Other programmers at CaBi Hack Night sought to better understand riding connections between stations and data accessibility issues
Bike rebalancing is one of the most costly and inefficient issues that bikeshare systems face.
The constant need for rebalancing manifests most frustratingly in “dockblocking,” those times when riders reach a station to find all of its docks filled and must keep riding on to the next one. It’s telling that many past Transportation Techies presentations have focused on how best to avoid being dockblocked, and that bikeshare operators are now focusing their efforts on addressing this and other demand problems.
Motivate, which operates Capital Bikeshare, is looking to riders to help address the problem in a pilot with New York City’s Citi Bike, which it also operates. The program, called Bike Angels, encourages participating customers to rebalance bikes as they ride. Those who ride a bike from a full station to one lacking bikes receive points that can be redeemed for raffle tickets and membership extensions. Motivate considers the program a success, with “angels” now providing more than 10 percent of the system’s rebalancing on busy days.
Alex Tedeschi, a GIS developer at bikeshare company Social Bicycles, dug into trip data for the Citi Bike fleet and mapped it to visualize how widespread rebalancing trips – rides taken against the commuting flow that return bikes to emptier stations – are for the system. While he found that rebalancing trips dropped 5 percent from 2013 to 2015, it is still a prevalent behavior.
According to Tedeschi, there are consistent patterns for availability and ridership for every station in the system. His breakdown of bike availability shows three distinct groups of station behavior, largely depending on their location in New York. Orange stations in the above map, for example, are places where riders will leave bikes in the morning, but not take them in the afternoon. While he calculated riders stand a 3.4 percent chance of being dockblocked overall, in neighborhoods like the Lower East Side, the likelihood of striking out is closer to 8 percent.
Social Bicycles, whose systems are “dockless” bikeshare, claims to have found an alternative to rebalancing vans by offering financial incentives to return bikes to areas that are considered “hubs. Colin K. Hughes, Social Bicycle’s director of strategic development, explained how the computer “brain” on a bikeshare bike’s fender collects an array of GPS data and allows customers to park and find SoBi’s bikes anywhere within its service area.
But to ensure the bikes spread themselves evenly and that people without smartphones can find them, Social Bicycles systems have established hubs, small geographic areas rather than limited-capacity stations. Some systems incentivize users to return bikes to reliable areas and charge a convenience fee for parking elsewhere. Because of this structure, Social Bicycles found the vast majority of bikes make it back to a hub within three trips, reducing the worry that bikes will become stranded in far-away areas.
Know your network
Riding conditions also play a major role in how customers use bikeshare systems. The availability of comfortable routes contributes to how or if people bike, and Capital Bikeshare data creates a useful starting point to understand a jurisdiction’s overall network.
Tracy Hadden Loh presented an analysis of how Capital Bikeshare stations in Arlington County connect with each other along comfortable routes. While the average station connected to 19 others via low-stress streets, there were 21 with no such comfortable connections at all. These areas remain inaccessible to many potential cyclists, but Loh also showed how relatively small, stress-reducing changes would better connect Capital Bikeshare stations, and therefore the bike network overall.
James Graham of the District Department of Transportation shared the agency’s efforts to make Capital Bikeshare’s live data as accessible to as many people as possible. Since not everybody is a developer, Graham explained, changes to the data feed can be confusing, especially during disruptive events like January’s inauguration, when several bikeshare stations were closed. Now, by combining the system information with GIS-compatible code, Capital Bikshare’s data is more useful to more people and agencies.
Michael Schade added the latest Capital Bikeshare ridership data to his visualization tool that examines system-wide and neighborhood-specific bikeshare usage. The new 2016 data includes the first few months of bikeshare activity in Fairfax County. The map’s “heat map” function displays high ridership areas and, not surprisingly, shows high activity at downtown D.C. Metro stations, but the view can be toggled by jurisdiction to closely examine other areas. Schade also added a boundary tool that helps focus on specific areas by capturing stations in a neighborhood or a transit corridor like 16th Street.
And at last, someone has answered a question that only the most ambitious bikeshare riders have considered: how long would it take to bike to every Capital Bikeshare station in the system? Jonathan Street determined the most efficient route and found that, in order to reach all 441 locations, the shortest route is 264.6 miles. Supposedly the ride should take 32 hours and 10 minutes. Just imagine the time overage charge on that ride.
Photo: Colin Hughes of Social Bicycles presenting at CaBi Hack Night (M.V. Jantzen, Flickr, Creative Commons).