London’s Tube tracked passengers’ phones via WiFi. What if Metro did the same?

London’s subway network recently tracked people's phones to collect data that it says will help improve train service and operations. Could WMATA borrow this idea? Doing so could help the agency get a better handle on Metro station crowding or the routes people take, but it’d also raise some serious questions about passenger privacy.

Transport for London (TfL for short, but also known as the Tube) conducted a phone tracking pilot at the end of 2016 for four weeks from November 21st through December 19th. During the trial, the agency tracked the physical addresses (MAC address, which is unique to each phone) of people's’ phones at 54 of the system’s downtown stations.

When the person exited the Tube, TfL had a good idea which stations the person traveled through and which trains they took (with the Tube, unlike Metro, there are a number of ways to go from various point As to point Bs). For instance, TfL was able to break down percentages of people who traveled between the Victoria and Liverpool Street stations, a trip with at least four viable routes. 70% of people preferred the two most direct routes, but a small percentage of people chose to either get out and walk in the middle or make a third transfer.

Being able to model how passengers actually use the system is hugely important, and can change what information the agency gives to people, or might tell the agency what it needs to tweak in order to get people to change their route to a different one (like going from a quick but overcrowded trip to a longer but uncrowded one).

Another use for the data: by having multiple WiFi access points throughout the stations (you couldn’t cover an entire train station with just one), TfL was able to essentially create 3D maps of where people waited in their stations — on the mezzanines, escalators, and platforms. Knowing this could allow TfL to give advertisers super-precise information on how many customers see their ads throughout the system. There are other potential uses for this kind of data, informing passengers which stations to avoid because of crowding.

Metro could use tracking technology for similar operational benefits

Of the ideas TfL looked at using the tracking data for, in-station tracking is maybe the most interesting and most applicable to WMATA.

An obvious first use of this kind of data at Metro could be to explore how people spread out on train platforms. Some trains are six cars long and others eight, and the WiFi tracking data could be matched up with train arrival/departure times to get a good overview of whether people crowd part of the platforms and not others. This might let Metro more easily identify which stations need better signage or crowd control.

This kind of data could also indicate which cars on a train are more popular (thus also showing which would be less crowded). In this case, knowing that could let Metro figure out where it might want to send an extra employee or two with a megaphone, or potentially place more eye-catching ads telling people where they might want to board a train.

Any mobile device tracking would have to be done carefully to assuage privacy concerns

With WiFi signal tracking comes some obvious privacy concerns. TfL mitigated these concerns by including posters in stations where this tracking occurred telling people what was going on. Though passengers were essentially opted in unless they chose otherwise, at least they were made aware of the need to make that choice. This is similar to how, when navigating in a car for instance, you choose whether you want to use a GPS navigation app that tracks where you are.

But what if WMATA didn’t provide that opt out option? An effort to monetize the data collected from tracking phones is not something everybody would be comfortable with WMATA doing. Some people might have real security concerns — or just concerns about the information being misused.

In terms of uses for the data, WMATA could tell vendors which advertisements are surrounded by the most people. Or similar to how some banner ads on websites can “follow” you around online, prying into the data set could potentially lead to even more money in ads, as they could follow people throughout the rail system. Maybe it would be worth Uber’s money to have ads radiate out from downtown in the evenings as people leave work. That way, once you’ve reached your destination station, you’ve seen their ad seven different times and have now decided to open the app instead of taking the bus.

The benefits of having a massive dataset of individual travelers’ paths and habits could be an incredibly useful tool for a transit agency like Metro, as it would give an understanding of how theoretical station and train designs mesh up with the reality of how people actually go about their travels. But a proper rollout would have to be managed such that the uses of the data are well-known and carefully controlled.