Cutting information overload on the MTA’s platform information screens

The Digital Services team at the MTA was overhauling their platform information screens down to the bare bones of the data feeds, and asked for my product and design consultation on the rider-facing screens.


This project consisted of overhauling three different types of screens found throughout the New York City Subway system:

The core goal was to get arrival information to more riders, using the same screens we already have access to.

A rectangular screen hanging from the ceiling of a Subway station showing train arrivals. A photo of orange brick wall of the 49th St subway station. There are two vertical screens hanging on the wall, one showing upcoming arrivals and one showing alerts. The street-level entrance to a Subway station. A large screen mostly displays an ad, but the bottom few inches show train arrival times.
Countdown clock at Union Square, Customer Information Screen at 49 St, and Subway Entrance Screens at Union Square

Re-setting expectations

Evidence suggested that riders felt that each screen type had a different data source, each with varying accuracy. Riders usually trusted the countdown clocks over other screens and felt they were the most accurate.

Thanks to some excellent data work by the engineering team that dug all the way down to analyzing the frequency of pings from a train’s onboard beacons as it enters the station, all screens in the system now show the same arrival times and update within seconds of each other.

We wanted to reset rider expectations and communicate this new shared data visually, using the same visual language across all of these signs.


I started my design explorations with the countdown clocks. Finding a visual language that works on these, the most trusted screens, and then applying it to all the others, would hopefully increase usage and trust of the others as well.

The updated designs address usability issues with the existing screens based on speaking with riders on the platform, while respecting technical constraints of the incoming data and agency requirements for how trains are represented on signage:

On the existing design, the second row alternated between the next five trains, regardless of how many lines stopped at the station.

This meant that if you missed seeing the second train, you might have to wait through the third, fourth, and fifth trains to see it again. This could be the difference between deciding to get on the arriving train or wait for the next one:

Pretend you’re looking for when the next N train arrives - how long do you have to wait?

We updated the logic for how many trains show on each screen to stop displaying trains after the last unique route is shown, meaning that the second and third trains to arrive will be on-screen much more often.

We also added a spatial rotating animation to better indicate that the second row is rotating, replacing the subtle instant change:

That N train is showing up much more often now.

Countdown clocks everywhere

Now that we had established a structure and visual language for a train arrival row on the countdown clocks, we wanted to extend that visual language across other screens.

Starting with the Customer Information Screens, we turned them into countdown clocks themselves, using the top half of the screens for arrival rows at eye-level, and the bottom half for detailed alerts.

A vertical screen showing a list of upcoming train arrivals, and service alerts
A vertical screen showing a route map of the next train to arrive.

When two screens are paired together, the right one gives a detailed route map of the next train to arrive.

Again, we use quick animations to spatially convey what’s happening when trains arrive or reorder.


Making good use of constrained space

The last screen real estate we had was the Subway entrance screens. While the initial instinct was to use the same visuals as all other screen, I encouraged us to consider the different context for these screens.

Because the other screens are on the platform, riders’ questions are constrained to confirming the time until their train arrives and whether an arriving train is for them. By contrast, the Subway entrance screens are used by people with many more options available to them.

People looking at a Subway entrance might be choosing between a cab ride and the train. They might spend some more time above ground if their train isn’t for a while, or even walk to another station.

Most importantly, they have an infinite number of other things competing for their attention as they navigate walking on a busy sidewalk.

To address this, we settled on a much simpler and bolder design that still mirrors the visuals of the other screens while removing most of the detailed information, favoring making the Subway routes and estimated time as large as possible.

For station entrances that have many more options in multiple directions, we fall back to a layout that’s more similar to the countdown clocks.


Reception

The screens have only been rolled out to a portion of the subway system so far, but riders have already taken note with overwhelmingly positive feedback.


The screens were designed and developed in collaboration with Will Fisher, Torrey Chiang, David Brulhardt, Sivakumar Ganesan, Teresa Ibarra, Ken Prakasam, Eric Shu, Grisha Temchenko, and Sunny Zheng on the MTA Digital Services team.


I previously explored countdown clock designs for the MTA - read the case study.

While those ideas still feel well-solved for the countdown clocks alone, the constraints of this project led to different conclusions which are interesting to compare.

Specifically, the need for these designs to be extendable across multiple contexts and noticeable as a significant visual change to reset riders’ assumptions about the screens ruled out some of the denser information design and layered visual approach of my earlier exploration.