Using Open Data to Model and Solve Car Traffic Jams

Car Traffic Jams

Presumably you've been stuck in a traffic jam. If you live in California, you might hear traffic congestion, in Texas, it's traffic problems. It's a universal experience that comedians like Seinfeld have been making light of for years.

Simulating Traffic and Alternative Routes

Dustin Carlino, project lead and software engineer of A/B Street, has been interested in improving traffic since working on the DARPA autonomous challenge. When he first moved to Seattle Dustin didn’t have a car, so he biked and used buses for transportation. Based on those experiences, it was clear to him that opportunity existed to model better routes for these modes of transportation.

Using a project he started in college as a starting point, he created A/B Street, a game for traffic simulation. “Traffic jams in Seattle suck,” says Dustin. “The way to fix it is not by adding more highways -  people like to use walkways, buses and bikes.”  His gamified solution creates simulations and proposes fixes to the Seattle City government.

Sharing Traffic Jam Solutions with Local Government

Simulations can then be shared with lawmakers and serve as easily understandable visual aids to advocate for better routes and pathways through the city. The goal is to improve traffic in Seattle then model additional cities, explains Dustin. So, what are some of the challenges? The first and foremost is data. A/B Street pulls from open data--mainly Open Street Map, King County GIS and City of Seattle--but what’s missing is parking data. Some of the available parking data is not correct, and there is no private parking data, leading to skewed, under-counted parking models. Traffic signal timing data has also proven difficult to track down. The other major challenge is that Dustin is a software engineer, not a sales and marketing professional, so getting contacts and meetings with SDOT to convince them to use this data is an area of opportunity.  

Volunteer to Help from Home

At St Hacktrick’s Day, the first-ever DemocracyLab virtual hackathon, Dustin and his team of 2 engineers aimed to make their bicycle simulation models more realistic by including MPH, elevation data (to ‘slow’ bikes uphill), and accident statistics to give “be aware” hints to players. Dustin has found a partner in DemocracyLab--he’s found multiple volunteers on the platform--but in order to keep the project moving forward at a healthy pace he needs more software engineers as the project grows. Additionally, he is looking for help with the marketing and outreach to SDOT. If you are interested in helping, please visit A/B Street on DemocracyLab and sign up! The next virtual hackathon will be on May 9th.