Using AI and Street View to Manage Road Infrastructure

Predictive maintenance is an increasingly common sight in industrial facilities around the world, but the ability for AI to detect when machinery is about to fail relies upon a steady stream of data. One would imagine the data provided by Google Street View might not be up to the job therefore, but new research from RMIT suggests otherwise.

The authors propose using data from Google Street View to spot when road furniture needs replacing. The authors highlight the often manual and laborious task of monitoring street furniture and traffic infrastructure.

In their paper, they reveal that their AI-based system was able to accurately spot the road signs in images 96% of the time whilst being able to identify the type of sign 98% of the time complete with its geolocation. Whilst the system was only trained to spot ‘stop’ and ‘give way’ signs, the authors believe this can easily be expanded to identify a range of road infrastructure.

“(Municipal authorities) have requirements to monitor this infrastructure but currently, no cheap or efficient way to do so,” they explain. “By using free and open source tools, we’ve now developed a fully automated system for doing that job, and doing it more accurately.”

Road Maintenance

The paper also highlighted how inadequate much of the information in street sign databases was, with the GPS location data often out by up to 10m, which makes maintenance that bit harder.

“Tracking these signs manually by people who may not be trained geoscientists introduces human error into the database. Our system, once set up, can be used by any spatial analyst — you just tell the system which area you want to monitor and it looks after it for you,” the authors say.

While Google Street View images are updated relatively infrequently, the authors reveal that a number of local authorities are installing cameras on garbage collection trucks to help ensure a ready supply of up to date information to feed the AI technology.

“This imagery is critical for local governments in monitoring and managing assets and with the huge amount of geospatial applications flourishing, this information will only become more valuable,” the team explains. “Ours is one of several early applications for this to meet a specific industry need but a whole lot more will emerge in coming years.”

If more authorities begin to capture data from garbage trucks, then this can easily be fed into the system to provide a richer data set to work from and ensure that infrastructure maintenance teams have all the help they need.

This UrIoTNews article is syndicated fromDzone