Early Pothole Detection

Road repairs are easier and as much as five times cheaper when problems are detected early. But sending crews to continually survey streets for damage isn't practical, and minor cracks often grow into gaping holes before workers can get to them. A group headed byMing Wang, a professor of environmental and civil engineering at Northeastern University, has developed a clever solution, called Versatile Onboard Traffic Embedded Roaming Sensors (VOTERS). Here's how it works.


Most potholes start small, some beneath the pavement where drivers can't even see them. When installed in a customized van, VOTERS can scan 80 miles of road per day, identifying potential issues before they expand into tire-busting asphalt chasms."Thirty-two percent of the roads [in the U.S.] are bad," says Wang, citing a national evaluation by the American Society of Civil Engineers. "We need to interrupt that cycle."


As the van rolls over the road surface, a sensor records changes in air pressure inside the tire caused by bumps. A carefully calibrated microphone senses any jostling sounds, and a unique radar system scans the surface between the wheels, checking for problems missed by the tires, and for pockets of air or pools of water beneath the pavement–likely signs of potholes-to-be. A video camera trained on the ground behind the van offers an overall picture of crack density and helps corroborate info from the other sensors.


As the van travels, information is sent to a computer, where it's sifted through by VOTERS software. Analysts then examine the results to determine when and where to send repair crews for the most effective prevention. The info reaches city planners in days rather than months, and at a frac­tion of the cost of traditional methods. Catching problems early, before disasters occur, means long-term savings as well.


Wang is in the process of spinning off a firm that cities can contract with to perform annual road evaluations. But the real goal is to persuade auto manufacturers to embed tire sensors in new cars, thereby crowdsourcing vast, continuously updated troves of data on road quality. The resulting info would be less detailed than what VOTERS can pick up, but it would still improve road maintenance dramatically. "This is not a crazy idea," says Wang. "It could be cheap. Every car could have it–it would be that easy."


View more at FastCompany.

Related Faculty: Ming L. Wang

Related Departments:Civil & Environmental Engineering