Bicyclists in the Washington, DC area encounter obstacles along most of their rides throughout the region. With blocked bike lanes, aggressive drivers, and potholed streets, fighting for a safe space to ride can sometimes feel helpless.
But some bicyclists who code are using their skills to nudge their cities toward policies that better protect cyclists, while also reminding each other to have fun and delight in the joys of biking (the joy aspect is an important component of ditching the car).
Coders at Transportation Techies’ tenth Bike Hack Night showed how they learn from each other and collaborate to build upon their work in order to advocate for better spaces for bikes.
Crowdsourcing data on bad driving
Daniel Schep led this off with an update from December’s Playing with Traffic Meetup, sharing that he is developing his How’s My Driving? Twitter bot into an app that will collaborate with public agencies to enforce parking laws. Schep has already spoken with the District’s Department of For-Hire Vehicles to work with cabs and ride-hailing (Uber/Lyft) drivers who block bike lanes. Eventually, he wants to connect the app to the Department of Public Works to ensure regular offenders get caught through the power of crowdsourcing.
In the meantime, Chris Nguyen is helping frustrated cyclists vent their anger with his web app “Bike Lanes WTF,” a version of what many cyclists likely are forced to shout multiple times a day. This app geocodes the locations of photos taken of cars blocking bike lanes and uploads them to Twitter. This effectively makes a giant, public database of the drivers getting in the way, likely with impunity, and is probably ripe for harvesting for future Techie/advocacy tools.
Other works shared by Techies have been evolving, as well. Chris Slatt’s Parking Dirty tool, which identifies how often bike lanes along certain road segments in Arlington are blocked (about 60 percent of the time), inspired Brendan Freehart to test an automated version of processing camera feeds of bike lanes through machine learning, with which a computer can learn from a large set of data to then make inferences and interpretations of similar datasets.
Because Slatt’s original version uses humans to tag whether or not a bike lane is blocked, Freehart had a reliable training set to teach a computer how to identify images of obstructed bike lanes. Freehart doesn’t consider this ready yet, but if computer vision can reliably identify blocked lanes, it could process far more information than advocates or academics could on their own, and therefore be a powerful tool for pushing for better infrastructure policies.
When cities fail to protect bike infrastructure, some studies have to focus on what Joseph Nelson did: bicycle crash rates in the District. Using the Metropolitan Police Department’s database of crash reports involving bikes – a notoriously underreported statistic, according to other attendees familiar with the issue – Nelson hoped to identify which parts of DC are more dangerous for bike riders. He also discovered that Fridays are the most dangerous day of the week, on average, to bike in DC.
Though Nelson hoped to use this to identify areas that cyclists should avoid for their safety, it raised the question that instead, perhaps it should highlight how unsafe streets are for anybody outside of a car, and if people are riding in dangerous spots when they have safe alternatives suggests significant need for a change in infrastructure.
Important presentations from Chris Leonberg and Daniel Schep grounded the conversation in why many people ride bikes: because it’s fun. As he introduced Leonberg, Michael Schade made a point that is worth holding onto in many conversations about human movement: “something we as coders sometimes lack is creative inspirations” for why we do what we do.
Leonberg brought in some of that creative inspiration by introducing the crowd to the Baltimore Kinetic Sculpture race, “part art pageant, part parade, part DIY festival, with some racing thrown in.” He explained the process of building the base of a kinetic sculpture – a moving piece of art – and one of the final products that he and his wife have raced around Baltimore’s Inner Harbor.
Schep also returned to remind the crowd that every day should be Bike to Work Day. In his effort, he built a Twitter bot that tweets “Happy Bike to Work Day” every day, serving up a small sliver of inspiration for its followers each morning.
Photo by Michael Schade