From small towns to major cities, road inspections are a big challenge in terms of dollars and distance. South Bend, Ind., turned to RoadBotics as an alterative to sending Department of Public Works employees down every road for visual inspections.
The company’s solution still relies on people to drive the vehicles, but the road inspection itself is performed via images taken by a smartphone mounted to the car’s windshield. Those images are uploaded to the cloud, where the company uses machine learning to identify where roads are in need of repair.
“This is similar to what we do,” said Jitin Kain, the city’s deputy director of public works, during an interview with a local TV station. “It just creates more efficiency in how the analysis of road conditions is done.”
The system has evolved significantly since North Huntingdon, Pa., first piloted it in 2016. “The level of detail and specificity used to diagnose a problem has dramatically increased,” said Mark DeSantis, CEO and co-founder of RoadBotics. “It can see objects [and] patterns in things and make the same kind of judgment that a human can make.”
South Bend and other cities could soon see even greater efficiencies, with the monitoring being performed by street sweepers and other vehicles that are “already driving the roads,” DeSantis said.
Kain added that such continuous improvements fit well with South Bend’s plan “to be a beta-test city for…new technology.”
Fuente de la noticia: https://gcn.com/articles/2018/11/01/psi_roadway-monitoring-with-ai-and-iot.aspx