GPS data informs transportation projects in Northern Virginia: SSTI study

By Chris McCahill
Transportation agencies, dependent for decades on traffic counts and travel demand models, are turning to new sources of data to understand the movement of vehicles and people. These include aerial photography, Bluetooth sensors, and cellular location data. Adding to that list, SSTI recently completed a study of vehicle trip-making patterns in Northern Virginia (NOVA), using commercially available GPS data. That study, presented to the Commonwealth Transportation Board in March, is now available for download.
Our study, done in collaboration with Michael Baker International and Virginia’s Office of Intermodal Planning and Investment, relied on four month’s worth of GPS data covering the entire region, provided by Streetlight Data. The anonymous data come from a variety of sources including navigational devices and mobile phones. The product includes only the highest resolution location data, which is growing quickly now that GPS-enabled devices are becoming more common.
Given the NOVA region’s transportation challenges and its limited ability to add significant new capacity, our goal was to scan for opportunities to improve access to destinations through relatively small investments, better connectivity, and demand management. We analyzed patterns of movement and identified places characterized by one or more of the following:

  • Short vehicle trips that could be made by walking, biking, or local transit;
  • Circuitous trips, indicating poor network connectivity; and
  • Common origin-destination pairs that present opportunities for transit connections or ridesharing.

By working closely with local stakeholders from the counties, transit providers, TDM groups, and other transportation agencies, our team developed 17 case studies highlighting different types of potential investments throughout the region. Many of these investments target short local trips that don’t always appear in travel demand models. Our team found that removing just a small share of those trips—through mode shifts or route changes—could produce major benefits at many of the sites.
One of those sites is George Mason University in Fairfax. The university serves an important function for commuters throughout the region and has invested heavily in parking, which surrounds its central campus. However, our GPS data reveals that roughly one-third of the vehicle trips to and from the campus each day fall within two to three miles (Figure 1), which suggests that much of the demand could be met by walking, biking, and local transit. We also estimate that the campus generates close to 20 percent of the traffic on nearby Route 123, half of which is for trips under five miles. This presents a real opportunity for university planners to think beyond the edges of its campus, improve connections to its surrounding communities, manage parking even more than it already does, and grow as a mixed-use center that can house future staff and students while limiting the amount of traffic it generates.

Figure1
Figure 1. Estimated share of trips to and from George Mason University (in yellow area) during typical weekday

To estimate the total number of trips represented by the GPS traces, we scaled up our relative numbers to match observed traffic counts. In another case study, this revealed several thousand daily car trips traveling between the Van Dorn Street Metro station in Alexandria and neighborhoods to the immediate north, less than a mile away (Figure 2). Most of that traffic uses Van Dorn Street, just west of the station. These data reflect a well known connectivity issue caused by a stream and rail line that separate the two areas. In being able to quantify the number of trips affected, our team estimates that a new connection prioritizing non-auto access could eliminate approximately 100,000 to 150,000 vehicle trips each year and achieve annual personal savings totaling $155,000 for those affected directly.
Figure2
Figure 2. Estimated number of trips to and from Van Dorn Street Metro station (in yellow area) during typical weekday

In addition to the 17 case studies outlined in our report, our team produced several customized data sets for local stakeholders to evaluate specific corridors, projects, and programs. We are also working toward making the data more easily accessible for planners and transportation providers in the region to help them better understand travel patterns and prioritize their investments.
Chris McCahill is a Senior Associate at SSTI.