2017 Hubway Challenge

As engineers and scientists with a keen interest in civic responsibility and public policy, we set out to find the unusual and the extraordinary in the last six years of Hubway data. With the curiosity of five-year olds, we asked "Where?" and “Why?” and "How?" at every turn of our exploration. To aid our investigation, we built several tools to query the full dataset and visualize results in realtime, and we've made these tools available for others to conduct their own explorations. Please join us on our adventure through five million Hubway trips as we share what we've asked, what we've learned, and what we've built.

Searching for outliers

To find interesting policy and operation questions, we focused on searching for outliers. For example, plotting total starts and stops for each station confirms an intuitive result: stations near the center of the Hubway network typically have the highest number of trips, and stations near the edge of the Hubway network tend to have the fewest. We wondered if any outlier stations break this pattern: are there stations in the middle of the network with unusually low usage? If so, why?

Do riders underutilize particular stations inside the city?

To answer this question, we measured each station’s “utilization” by summing its total trip starts and ends over a fixed time period and then dividing by the total number of docks and hours.

Then, we searched for stations with unusually small “utilization” rates by searching for small circles surrounded by large circles. Grouping (clustering) stations together by their utilization rates and assigning distinct colors to each group made it easier to find stations with unusual usage. Considering only stations that have been in operation for at least one year, we found the following underutilized stations:

  1. Dana Park, in Cambridgeport;
  2. Binney St./Sixth St and Kendall St. in East Cambridge; and
  3. Brookline at Burlington St., near Fenway.

Underutilization: Dana Park, Cambridgeport

Relative to other Hubway stations in Cambridgeport and Central Square, the utilization rate for Dana Park seems unexpectedly low—especially after taking into consideration the fact that transit by Hubway in this area is often the fastest mode of transportation.

To explore how other modes of transportation compare, we estimated* average commute times based on the Hubway data and combined these estimates with transit information from the MBTA and an average walking speed of three miles an hour to determine the fastest mode of transportation from any arbitrary start location to any arbitrary end location in the greater Boston area. From Cambridgeport, Hubway often is the fastest mode of transportation to Somerville, Cambridge, Boston, and Brookline. Only a few distant destinations, like Chelsea and East Boston, are faster via subway. Given this time advantage, why does the Dana Park station seem to be underutilized relative to nearby stations?

*Our model is pessimistic and always assumes that someone shows up halfway between buses during normal (median) operating schedules. During peak-time travel times or with well-planned trips, transit rides can still be more convenient.

Underutilization: East Cambridge

East Cambridge exhibits similar trends to Cambridgeport: it is relatively underserved by public transportation based on commute times, and Hubway often is the fastest mode of transit. Given this time difference, why are the Kendall area stations near Binney St. underutilized relative to other nearby stations?

Underutilization: Brookline at Burlington St., Boston

In Boston, the Brookline at Burlington St. station near Fenway also appears to be underutilized relative to nearby stations. This particular station also has an additional unusual trait: when riders do use this station, they tend to bike to distant stops. Perhaps the location of this station on a dead-end street makes it inconvenient to access or difficult to discover, and the primary users are a small but loyal group of Hubway commuters?

Underutilization: Winter Availability in Boston

In addition to searching for stations that users seem to visit less than one might expect, given that station's dock size, analyzing utilization rates also led us to wonder why Boston shuts down some of its Hubway stations in the winter when Cambridge does not. Utilization rates for Cambridge and Boston are fairly similar. Although winter utilization is the lowest of any time of the year, riders still use Hubway throughout the winter in Cambridge, and Boston riders might do the same if they had the option. Why not keep the Boston stations open?

Outlier Regions: Why Not Put a Station Here?

While combing through the Hubway/MBTA travel time data, we also wondered if any communities might benefit from the installation of a new Hubway station. For example, because there are relatively few stations in East Somerville, walking currently is the fastest way to travel from Inman Square to large parts of East Somerville. Are there other areas where walking is often the fastest mode of travel? Could these areas benefit from a Hubway station?

Based on local commute times, we identified at least two potential candidates that might merit further consideration for new stations: East Somerville near Winter Hill and Brookline along Kent Street.

Fun Facts

Although we've asked (hopefully interesting and useful) questions about potentially underutilized stations and underserved areas, we've left a lot to explore! Below, we've included some of the fun facts that we stumbled upon during our adventure to find the extraordinary in the Hubway data. Read on to learn what we've found, or conduct your own exploration!

Fun Facts: The Greater Boston Area in Transit Maps

Fun Facts: Where do Hubway riders bike to?

Fun Facts: Work-Life and Gender Balance

Your turn!

All of the maps that you've seen and all of our findings can be recreated with our exploration tools! Start your own adventure with all six years and five million trips of Hubway data.