Transit data reveals patterns about our community, but no other system in the Washington D.C. region (or perhaps in the nation) makes it as easy for people to access the raw data and crunch the numbers on their own as Capital Bikeshare.
Ever since CaBi started releasing trip-history data, professional analysts and hobbyists alike have been pouring through the data to sift out nuggets of information about our lives.
Mobility Lab sponsors CaBi Hack Night, part of the Transportation Techies Meetup group, in order to encourage people to share their creations and discoveries.
Our second-ever CaBi Hack Night was August 7 here at Mobility Lab. Seven speakers gave “show and tell” demonstrations of their work. ShopHouse Kitchen, Capital Bikeshare, and BikeArlington provided food and swag, making for a fun evening.
Michael Alvino from the National Park Service started us off with a discussion of the history of CaBi on the Mall. The stations on and near the Mall are in some ways the reverse of the system as a whole: on the Mall, most users are casual members (with one- or three-day memberships). In fact, 20 percent of all casual trips either arrive or depart from the six stations on the National Mall. Since casual users pay higher prices, the Mall stations help make the entire system more profitable.
Inspired by Ben Wellington’s iQuantNY blog, Charley Dingboom dug through all 6,333,624 trips available in CaBi’s open data, covering September 2010 to March 2014. Bike number W00905 spent the most time checked out by riders: 58 days, 8 hours, 34 minutes, and 4 seconds. But another bike, W00632, had the most overall trips: 4,863. The average CaBi trip is 17 minutes, 29 seconds.
During the Cherry Blossom Festival, CaBi painted a single bike pink and dubbed it the “Bike In Bloom.” When civic hacker Justin Grimes found it, he decided to track its travels throughout the festival. His BikeInBloom data analysis included an animation tying its whereabouts to tweets about the bicycle.
Yinyue Hu from George Mason University used statistical methods to show how certain stations form clusters, and how the clusters vary between casual riders and registered riders, as well as between summer and winter and weekdays and weekends. Her analysis also highlighted that the busiest time for CaBi is Saturdays in June.
Dwight Martino talked about cabistations.com, CaBi’s official “suggest a station” crowdsourcing map. Local planners have used this data to make decisions about where to invest in new CaBi stations. And now that set of data is available to the public, at cabistations.com/opendata, so you can sift through the data to discover your own findings.
Andrew Zalewski of Foursquare Integrated Transportation Planning shared with us the Arlington County Capital Bikeshare Transit Development Plan. We saw how the planners used data, including survey data, to make decisions about how to grow the bikeshare system.
I had my own demos, two little JavaScript apps that visualize CaBi’s trip-history data in different ways. The Birds-eye view of CaBi stations app takes a mix of stats and plops them on aerial photos. The 2013 daily ridership statistics for Capital Bikeshare lets you compare how different stats correlate to each other. I used the tool to see how weather affects CaBi ridership (see Explore the Links Between Weather and Capital Bikeshare Ridership).
We’re proud that our local bikeshare agency makes it easy for everyone to access its data. CaBi Hack Nights bring people together to share their knowledge and meet other data nerds. Mobility Lab is happy to help build this community. It was just last December when the Transportation Techies Meetup group debuted with the first CaBi Hack Night (see Capital Bikeshare Hackers Pedal Their Wares at Mobility Lab). Hope to see you at the next event!
See lots more photos of the event here by M.V. Jantzen.