Beyond the Turnstile: How Subway Riders Use Transit
Knowing how many people enter and exit a subway system is a relatively straightforward affair. By their very nature, such systems generally have limited access points, so counting each body that passes through can be accomplished with a high degree of accuracy. Technologies vary from the tried and true mechanical turnstile to fancier infrared systems like those in the new Amazon stores. Here in New York City, the MTA uses mechanical turnstiles and publishes data on MetroCard swipes by the day, the week, and the year, that can be used by policy makers and designers alike. However, since the MTA does not require riders to swipe upon exiting the system (unlike San Francisco’s BART), it is impossible to tell where riders go. More importantly, it is impossible to determine how a rider transfers within the system to reach their destination.
To illustrate this blind spot, consider Broadway Junction, which is a large subway station in eastern Brooklyn where three subway lines meet. The MetroCard swipe data shows that entries are relatively low at just over 3 million in 2016, but informal observation describe a station filled to capacity, which is likely to increase with the looming L train shutdown. To better understand the dynamics at play, the NYC Department of City Planning used 11 paid staffers placed throughout the concourses to manually count traffic flows. What they found was that just 16.87 percent of all observed passengers had entered from the street. The vast majority of traffic was being generated by riders transferring between lines, which was effectively invisible in the MetroCard swipe data. (See Broadway Junction Transportation Study, 2008, page 42 for more information)
Like the MTA, Transportation for London (TfL) manages a sprawling and interconnected subway system that millions of people every day. TfL also faces the challenge of not knowing exactly how riders traverse the system; where they transfer; which routes are preferred; what routes could be added. Replicating a staffed study along the lines of the one described above would be astronomically expensive if conducted throughout the entire system, but by using the public Wi-Fi network, there was no reason to.
TfL recently conducted a pilot study in 54 stations to test the viability of using Wi-Fi tracking, which uses depersonalized pings from people’s devices — essentially your phone checking for the nearest Wi-Fi point just in case you should decide to use it — to paint a picture of passenger flows. The TfL have been releasing the data for external analysis and the results have been very impressive so far, offering a level of insight into how the network is used that has never been seen. This graphic really sums it up. It shows that riders who traveled between King’s Cross St Pancras and Waterloo took at least 18 different routes! This type of data would be nearly impossible to discern from ticketing data or paper-based surveys.
The implications for this type of insight is relatively straightforward in a subway system. Ways to improve passenger’s journey can be informed by providing (potentially real-time) understanding of traffic peaks, estimates on the amount of time transfers will take, the degree to which any disruptions will impact travel time, and visibility into how busy various line and route options are. Taking this technology outside the subway system could inform other municipal entities. For example, parks could “see” where “desire lines” are before people create paths through open fields. Business improvement district could decide where to dispatch food trucks to an area that people are congregating at around lunch time. City governments could use the data to understand the utilization of new amenities. Former New York City Mayor Michael Bloomberg was famous for saying, “If you can’t measure it, you can’t manage it and you can’t fix it.” Using Wi-Fi tracking would finally allow municipalities to really measure how people used the city around them. Unfortunately, this type of thinking ultimately leads to fairly dystopian scenarios of Big Brother tracking us step by step. Everyone may need to get tin foil hats!
Setting the dystopian future aside, and recognizing the potential for privacy concerns, TfL plastered its stations with posters informing riders of the study and advised anyone who expressed apprehension to just shut their phone off. In the end, everyone can control how they are tracked if they are made aware of it.
