After a heavy winter, roads in cities become uneven and potholed, and officials need to send out the repair trucks. But where exactly should they go? Which roads need fixing most?
To help answer that question, Argo Labs has the Street Quality Identification Device, or "SQUID," an accelerometer-camera combo that produces a complete map of road surfaces in a matter of days. Mounted on the back of municipal vehicles, the device measures bumps in road surfaces, while taking a picture every second.
"You get a complete data set for a city, so you're able to see the streets in a more equitable manner," says Varun Adibhatla, Argo's co-founder. "With a traditional approach, you're playing whack-a-mole, and it's 'Whoever shouts the loudest gets their street fixed.'"
Argo, described by Adibhatla "as a skunkworks for local government," is currently testing the SQUID in Syracuse, New York. A single vehicle can record up to 400 miles of road in a week. "When you're able to do complete street surveys that quickly, you make preventative maintenance possible and you can start anticipating street defects before they happen," he says.
The camera and accelerometer sit outside a sedan and feed data to a Raspberry Pi computer inside. The equipment costs no more than $300 all in—which is less than more sophisticated laser-based systems, though not quite as granular. What SQUID loses in spot-by-spot detail, Adibhatla says, it makes up for in comprehensiveness.
Boston has a similar project, called Street Bump, that uses measurements from citizens' phones. But Adibhatla says Argo's method offers a more complete view of street conditions and isn't reliant on the co-operation of strangers.
Argo was founded by Adibhatla and two other graduates from NYU's Center for Urban Science and Progress. They're also interested in measuring ozone levels with car-mounted sensors.
"We want to develop technology with city managers rather than for city managers. Often they have a lot of ideas, but nobody to implement them," Adibhatla says.
Cover Photo: Flickr user Lee Cannon