Skip to the content

Thurrock to use deep learning to prevent pothole problems


Council set to run trial using intelligent software and cameras set on its bin lorries

Thurrock Council is to run a trial on the use of intelligent software to identify spots of deterioration in its roads before they develop into potholes.

The council has received £183,000 from the Department of Transport to support the trial, which it is running with software developer Gaist and consultancy Social, Environmental & Economic Solutions (SOENECS).

The trial involves fitting high definition cameras to the council’s refuse collection vehicles to build a library of pictures of roads and pavements in the borough. They will be fed into an integrated navigation and intelligent software system and subjected to deep learning to help council officers understand the advanced signs of potholes.

This can be used in identifying vulnerable spots and ensuring they are repaired before they become potholes.

Advanced data

Dr Stephen Remde, director of innovation and research at Gaist, said: "This project is really exciting and will capture the highest ever levels of technically advanced data that will provide us with a real insight into how roads deteriorate and defects form such as potholes, surface durability and day to day traffic volume damage.

"Computer vision technology is advancing rapidly and we seek to capitalise on new 'Deep Learning' data analysis techniques we have, to analyse and manage the huge volumes of video and related data that can be used to improve the safety of roads and provide more cost-effective repairs."

Thurrock’s leader, Rob Gledhill, said it is the first initiative of its kind and that the council share what it learns with other local authorities.

 Image by Andrew Skudder, CC BY 2.0 through flickr

Register For Alerts

Keep informed - Get the latest news about the use of technology, digital & data for the public good in your inbox from UKAuthority.