40 years to incorporate radar and video analytics and, most recently, communications technology.

 

During recent years, motorists buying GPS units like TomTom, trucks with their own locators and smartphones have increasingly painted a real-time picture of traffic flow on every road during every minute of the day.

 

South Carolina started using so-called “big data” in 2008, Patel said, and the quality of it has improved in recent years — just as the state was emerging from the Great Recession and traffic volumes across the state began growing again.

 

Voigt said the big-data measurements have really found their way in the last several years. “It’s relatively new,” he said.

 

Patel said the state began using this traffic-flow data in 2008. More recently, his office meshed it with more traditional traffic-count information as well as its data on the features of each stretch of road, such as the number of lanes, lane width and other data to come up with a new way to measure which sections of which roads were the most congested. This approach allows for apples-to-apples comparisons between a stretch of highway in Greenville and one in North Charleston.

 

“This is real data, not assumed data,” Patel said. “You can’t get more real than this.”

 

Freeimages.com/ John PilgeFreeimages.com/ John Pilge

 

Following up

The data also is used to evaluate fixes.

 

For example, the state spent about $66 million several years ago to improve Interstate 26 interchanges at Remount Road and Aviation Avenue in North Charleston.

 

The data showed how dramatic an effect it had — easing congestion by about 47 percent.

 

Michael Dennis, who does most of the state’s traffic projections for big projects, said he tries to take a conservative approach to estimating how much congestion relief a project will offer. “Some days, the crystal ball is clear,” he said.

 

But big data is showing how some projects that might have eased congestion for as much as a decade in the past now offer far shorter relief. In just three years, the data showed congestion at Remount and Aviation had returned to the same level before the interchange was rebuilt.

 

“It’s not the project’s fault,” Patel said. “It’s people changing routes.” When Interstate 26 improved after the Remount and Aviation project, commuters and others switched back to it instead of using Rivers Avenue or Dorchester Road.

 

New solutions

Congestion is only one of nine factors considered when the state Transportation Department evaluates which road projects to pursue. Act 114 says the state also should consider financial viability, public safety, economic development, pavement quality, environmental impact and local zoning.

 

And managing congestion is about more than building projects: It increasingly will involve managing travel demand through such steps as carpooling and staggering work schedules; promoting public transit, biking and moving more freight by rail; and modifying highway signs, intersections and traffic lights.

 

Dennis said if the state can get 10 different strategies to reduce traffic by 1 percent each, “then you’ve got a 10 percent reduction.”

 

Locally, the consulting firm Stantec used probe cellphone data for its recent study of 8 miles of Interstate 526, said Vonie Gilreath, mobility manager with the Berkeley-Charleston-Dorchester Council of Governments. She said that study also highlighted the need for low-cost congestion relief, such as carpooling, telecommuting and new striping and signs.

 

In previous decades, the answer to traffic congestion mostly involved finding ways to add more lanes, but that solution is often not practical.

 

“We’re getting to a point now where cost of construction has gone up and right of way is almost impossible,” Patel said, adding that Ashley Phosphate Road is a prime example. “We couldn’t widen that without buying those hotels. We’re running out of room.”

 

Patel said the state is starting to use big data not only to identify congested areas but also to develop the strategies to address them — and to monitor how well those strategies are working.

 

“There’s not enough funding to build out of congestion because the economic growth grows much faster,” he said. “There are no more silver bullet solutions.”

 

How it works

A growing number of motorists have GPS devices for navigation or other reasons, and these devices are switched on as they drive down the road.

 

Increasingly, these devices are being monitored — and their information collected, then massaged for insight into how traffic flows.

 

The state of South Carolina has bought data from INRIX, a private company founded a decade ago that collaborates with hundreds of companies to collect traffic flow data every second of every day.

 

The INRIX data compiled devices in about 275 million vehicles, smartphones, cameras and other sensors over almost 5 million miles of roads, ramps and interchanges in more than 40 countries, according to its website.

 

Before going to the S.C. Department of Transportation, the data is first sent to — and cleaned up by — the Center for Advanced Transportation Technology at the University of Maryland, said Dipak Patel, director of DOT’s System of Performance Management.

 

“We can’t maintain the servers to keep this data in South Carolina,” he said. “It’s just too much.”

That said, the state isn’t monitoring who is driving where. Patel said the state only gets the average of travel times from many devices traveling between two points on a highway.

 

“We don’t have the ability or access to the individual device information,” he said. “No, Big Brother is not looking over you.”

 

John Voigt, a research engineer with the Texas Transportation Institute, said while Bluetooth signals are tracked, their unique numbers are changed before they are sent along, so no one can identify the owner.

 

“When it’s sent back to a location to wait for a match down the road, you can’t re-create it, and even if you could re-create it, Bluetooth in particular is not tracked in the sales chain,” he said.

 

But Voigt said privacy issues likely will continue to arise as a growing number of cars talk to the infrastructure they are driven on.

 

“With the future of automobiles and ever-increasing connectivity, the technology that we use today will be completely transformed within 10 or 15 years from now,” he said. “We think we have big data, but that will really be big data.”

 

 

Visit the Post & Courier for the original article. 

Top Kudoed Posts