Big data enables new tool for analyzing and diagnosing traffic congestion

By Chris McCahill

StreetLight Data, which provides trip-making data from mobile devices and smartphone apps, has just launched a new interactive Congestion Analysis tool. The tool lets subscribers identify congested roads by time of day, break down the traffic in terms of trip length, trip purpose, and other characteristics, and then focus on specific strategies to relieve demand.

SSTI used StreetLight Data’s earlier products in Northern Virginia and more recently in Sacramento to identify opportunities for transportation demand management and last-mile transit connections. In many cases, the data revealed pockets of congestion caused partly by people driving short distances on already-busy streets because they have no reasonable alternatives. This let us recommend relatively small investments—network connections, bicycle and pedestrian facilities, and minor transit enhancements—that could help cut the number of cars just enough to let the roads flow more smoothly. The new tool formalizes this process into one platform.

The alternative, without such detailed information, is often simply to increase the capacity of the road.

The data behind these analyses come from in-vehicle GPS devices and, more recently, data from location-based services used mainly in smartphone apps—an improvement over traditional cell phone data. According to StreetLight Data, this newer data source grew their sample size to cover more than 20 percent of the adult population.

A webinar recorded earlier this month highlights the new Congestion Analysis features. After identifying a congested road, for example, users can break down traffic volumes and congestion by time of day. Then, for any given period, they can see the trip length distributions, trip purpose distributions, and income distributions, which StreetLight Data derives from available demographic and land use data.

 

Congestion Analysis also scores different “congestion busting” strategies based on what’s known about the traffic. A large number of short trips, for example, indicates potential walking and biking opportunities. Circuitous trips indicate network improvement opportunities (e.g., new connections). Trips concentrated in common origin and destination zones indicate transit and carpooling opportunities. There’s also a score indicating congestion caused by commercial trucks. These scores will eventually be customizable.

Congestion Analysis is now available to subscribers in the StreetLight Insight platform.

Chris McCahill is an Associate Researcher at SSTI.