How real-time data and analytics are transforming traffic modeling

Traffic modeling has been practiced for decades, but now real-time traffic data and cutting-edge analytics are transforming the digital tools transportation planners, engineers and policy makers have at their disposal.

It all started during the early 1950’s when transportation authorities began to examine road networks on a system-wide scale. Policy makers and engineers understood that future construction decisions had to be in the context of the impact they would have on the entire road network. In other words, if a new road segment was added, how would it impact vehicle movements on adjacent roads?

Possible scenarios, according to HERE product manager Joe Guthridge, could be: What happens if we build a bridge from here to there? What if we close this road or if we add a new lane to that road?

“Traffic modeling helps planners prepare for the what ifs. It brings to light where it’s best to invest,” he says. “For example, we recently worked with a government traffic bureau trying to improve traffic on Saturdays in a certain area. One option was to take away curb lane parking during that day, but that meant spending a significant amount on buying a parking lot. The traffic model allowed the agency to understand if the traffic savings would offset the cost of the lot.”

Traditional traffic modeling

To set up a model, a planner needs a view of the road network and the demand on it. “Where do people want to go and how do they want to get there?” says Joe, “Once you apply the traffic demand to a map of the available road network you get results. Then you can devise proposed changes and identify the impact of various modifications on drivers.”

As we recently explored in a blog post on vehicle speeds around major cities, comprehensive traffic modeling is a complicated exercise that boils down to forecasting a road network’s ability to balance supply (capacity of transportation infrastructure across all modes) and demand (volume of individuals traveling particular routes at given times).

While the process is tried and true, it’s lagging in the context of our connected world. For one, conducting a traffic study costs millions of dollars, so most agencies commission them once every 10 years.

The laborious and costly process involves getting drivers to fill out surveys, so agencies must either put money behind marketing to entice the public to participate; or, authorities have to set up roadblocks where cars are stopped and officials ask drivers where they are coming from and going to. This manual intervention actually creates traffic and become a safety issue.

Smarter traffic modeling

This is where HERE Trip Data comes in and provides a more effective and economical means to measure traffic. HERE collects vast amounts of anonymized probe data (80 billion a month) from vehicles. Probes are minute samples of car journeys — information that only identifies what happened on a trip, not who was in the car. The combined data results in an immediate view of the entire trip. “You can get data today and build a traffic model tomorrow,” says Joe.

Also, as data is amassed, historical information is thereby aggregated so trends over time can be recognized for more sophisticated modeling and future proofing.  Joe explains, “You are able to detect growth hot spots and solve problems before they happen.”

Though traffic modeling is mainly employed by transportation agencies, there’s a waterfall effect where businesses with fleets and everyday drivers benefit. Joe says, “In another example, Trip Data can help retailers learn about the type of traffic that passes by a billboard at a given time, which helps inform ad campaigns.”

Looking ahead to smart cities and autonomous vehicle deployments, Joe believes some of the fundamental principles of traffic modeling will have to be reworked. “Before vehicles are automated, they will be connected, so the range of data will grow exponentially,” he says. “This volume will add to the robustness Trip Data and enable HERE to be an even better partner — helping citizens get where they want to go using big data rather than disrupting their journeys.”