France Road Surface and Passability Data for 229,100 km of Arterial Roads

This dataset maps 229,100 km of arterial roads in France, providing AI-derived surface types, widths, and passability scores produced by HeiGIT. It utilizes PlanetScope satellite imagery from 2020 and 2024 to classify motorways, trunks, and primary/secondary roads. The data fills critical gaps where OpenStreetMap (OSM) surface tags are missing, achieving 89.2% classification accuracy.

Use Cases

• Evaluating logistical constraints for heavy transport using the Road Width Class and HPI scores
• Identifying infrastructure improvements by tracking the Road Surface Change indicator
• Enhancing navigation datasets by substituting missing OSM surface tags with DL Surface Prediction values

Strengths

• 89.2% surface classification accuracy
• Covers 229,100 km of road network
• Multi-temporal comparison between 2020 and 2024
• Fills surface data for 39.7% of roads missing OSM tags

Limitations

• Geographic focus restricted to arterial road classes
• Significant class imbalance with only 0.28% of roads identified as unpaved
• AI-derived predictions may have errors in areas with dense canopy or cloud cover

Provenance

Source

Heidelberg Institute for Geoinformation Technology (HeiGIT)

Collection Method

AI-derived predictions from PlanetScope satellite imagery combined with OpenStreetMap vector data

Time Range

2020-2024

Freshness

Last updated March 2026; imagery covers 2020 and 2024.

Geography

France