This is a follow-up article about Washington Metropolitan Area Transit Authority’s trip data collection and analysis program, Trace. Read the first article here.
People can either use buses as a complement to a region’s rail network or a substitute. As a complement, a passenger uses buses to connect to rail, but as a substitute, they use buses exclusively even when rail would be more convenient for their trip.
In Washington, DC, people might use buses as a substitute because of the fare difference. One Metrobus ride costs a flat fare of $2 (including a free transfer to another bus within one hour of tapping onto the first bus), while one Metrorail ride during peak could cost as much as $6.
WMATA has a data collection software that can determine if people use buses to complement or substitute rail, according to Catherine Vanderwaart from the agency’s Office of Applied Planning. But will they use this data to make the case for free rail-to-bus transfers?
Not yet. “This data is used right now to back up plans already in place,” Vanderwaart said. Fare changes are a Board of Directors-level decision.
For most transit agencies, where people get off the bus is a very important mystery. It affects everything from where bus bays are located to how schedules and routes are made.
But unfortunately, there’s no easy way to determine people’s travel patterns. “In the past, we would assign people to trains and bus stops to people-count, but neither of these efforts turned into anything big.”
That’s where data comes in. Vanderwaart is a graduate of MIT’s Transit Lab, a department of the university that uses “big data” to estimate travel patterns. The lab analyzes data for the Chicago Transit Authority, Hong Kong’s Mass Transit Railway, and Transport for London.
Here’s how it works: researchers pair data from fare cards showing where passengers enter the transit system with bus and train location data. By combining these data sets, researchers approximate which anonymized passengers were on which vehicle, and from there – this is the part that is far too complicated to concisely explain here – guesstimate where riders exit the system.
“The real benefit is collecting overall trends,” Vanderwaart said.
According to Vanderwaart, this program – called Trace – can infer the destination of 70 percent of bus trips. This knowledge can massively inform bus planning, in which schedules and routes are re-evaluated every three to six months – unlike rail, where schedules rarely change.
“If a bus terminates at a rail transfer station, we wouldn’t know which color lines people were transferring to and which directions,” Vanderwaart said. “Now we can locate the bus bays better for how people are transferring, and see the effects of limited-stop bus service,” which could improve bus speeds.
WMATA will give this data to the research contractors working on the Washington Area Bus Transformation Project.
The entire transit agency is excited about Trace, though. “It’s been really fun to see how excited people are in the agency when we show them this new level of data, all of the lightbulbs that go off, all of these questions that people now have access to,” Vanderwaart said.
Photo by Sam Kittner for Mobility Lab.