Marshall Islands Road Surface and Passability: AI-Derived Data for 100km of Arterial Roads

Marshall Islands arterial road network data featuring AI-derived surface and width attributes generated by HeiGIT from 2020 and 2024 PlanetScope satellite imagery. It covers approximately 100 kilometers of roads, providing surface predictions for segments missing OpenStreetMap (OSM) surface tags. The analysis focuses on motorway, trunk, primary, and secondary road classifications.

Use Cases

• Evaluating logistical accessibility for heavy vehicles using the Road Width Class (1-3) and HPI scores
• Monitoring infrastructure modernization by analyzing the Road Surface Change column for unpaved-to-paved transitions
• Filling data gaps in OpenStreetMap by utilizing DL Surface Prediction for segments lacking surface tags

Strengths

• 89.2% surface prediction accuracy
• Temporal road surface change detection between 2020 and 2024
• Standardized Humanitarian Passability Index (HPI) for logistics planning

Limitations

• Small sample size of approximately 100km of roads
• Geographic bias towards arterial road networks
• AI-derived predictions may contain errors compared to ground-level surveys

Provenance

Source

Heidelberg Institute for Geoinformation Technology (HeiGIT)

Collection Method

Deep learning models applied to PlanetScope satellite imagery (2020 and 2024) combined with OpenStreetMap data.

Time Range

2020-2024

Freshness

Last updated March 2026; includes satellite-derived data from 2020 and 2024.

Geography

Marshall Islands