Robotaxis and road data: how AI helps detect potholes

Robotaxis can become a mobile road-monitoring network: sensors and AI detect potholes and turn the data into an operational advantage for cities and mobility.

Robotaxis and road data: how AI helps detect potholes
Robotaxi detecting road potholes and sharing data for urban maintenance

Autonomous vehicles are evolving from simple means of transport into mobile data-collection platforms. As an automated driving system travels through the city, it can observe the quality of the road surface, pick up anomalies and return useful information to those who manage infrastructure.

This approach raises a strategic issue: using connected fleets not only to move people, but also to produce operational insight on safety, maintenance and urban planning. For decision makers and tech teams, the value lies not only in detection, but in the ability to integrate, validate and distribute the data reliably.

From sensor to operational indicator

A robotaxi has a combination of cameras, lidar, radar and vehicle dynamics signals. By cross-referencing these inputs, the system can recognise changes in the road surface, abnormal jolts and critical points along the route.

Pothole detection is a concrete example of how machine-based perception goes beyond the single event: multiple passes over the same stretch of road produce a statistical picture useful for distinguishing a temporary defect from a persistent problem.

Why it matters for cities

Local authorities often depend on manual reports or periodic inspections, which have clear limits in coverage and timeliness. A network of vehicles already on the road can fill these information gaps and create an almost continuous map of problem areas.

For cities, this means clearer priorities for interventions, shorter response times and a more robust data base for allocating budgets and maintenance crews.

Data quality and validation

Sharing this data only makes sense if the collection process is controlled. A useful road report must be geolocated accurately, de-duplicated and confirmed by several independent observations.

In this context, the contribution of end users can be an additional layer of verification. The goal is not to replace public maintenance, but to increase the reliability of the data before it enters decision-making systems.

The metrics to manage

For those designing the solution, the critical points are clear: false positives, geographic coverage, update latency and interoperability with municipal systems. Without these elements, the perceived quality of the service risks falling short of its technical potential.

Transparent governance is also needed over data ownership, access, retention and permitted uses, especially when the information comes from assets operating on public roads.

Implications for mobility and business

The strategic value of this logic goes beyond maintenance. An autonomous mobility operator can turn a safety requirement into a relationship asset with institutions, improving its credibility during expansion.

At the same time, the collected data can feed map platforms, driver services and urban monitoring systems. In other words, a single event detected by one vehicle can have effects across several products and several stakeholders.

For decision makers, this model highlights a broader trend: connected transport systems do not only sell rides, they also sell infrastructure intelligence.

Conclusions

The case shows that autonomous mobility can become a structured source of data for the smart city. The combination of sensors, distributed validation and integration with navigation platforms creates an information flow with high operational impact.

  • Autonomous fleets can detect road problems continuously.
  • Data quality depends on validation, geolocation and governance.
  • Cities benefit from faster prioritisation and more targeted maintenance.
  • For operators, sharing data can strengthen institutional relationships.
  • Connected transport is also becoming information infrastructure.