The San Francisco region has joined the bikesharing movement, with the introduction of Bay Area Bikeshare in August 2013. I wanted to see if I could adapt any of my CaBi tools for the “BABS” system, but their open data is too limited to be of much use. They have a System Metrics page which offers only ridership and membership data, which is not very interesting. To analyze the system we need trip history data, like Capital Bikeshare shares every quarter.

Luckily, I discovered Eric Fischer, who has been tracking station statuses since late August. Every minute, he records the number of available bikes and docks at each station. While not as valuable as trip history data, this data does let us discover when stations are either full or empty.

The data he records is a copy of the current station data, available at bayareabikeshare.com/stations/json. I had to reduce the size of the file by writing a Java program to remove redundancy and unnecessary fields. Still, storing data for a single day takes a megabyte of space for even the condensed JSON file.

You can visit my Bay Area Bikeshare Status History app to see the data collected for a single day, Tuesday, December 3, 2013 (chosen for no particular reason). A slider lets you choose the time to show statuses for. Click on a station to show the number of available bikes and docks. There are two coloring schemes you can choose: a gradient changes gradually from red (empty) to blue (full); the other color scheme shows balanced stations in white, but when the number of available bikes or docks goes to one or zero, the color changes to red or blue.

Because the Bay Area Bikeshare system is so spread out, there are five buttons to zoom into local clusters (San Francisco, Redwood City, Palo Alto, Mountain View, and San Jose).

The app isn’t very interesting (I hate to admit) because the system seldom has long-lasting imbalances. Without more data, it’s impossible to say if that’s because the system is naturally balanced via ridership patters, or if the BABS rebalancing team is so awesome, or if no one uses the system. But hopefully the tool allows us a peek into the system, for one day in December. Let me know how it can be improved to help both planners and customers.

Probing Data from Bay Area Bikeshare

Leave a Reply