Fixing city roads can be a long and costly ordeal, but Sedona has piloted a system to improve the process.

The city hired RoadBotics Inc., and its specialty is assessing road conditions via data collection and machine learning. 

“They go out in regular passenger vehicles and they have video cameras in those vehicles and they go through the streets and video the streets," said public works director and city engineer Andy Dickey. "They then take that recorded video data and run it through some algorithms that assess the pavement condition”

Dickey says the process usually takes humans six months and costs twice as much as the artificial intelligence system, which took a few days and cost $6,000.

“The observations are completed in a matter of a few days and then the data is sent to us within a couple weeks, so it’s much faster and therefore much cheaper than having people do it.”

The system uses passenger vehicles to drive on the roads and collect video footage, which is then processed with deep-learning algorithms.