In Brief
Citroën proposes using artificial intelligence, via its STAN app, to detect pothole nests early and improve road maintenance in the United Kingdom. This approach, embedded in the “Better Roads Manifesto,” relies on collecting and analyzing road imagery to help authorities intervene faster and more effectively. The goal is to reduce the societal costs of accidents and repairs while delivering safer, better-maintained roads. There are international examples that already prove the effectiveness of such systems. However, adoption will depend on local governments’ willingness to weave these new technologies into their daily management.
Who hasn’t cursed a pothole that makes the whole cabin tremble or wrecks a rim? In the UK, these road defects do more than irritate drivers. They would be involved in 6% of fatal or serious crashes. The total cost is estimated at £54.7 billion per year, roughly €64 billion for society (about $68–70 billion USD).
In light of this, Citroën is bringing the debate into territory unusual for an automaker: data and artificial intelligence. With its “Better Roads Manifesto,” the brand with the chevrons urges British road authorities to lean on the STAN app. Objective: detect potholes much earlier and move to true preventive maintenance. Behind this push lie billions in potential savings, as well as a concrete promise for fleets and everyday drivers alike.
Rough Roads, High Costs: Citroën’s AI Aims to Change the Game
The Citroën manifesto starts with a chiffre that resonates with fleet managers. According to the automaker, faster detection of road defects could translate into substantial savings. A better organization of repairs would push total savings to at least £5.5 billion (about €6.4 billion; around $6.8–7.0 billion USD). The gains would come from fewer accidents and more efficient maintenance. For fleet operators, this means suspensions and tires needing replacement less often, more available vehicles, and easier planning of routes.
To back its manifesto, Citroën has filed information-access requests with the 154 local road authorities. The responses reveal a highly fragmented system. Only 49% of services use geolocation to log potholes. Nearly half do not precisely map the observed defects. Repair times range from a few hours to three months. One extreme case even shows a wait approaching a full year. The very definition of a pothole varies as well—some are addressed at 40 mm depth, others only when the hole is significantly larger. The perception data are telling too. A study commissioned by the automaker shows that 28% of motorists have already lost control of their vehicle to avoid a pothole or debris on a damaged road. 56% would be willing to pay higher local taxes if the money were earmarked entirely for road maintenance. 49% would also accept a rise in vehicle tax if all funds went toward road improvements.
Citroën even cites £22.7 million in compensation paid to drivers in 2022 because of potholes. That amounts to roughly €26 million, enough to fund hundreds of thousands of repairs.
How AI and the STAN App Plan to Track Potholes in Real Time
To translate these observations into action, Citroën relies on the AI-powered road-tracking platform STAN. The concept is straightforward: a smartphone app lets drivers automatically capture images of the roadway as they drive. The artificial intelligence analyzes these images, spots cracks, holes, and other degradations, and then links each defect to a precise GPS location. All this data flows into a database accessible to the authorities, enabling them to identify budding problems and prioritize repairs before potholes become truly dangerous.
In other cities, similar setups have already proven their worth. In Memphis, Tennessee, a road-analysis system using video driven by AI filled 63,000 potholes in one year. In Orléans, France, vehicles equipped with sensors and vision algorithms are already patrolling the streets. Their use has avoided roughly 30,000 agent trips per year. The idea is the same: harness mobile sensors and AI to continuously monitor roadway conditions, aiming for earlier intervention, on smaller surface areas, and at lower cost.
The “Better Roads Manifesto” isn’t limited to touting an app. The document outlines six recommendations to improve the British local road network. These include the systematic use of AI for inspections, stronger oversight of roadworks contractors, and the creation of a national pothole repair standard. The objective is explicit: move away from a costly emergency-fix mentality toward planned, multi-year maintenance based on shared data. “Our roads are a constant source of concern for drivers,” says Greg Taylor, UK Director of Citroën. “Raising more funding alone isn’t realistic, so we propose a few simple, economical, and logical steps that the national government, local road authorities, and the public can back to give the country the quality of roads we all deserve,” Taylor adds.
For businesses and individuals alike, the promise is very tangible: fewer rim damages, fewer unexpected breakdowns, and safer trips. The lingering question is one no one can answer for the authorities: how quickly will they adopt these AI tools to transform their roads on a daily basis?
Comparison Table
| Indicator | Explicit Value | Scope |
|---|---|---|
| Share of fatal or serious crashes involving potholes | 6% | United Kingdom |
| Annual total cost of degraded roads | £54.7 billion (about €64 billion; about $68–70 billion USD) | United Kingdom |
| Potential savings from early detection | £5.5 billion (about €6.4 billion; about $6.8–7.0 billion USD) | UK, per Citroën |
| Proportion of services using geolocation to log potholes | 49% | British local authorities |
| Drivers who lost control to avoid a pothole | 28% | Study commissioned by Citroën |
| Drivers willing to pay more local tax for road maintenance | 56% | Study commissioned by Citroën |
| Drivers accepting a vehicle tax increase if all funds go to road improvements | 49% | Study commissioned by Citroën |
| Compensations paid in 2022 due to potholes | £22.7 million (about €26 million) | United Kingdom |
| Potholes filled in Memphis in one year (AI/video) | 63,000 | Memphis, USA |
| Avoided trips in Orléans thanks to AI sensors | 30,000/year | Orléans, France |
Key Takeaways
- Citroën aims to revolutionize road maintenance through AI and the STAN app.
- Damaged roads cost roughly €64 billion per year to UK society.
- Preventive pothole detection could generate up to €6.4 billion in savings.
- Only 49% of local authorities actively use geolocation to track potholes.
- The initiative is part of the Better Roads Manifesto, which outlines six road-improvement measures.
- Memphis and Orléans examples already show the potential of AI to monitor and repair the road network.
FAQ
How does Citroën’s STAN app detect potholes?
The STAN app enables drivers to automatically capture images of the roadway, which the AI analyzes to identify defects and geolocate them in real time.
What are the benefits expected from AI for road maintenance?
AI should enable faster detection of deteriorations, promote preventive maintenance, reduce accidents, and deliver substantial savings for society and fleets.
What is the estimated annual cost of degraded roads in the UK?
The total cost is estimated at £54.7 billion per year, about €64 billion for society.
What are the main recommendations of the “Better Roads Manifesto”?
The manifesto calls for, among other things, the systematic use of AI for inspections, better oversight of roadwork contractors, and the creation of a national pothole repair standard.
How many potholes were filled in Memphis thanks to an AI-driven system?
In Memphis, an AI-powered video analysis system filled 63,000 potholes in one year.
