Last week, coders presented their apps and data projects using bike movement – from both shared and personal bikes – at Bike Hack Night IX for Mobility Lab’s Transportation Techies. What did all of these unusual projects have in common? They use open and crowdsourced data to understand the options cyclists have and how they could be better.
(And this is exactly why open bikeshare data is important.)
Functional art
Matt Fowle opened with his “Live Light,” an LED bulb connected to a heart rate monitor. The light flashes or changes intensity in tandem with the rise and fall of users’ heart rates. As they work harder, the light shines brighter and blinks faster. What makes it so fun, according to Fowle, is that he can “transform characteristics of the heart into something can see.”
Daniel Schep talking about his projects.
In addition to urban cycling, mountain biking got some attention, too. Daniel Schep’s project to make trail exploration easier combines the stylization that property owners add to their paper trail maps with the deeper knowledge that digital platforms allow – a combination that hadn’t existed before. He now has a digital platform for exploring mountain bike trails that offers interactive elements to help users better understand trail characteristics, such as points of interest and elevation changes.
Schep also followed up his February Bikeshare Hack Night presentation with Bike Hero, an app that shows users where to find publicly accessible bike pumps and Fix-It stands around Washington, D.C. Though the app uses data from Fix-It stands’ manufacturer as a “seed” data set, Schep believes the true power of the app lies in crowdsourcing resources.
Ultimately, he hopes it will encourage community members to leave their own pumps or tools (tethered to their homes) in publicly accessible areas to add resources and a stronger sense of community among cyclists around the region.
Bike to the future
While Techies past have analyzed historical data of bike counters, George Silva has turned this information around to forecast ridership at the Bikeometer in Arlington’s Rosslyn neighborhood. Silva has applied automated machine learning – where an algorithm teaches itself how to interpret data as its “knowledge” grows – to BikeArlington’s data from the sensor and combining it with factors such as days of the week, holidays, weather, and even daylight hours and Daylight Savings Time shifts. Despite having difficulties with consistent datasets, including power cuts to the Bikeometer and subpar weather feeds, Silva has built a pretty accurate predictor of how many people will bike past the counter a few days ahead of time.
Jacob Baskin presenting about his company, Coord.
Finally, Jacob Baskin of Coord shared how his company is working to improve multimodal trip planning. Importantly, they have created a tool that pulls in data feeds from dockless bikeshare providers, giving users the ability to plan a trip through D.C. using the best possible combination of bike and transit options most immediately available to them. For example, Baskin generated a trip that suggested he ride an Ofo bike to a Metro stop, from which he would complete his trip via transit. And the tool uses live data, showing what is nearby as users generate their trip, and updating if anybody takes a bike away before the trip starts.
Data collection for dockless bikes, as opposed to the station-based Capital Bikeshare, is a heavy lift. Opening dockless bikeshare data gives coders like Techies the opportunity to build their own apps, maps, and analyses to serve their community. By integrating data sources and feeds into one place for others’ use, perhaps future Techies will be able to dig even further into bikeshare data and how people use their local transportation network to move around.
All photos by M.V. Jantzen.