Road Repair via Crowdsourcing

Road Repair via Crowdsourcing

A contest could help make Boston's pothole-spotting smart-phone app more accurate.

Bumpy ride: This screenshot from the Street Bump app for Android shows circles that represent suspected potholes.
Credit: New Urban Mechanics

In February, the city of Boston developed an app called Street Bump, which collects data from drivers with Android smart phones. Now, after identifying some problems, the city plans to have volunteers refine the app's accuracy.
Traditionally, Boston has relied on public works inspectors to locate potholes. Since 2009, residents have been able to report potholes, graffiti, and other nuisances using the Citizens Connect app for iPhone and Android devices. Users snap a photo and submit the image, which is routed directly to city repair teams. Some other cities already have similar citizen-reporting programs. For example, Minneapolis provides a website where citizens can file pothole complaints.
Street Bump, released by the city in February, requires no direct user involvement. The app, for Android devices only at the moment, uses GPS data to track a device, and detects potholes by using the phone's built-in accelerometer to sense sudden jolts. When multiple phones report the same jolt, the app identifies a pothole that needs to be repaired.

Road tested: An area of Boston that was tested for potholes. Credit: New Urban Mechanics

Nigel Jacob, cochair of the mayor's office of New Urban Mechanics, an organization set up to explore new approaches to civic engagement—and which created both Citizens Connect and Street Bump—says his team was unhappy with the number of false positives that Street Bump produced. "We need someone to do deeper analysis of our data," he says. "The app works well, but it can't tell the difference between a real pothole and a train track."
To improve the app, the city has now posted a bounty of $25,000 on Innocentive.com, a marketplace for crowdsourcing innovation, for a developer to create algorithms that report potholes accurately.
Thilo Koslowski, a Gartner analyst who studies automotive technology, says the success of the app will depend greatly on the accuracy of the data.
"These types of applications are the first examples of self-aware devices and applications that will become an integral part of future smart infrastructures and smart cities," he says.
Jacob says he could see Boston collecting data from drivers in real-time, leading to faster response times and better roads. The city plans to roll out the app to other platforms as well, including the iPhone, in the future.