What is Road Damage Detection data?
Description
Road Damage Detection data provides insights into where and what type of road surface issues are present along road networks. This includes both road damage (e.g. potholes, cracks) and road assets (e.g. speed bumps and other obstacles on the road surface).
The dataset is an aggregated view of persistent road obstacles. Each record corresponds to a unique obstacle location detected over time. As a result, an obstacle first detected in February and last seen in March will still be included in a July delivery as a known obstacle location, even if it has not been detected recently. This approach allows for analysis of both recently active and historical obstacles.
Data Source
The data is sourced from millions of vehicles globally via our vetted partners including:
Automotive OEMs; vehicles with embedded connectivity from the car manufacturer
Fleets with after-sales in-vehicle hardware telematics devices
Use Cases
Key applications include:
Road Maintenance & Infrastructure Planning: Road operators and municipalities can identify and prioritise road sections requiring repair, enabling cost-effective maintenance scheduling.
Road Safety Analysis: Traffic safety analysts and government agencies can assess the distribution and severity of road surface hazards to support targeted safety interventions.
Navigation & Routing: Navigation platform providers can incorporate road damage and obstacle data to improve route quality and driver experience.
Smart City & Mobility Analytics: SaaS analytics platforms and urban planners can monitor road condition trends over time and correlate infrastructure quality with traffic and mobility patterns.
Road Damage Detection Data Schema
Main Attributes
The following table provides the main data attributes available via the Mobito Road Damage Detection Data Product:
Attribute | Type | Description |
latitude | float | Estimated latitude of the road damage location |
longitude | float | Estimated longitude of the road damage location |
heading | integer | Estimated direction of driving in degrees (0–359, 0=North, 90=East) |
first_detection_date | date | Date of the first time the event was detected (ISO 8601) |
last_detection_date | date | Date of the last time the event was detected (ISO 8601) |
no_of_vehicles | integer | The number of unique vehicles that detected the damage |
no_of_detections | integer | The total number of detections |
avg_severity | float | The average severity of detected damage (scale 0–3) |
road_type | string | The type of road |
road_name | string | The name of the road |
city | string | The name of the city |
municipality | string | The name of the municipality |
state | string | The name of the state / province |
class | string | The road damage classification. Possible values: |
Data Sample
A small sample is available at the link below for download, which can help you better understand the data schema.
Road Damage Detection Data Specs
Historicity
Historical data availability as of January 2025. Availability differs per country.
Location Granularity
Each obstacle is mapped to an exact GPS location, expressed as latitude and longitude coordinates, along with a heading, indicating the estimated direction of travel at the point of detection.
Delivery Frequency
Data is delivered on a daily basis or lower frequency, depending on the agreed delivery configuration.
Vehicle Types
The following vehicle types are available in the datasets (availability varies per country):
Passenger car
Light commercial vehicle (LCV). This is a van type of vehicle. Examples:
(Renault Master, Renault Trafic, Mercedes-Benz Sprinter, Peugeot Boxer, Volkswagen Crafter, Fiat Ducato, Opel Crossland, Citroën Berlingo etc.)
Heavy truck (more than 3.5 tonnes)
Bus
Other (unclassified)
FAQs
What is the difference between road_damage and road_asset?
What is the difference between road_damage and road_asset?
road_damage refers to unintended deterioration of the road surface, such as potholes or cracks. road_asset refers to intentional road infrastructure features that may act as obstacles, such as speed bumps.
How recent is the data?
How recent is the data?
Each record includes a last_detection_date field indicating when the obstacle was most recently observed. Records are retained in the dataset even if not recently detected, allowing you to distinguish between active and historical obstacles.
Can I access a sample of this data product?
Can I access a sample of this data product?
Yes! Mobito offers a free 3-day data sample for your preferred location. Send us a message and a Mobito expert will assist you in arranging the details.
For details on how to access the sample data, see Mobito Data Delivery Methods