An autonomous robot designed to tackle potholes using AI is to venture out of a controlled environment and be tested on actual roads for the first time.
The machine called ARRES (Autonomous Road Repair System) PREVENT can identify and characterise potholes and cracks using artificial intelligence.
It can then automatically fill them up to keep out surface water – which otherwise can seep through causing further damage.
Should it be successful, ARRES could save time and money identifying potholes that could worsen due to neglect, and reduce the disruption they cause to motorists.
The robot was developed by tech company Robotiz3d and academics at the University of Liverpool in partnership with Hertfordshire County Council [HCC] Highways Engineers.
It will soon roam the roads of Hertfordshire for a “real-life” road repair on a residential street in the county.
So far, it is still in its pilot phase tested extensively under lab conditions.
In development since 2020, the robot is the first of its kind in the world.
Read more:
Robots may improve mental wellbeing
Artificial intelligence used to deter deer from rail lines
Phil Bibby of Highways at Hertfordshire County Council, said the technology “could be exactly what we need to ensure our road network remains one of the best in the country”.
“We know this issue matters to our residents, so it matters to us too,” Mr Bibby added.
What makes a pothole?
Potholes are formed in multiple stages, beginning with the cracking of the road surface due to age, poor drainage and utility works among other reasons.
Water enters the surface and, in winter, ice can seep through the cracks as the roads become too cold.
As traffic continuously pounds the roads, small blocks of asphalt are flicked out.
Be the first to get Breaking News
Install the Sky News app for free
The final nail in the tarmac is water ponding – where rain floods an already-formed hole – which expands and deepens it.
The UK is burdened with potholes due to high traffic levels and its cold and wet climate during winter months.
