Georgia Road Surface, Width, and Passability for 7,300 km of Arterial Roads

HeiGIT produced this dataset covering approximately 7,300 km of arterial roads in Georgia using PlanetScope satellite imagery from 2020 and 2024. It provides AI-derived classifications for road surface, width, and passability to supplement OpenStreetMap data. The analysis focuses specifically on motorway, trunk, primary, and secondary road classes.

Use Cases

• Assessing logistical accessibility for heavy vehicles using the 'Humanitarian Passability Index (HPI)' and 'Road Width Class'
• Monitoring infrastructure development by analyzing 'Road Surface Change' between 2020 and 2024
• Improving routing algorithms by integrating 'DL Surface Prediction' for segments where the 'surface tag' is null in OSM

Strengths

• High model accuracy of 89.2% for road surface classification
• Multi-temporal coverage (2020 and 2024) allowing for infrastructure change detection
• Provides surface data for 7,100 km of roads, filling gaps in OpenStreetMap tags

Limitations

• Limited to arterial road classes, excluding the majority of the local street network
• AI-derived predictions may contain classification errors compared to ground-truth surveys despite high reported accuracy

Provenance

Source

HeiGIT (Heidelberg Institute for Geoinformation Technology)

Collection Method

AI-derived analysis of PlanetScope satellite imagery and OpenStreetMap data

Time Range

2020-2024

Freshness

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

Geography

Georgia