Capturing Boston's Subway System in Stunning Detail

What a month's worth of train data can reveal about delays and commute times.

Image
Michael Barry/Brian Card

Two students at Worcester Polytechnic Institute have raised the bar high for train data visualizations.

Courtesy of the Massachusetts Bay Transit Authority and its Developer Relations program, Michael Barry and Brian Card got their hands on information like train schedules, real-time train locations in the Boston metro area (only available for the Red, Blue, and Orange lines), and per-minute entry and exit counts at each station based on turnstile measurements.

The pair pulled all the data from the month of February 2014. And then, taking cues from information visualization pioneers like Edward Tufte, they crafted the data into a stunning set of interactive visualizations. They hope their work will help people better understand how Boston’s transit system works and how their behavior fits in to overall patterns.

For example, the graphics in the screenshot above depict all subway trips on the Red, Blue, and Orange lines on February 3, 2014.  In the "timetable" on the right, each vertical line represents a station, so the colored lines represent each train's position as time passes by. Steeper lines means slower trains. In the interactive version, mousing over a single train also brings up its position on the subway map (shown below in the center).  

On the very left of this screenshot is a zoomed-out look at the whole day. In the concentration of colored lines, we can make out the more frenetic morning and afternoon rush hours, the mid-day and evening lulls, and complete lack of service in the wee hours.

Below, we zoom in at around  5:00 p.m. on the "timetable," where you can see that a disabled train (bolded) on the Red Line delays all the trains following it for over an hour (which is observed in the concentrated strip of extra steep lines). Major delays due to disabled trains make headlines quite regularly in Boston—just this week a disabled train at Davis Square created mayhem on the Red Line. Now, we can see the extent of the impact of such disturbances. 

The project also tackles the people component. The graphic below for example, illustrates the relationship between congestion and passenger volume. You can see how overall, the biggest delays, highlighted in red, correspond to morning and evening rush hours, when turnstile entries recorded across all the stations are highest.

Another visualization also lets you explore specific trips on each line. Just highlight the start and end positions of a trip and see how long that commute would take at different points of the day. You can begin to see what time in the morning might give you the best shot at a fast trip.

For example, as shown in the screenshot below, if you get to the Harvard Station platform at 7:56 a.m., the typical total commute time to South Station—including wait and transit time—is 18 minutes. Hover over the graphic and you can see that if you get there at 8:25 a.m., that number goes up to 21 minutes.

Barry and Card also observed trends like the fact that increased delays during rush hour are generally offset by trains arriving more frequently, and that transit times on the Blue Line are less variable than on the Red Line. To explore the detailed data in all the interactives, click through to the project page.

Inspired to make something similar?  Barry has promised to share the code behind these visualizations soon on GitHub.

About the Author