Skip to main content

Road Health Data Product

Road quality data generated from connected car data

Updated over 2 weeks ago

What is Road Health Data?

Road Health Data offers insights into road quality through connected car vehicle data. It empowers road operators, municipalities, and infrastructure managers to monitor road conditions in near real-time and to plan maintenance activities more efficiently.

Typical applications include:

  • Asset Inventory: Maintain an up-to-date view of the road network, assess seasonal wear (e.g., winter impact), validate maintenance plans, and address citizen complaints.

  • Maintenance Follow-up: Evaluate newly paved roads, confirm proper execution of roadworks, and receive early alerts when roads begin to degrade.

  • KPI & Trends: Objectively assess road network conditions, establish trends from daily measurements, and benchmark city versus national levels for budget planning.

  • Sudden Change Alerts: Get notified of hazardous conditions such as potholes, unexpected roughness changes, or road obstacles, and respond promptly using integrated online tools.


Data Source

The data is sourced from millions of vehicles globally through Mobito’s vetted partners, including:

  • Automotive OEMs – Vehicles with embedded connectivity directly from the manufacturer.


Data Specs

  • Location granularity: 25-meter road segments

  • Geographies: Belgium, Greece

  • Anonymisation: Road surface data is segmented and anonymised. It is GDPR-compliant and completely free of personal information.


Datasets Overview

The Road Health Data Product consists of three complementary datasets. Customers can select individual datasets or receive all three combined.

Dataset 1: Roughness Data

Provides daily and long-term measurements of road roughness across the network.

Indicative fields include:

  • roadName: The road’s name

  • segmentId / subSegmentId: Identifiers for each 25-meter road segment

  • roughnessDailyAverage: Average roughness for the day

  • roughnessLongTermEstimate: Smoothed estimate for long-term evaluation

  • geometryWkt: WKT LineString geometry for the segment

Dataset 2: Roughness Change Alerts

Highlights improvements or deteriorations in road conditions, enabling proactive maintenance.

Indicative fields include:

  • alertId: Unique identifier of the alert

  • alertType: Improvement or degeneration

  • severity: Severity level of the change

  • alertLevel: Classification of the alert

  • geometryWkt: WKT LineString geometry of the affected segment

Dataset 3: Road Health Anomalies

Captures unexpected or hazardous conditions detected on the road network.

Indicative fields include:

  • id: Unique anomaly identifier (UUID)

  • firstDetection / lastDetection: Timestamps of anomaly occurrence

  • position: Anomaly position reference

  • type: Anomaly type (e.g., pothole, obstacle)

  • uncertaintyInMeter: Uncertainty area of the anomaly

  • count: Number of detections

  • hitRatio: Ratio of confirmed detections

  • bearing: Bearing of the road segment

  • severity: Severity level of the anomaly

  • geometryWkt: WKT LineString geometry


Dashboard Visualisation (optional add-on)

In addition to raw datasets, the Road Health Data Product can be delivered with an interactive dashboard. This offers an intuitive way to explore road conditions without additional data processing.

Key features include:

  • Up-to-date IRI measurements for every 25-meter road segment

  • Heatmap visualisation of road health conditions on an interactive map

  • Road health historicity: Trends and history for each road segment

  • Highlighted problematic areas across the network

  • Road network health scoring for quick assessment

  • Smart data filtering to refine by geography, time, or road type

  • Data export for offline analysis or integration with other tools


Anonymisation & Privacy

All data is anonymised and aggregated at the road-segment level. The product is fully GDPR compliant and does not include any personal information. More details are available in our Data Anonymisation Rules article.

Did this answer your question?