Liberia Arterial Road Surface and Passability Data (2020-2024)

HeiGIT produced this dataset of AI-derived road surface, width, and passability metrics for approximately 3,400 km of arterial roads in Liberia using 2020 and 2024 PlanetScope satellite imagery. It integrates OpenStreetMap (OSM) segments with deep-learning predictions for motorway, trunk, primary, and secondary road classes to fill data gaps in national infrastructure maps.

Use Cases

• Monitoring national infrastructure progress using the Road Surface Change indicator to identify unpaved-to-paved transitions
• Optimizing heavy vehicle routing by filtering for Road Width Class 3 (>5.5 m) segments
• Assessing regional accessibility for aid delivery using the Humanitarian Passability Index (HPI) scores

Strengths

• AI surface model achieves 89.2% accuracy compared to 64.7% for OSM tags
• Provides surface data for 92% of road segments missing tags in OSM
• Includes multi-temporal predictions (2020 and 2024) to track infrastructure changes

Limitations

• Geographic scope is limited to arterial road networks
• AI-derived predictions have a documented 10.8% error rate
• Humanitarian Passability Index is a derived model and may not reflect real-time weather-related closures

Provenance

Source

HeiGIT (Heidelberg Institute for Geoinformation Technology)

Collection Method

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

Time Range

2020-2024

Freshness

Last updated March 2026; contains imagery-derived data from 2020 and 2024.

Geography

Liberia