Sri Lanka Road Surface and Passability Data for 14,000 km of Arterial Roads

Supplying AI-derived road surface, width, and passability metrics for approximately 14,000 km of arterial roads in Sri Lanka, produced by HeiGIT using 2020 and 2024 PlanetScope satellite imagery. It covers motorway, trunk, primary, and secondary road classes, filling gaps for the 39.8% of roads lacking surface tags in OpenStreetMap.

Use Cases

• Monitoring infrastructure progress by identifying transitions in the Road Surface Change column
• Optimizing disaster response routes using the Humanitarian Passability Index (HPI) to identify high-passability corridors
• Enhancing OpenStreetMap metadata by populating missing surface tags using DL Surface Prediction 2024

Strengths

• 89.2% accuracy for the PlanetScope surface model
• Covers 14,000 km of arterial roads
• Provides temporal change detection between 2020 and 2024

Limitations

• Restricted to arterial road classes only
• Geographic scope limited to Sri Lanka
• AI-derived predictions may contain 'Unknown' classifications

Provenance

Source

Heidelberg Institute for Geoinformation Technology (HeiGIT)

Collection Method

Deep learning analysis of PlanetScope satellite imagery combined with OpenStreetMap road network data.

Time Range

2020-2024

Freshness

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

Geography

Sri Lanka