Sierra Leone Road Surface and Passability Data for 3,700 km of Arterials

AI-derived road surface, width, and passability metrics cover approximately 3,700 km of arterial roads in Sierra Leone. Produced by HeiGIT, the data integrates OpenStreetMap attributes with deep-learning predictions generated from 2020 and 2024 PlanetScope satellite imagery. It specifically targets the national road backbone, including motorway, trunk, primary, and secondary classifications.

Use Cases

• Evaluating logistical accessibility for humanitarian aid delivery using the Humanitarian Passability Index (HPI) scores
• Monitoring national infrastructure development by analyzing the Road Surface Change column between 2020 and 2024
• Planning transport routes for heavy vehicles based on the Road Width Class (1-3) designations

Strengths

• Surface model achieves 89.2% accuracy, significantly higher than the 64.7% accuracy of standard OSM surface tags
• Includes multi-temporal predictions (2020 and 2024) allowing for infrastructure change detection
• Provides standardized width classes (1-3) to assess heavy vehicle flow capacity

Limitations

• Excludes minor, residential, and unclassified roads not part of the arterial network
• AI predictions may be affected by environmental factors like cloud cover or dense tropical canopy
• Geographic scope is restricted to Sierra Leone

Provenance

Source

Heidelberg Institute for Geoinformation Technology (HeiGIT)

Collection Method

Deep learning models applied to PlanetScope satellite imagery combined with OpenStreetMap road network data.

Time Range

2020-2024

Freshness

Contains data updated as of March 2026, reflecting satellite observations from 2020 and 2024.

Geography

Sierra Leone