Sao Tome and Principe: AI-Derived Road Surface and Passability Data

HeiGIT generated this geospatial dataset covering approximately 200 km of arterial roads in Sao Tome and Principe using 2020 and 2024 PlanetScope satellite imagery. It integrates OpenStreetMap (OSM) data with AI-derived predictions for road surface type, width, and logistical passability. The analysis focuses on motorway, trunk, primary, and secondary road classifications.

Use Cases

• Logistics planning using the Humanitarian Passability Index (HPI) to assess route accessibility for aid delivery
• Infrastructure development tracking by analyzing 'Road Surface Change' between 2020 and 2024
• Fleet routing optimization based on 'Road Width Class' to identify segments suitable for heavy trucks

Strengths

• 89.2% surface prediction accuracy
• Multi-temporal coverage for 2020 and 2024
• Fills surface tag gaps for 23.6% of OSM segments

Limitations

• Geographic scope limited to Sao Tome and Principe
• Excludes minor and residential road classes
• AI-derived predictions may contain errors despite high reported accuracy

Provenance

Source

HeiGIT (Heidelberg Institute for Geoinformation Technology)

Collection Method

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

Time Range

2020-2024

Freshness

Last updated March 2026; imagery analysis covers the period between 2020 and 2024.

Geography

Sao Tome and Principe