The Asphalt Industry Alliance published its annual report in March 2026 with a stark headline: the cost of bringing local roads in England and Wales up to standard has risen to a record £18.6 billion. Roads are being resurfaced on average once every 97 years. Only 51% of the local road network is in good condition. In the last year, local authorities filled 1.9 million potholes. And still the backlog grew.
Traditional survey methods are part of the problem. Manual inspections are slow, infrequent and inconsistent. The result is that highways teams are always reacting rather than preventing. AI is beginning to change that calculation.
How the technology works
The approach is straightforward. Cameras are mounted on vehicles that already travel the road network. Fleet cars, bin lorries, dedicated survey vehicles. As they drive, AI processes the footage in real time, detecting surface defects, potholes, cracks, faded lines and damaged signs. The data is geo-tagged and uploaded to a management platform, giving highways officers a continuously updated picture of their network.
Gaist Solutions has been doing this across England, Wales and Scotland for several years. In 2021, it secured the first national AI-powered road survey contract in the UK and Ireland, with the Department for Infrastructure in Northern Ireland. The system captures a digital twin of every road in the country, generating billions of data points and mapping carriageway defects to centimetre accuracy.
UK Roads: The Scale of the Problem
£18.6bn: Cost to bring local roads up to standard in England and Wales (AIA, 2026)
1.9m: Potholes filled by local authorities in the past year
97 years: Average resurfacing frequency for local roads
£1.6bn: Government funding for local road maintenance in 2025/26
What councils are already doing
Surrey County Council became the first UK authority to fully replace manual inspections with AI-powered video analytics. Councillor Matt Furniss told Route Reports the technology "means we can proactively log and fix potholes, helping ensure we have well-maintained roads across the country."
Cambridgeshire County Council, facing an £800 million road maintenance backlog, awarded a £100,000 contract in early 2026 to deploy AI cameras on inspection vehicles. The technology automatically detects road defects and feeds them into the council's work ordering system.
A new standard, PAS 2161:2024, published by BSI in partnership with the DfT, has opened the door further. It is technology-neutral, meaning any compliant method, from laser scanners to AI cameras, can be used for national reporting. The DfT estimates adopting the data standard could unlock £300 million of efficiency savings annually through risk-based maintenance planning.
"It's the speed that the images come back to us. It's 30 seconds. Inspectors can call me from the road and say 'have a look at this,' and I'm already going through the images."
The responsible deployment question
AI road surveys generate enormous amounts of data about public infrastructure and, incidentally, about the vehicles and people on the roads when surveys happen. Local authorities adopting these systems need clear data governance policies covering retention, access and secondary use.
There is also a procurement question. Several companies now offer AI road survey products, and the market is developing quickly. Councils need to ask about data portability, because the value of survey data accumulates over time and switching vendors should not mean losing historical records. They should also ask how model accuracy is validated, since AI defect detection performance can vary significantly between road types and lighting conditions.
The £18.6 billion backlog is not going to be solved by AI alone. But reacting to potholes one at a time, after residents complain, is not going to solve it either. Smarter data, deployed responsibly, is what makes the shift from reactive patching to preventive maintenance possible.
Building AI Competence in the Built Environment
The Responsible with AI programme gives built environment professionals the tools, frameworks and practical guidance to govern AI confidently. Including responsible procurement of AI systems and data governance for public sector teams.
Explore the Programme → Responsible with AI
