Lesotho Road Surface Classification from OSM and Deep Learning

Detailed information on road surfaces in Lesotho, distinguishing between paved and unpaved roads. The dataset originates from OpenStreetMap and is augmented with deep learning predictions based on Mapillary imagery, created by the Heidelberg Institute for Geoinformation Technology (HeiGIT). It was last updated on March 2, 2026.

Use Cases

• Transportation planning based on road surface type and length statistics.
• Infrastructure analysis using attributes like highway type, surface, and width.
• Climate emissions assessment leveraging the distinction between paved and unpaved road networks.
• GIS applications utilizing the geospatial road network data with urban/rural classification.

Strengths

• Includes AI-derived surface classification for approximately 0.067 million km of roads.
• Provides specific statistics: 0.0053 million km paved (7.9664%) and 0.0131 million km unpaved (19.589%).
• Combines multiple data sources: OSM, Mapillary imagery, and urban settlement layers.

Limitations

• A significant portion (0.0486 million km, 72.4446%) of road surface information is missing in the base OSM data.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and file size are unknown, which may limit suitability assessment.

Provenance

Source

OpenStreetMap (OSM), augmented with deep learning predictions from Mapillary imagery and urban layers from GHSU and AFRICAPOLIS.

Collection Method

Surface classification derived using a hybrid deep learning approach, with methodology detailed in a referenced arXiv paper.

Time Range

The mean timestamp of predictions used is included (DL_mean_timestamp), but the overall temporal coverage is not explicitly stated.

Freshness

Last updated 2026-03-02 17:19:47.836862; freshness should be verified.

Geography

Lesotho.