Slovenia Road Surface and Passability Metrics for 5,500 km of Roads

Produced by HeiGIT, this dataset contains AI-derived surface, width, and passability metrics for 5,500 km of Slovenian arterial roads based on 2020 and 2024 PlanetScope imagery. It supplements OpenStreetMap data by providing surface predictions for segments lacking tags, achieving a validated accuracy of 89.2%.

Use Cases

• Logistics planning for heavy transport using 'Road Width Class' to identify segments wider than 5.5 meters
• Humanitarian response prioritization based on 'Humanitarian Passability Index (HPI)' scores from 1 to 6
• Infrastructure development monitoring by identifying transitions from unpaved to paved via the 'Road Surface Change' indicator

Strengths

• 89.2% accuracy for AI-derived surface predictions compared to 64.7% for standard OSM tags
• Provides surface data for 20.8% of roads that lack tags in OpenStreetMap
• Includes multi-temporal comparisons between 2020 and 2024 for infrastructure monitoring

Limitations

• Geographic scope is restricted to arterial roads only
• AI-derived width and surface predictions may have errors in areas with dense canopy cover or extreme terrain
• License is listed as hdx-other, which may imply specific usage restrictions

Provenance

Source

Heidelberg Institute for Geoinformation Technology (HeiGIT)

Collection Method

Deep learning models applied to PlanetScope satellite imagery (3m resolution) integrated with OpenStreetMap vector data.

Time Range

2020 - 2024

Freshness

Last updated March 2026; utilizes satellite imagery from 2020 and 2024.

Geography

Slovenia