Hubway, a bike sharing system in Boston, was launched in July of 2011. In the past 8 years, they have expanded to over 150 locations throughout the city.

In 2014, as a part of a data science challenge, Hubway made 3 years of its data public. This reflected every time a user started or ended a trip. The main data features included of:

• Start Time and Location
• End Time and Location
• User Type
• User Gender (if user is a subscriber)

From these data points we can extrapolate information on how Boston uses this new transportation service, and how Hubway can more effectively market to its customer base.

One of the questions I hoped to answer in this project was which stations service commuters more heavily than tourists, and at which times. I reduced the data set to only include the year 2012, both to speed the computation and capture an entire year. I also removed all trips that started and ended within a few minutes of each other, as these likely reflect false starts.

I then built an interactive Shiny app that allows users to examine the flow of riders through selected stations. The data is averaged out throughout the year, with a toggle for whether to include weekend data in the aggregate (I found that weekends are almost entire non-commuter traffic). Users will see these features for any given station:

• Graphs show the times that station experiences peak usage.
• The fraction of users that are subscribed vs casual.
• The net flow of traffic of a given time period. Positive indicates more bikes arriving, while negative indicates bikes leaving.
• The top 10 stations users are arriving from / traveling to.

The ideal user of this app would be a station manager or marketer looking to improve the quality of advertisement directed to its user base. I encourage you to interact with the Hubway Station Metrics app yourself and see how the city of Boston is moving with Hubway. You can also check my Github repository to see the code that went into making this app.