DRC Road Surface Classification from OSM and Deep Learning

A 2026 dataset from HeiGIT provides AI-derived road surface classifications for the Democratic Republic of the Congo, covering approximately 0.6182 million kilometers of roads from OpenStreetMap. It distinguishes paved and unpaved roads using a hybrid deep learning model based on Mapillary imagery, augmented with urban classification layers. The data is intended for transportation planning, infrastructure analysis, and GIS applications.

Use Cases

• Transportation planning based on paved vs. unpaved road classifications.
• Infrastructure analysis and investment prioritization based on road surface quality.
• Climate emissions assessment for vehicle fleets based on road surface type.
• GIS applications for routing and logistics based on combined surface classifications.
• Urban/rural development studies based on the integrated urban classification attributes.

Strengths

• Covers approximately 0.6182 million kilometers of roads in the DRC.
• Provides AI-derived estimates for paved (0.0114 million km) and unpaved (0.2246 million km) roads.
• Integrates multiple data sources: OSM, deep learning predictions from Mapillary, and urban layers (GHSU, AFRICAPOLIS).
• Includes multiple classification schemes (OSM-priority, DL-priority) for flexibility.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Approximately 0.3822 million km (61.8246%) of road surface information is missing in the underlying OSM data.

Provenance

Source

HeiGIT (Heidelberg Institute for Geoinformation Technology)

Collection Method

Derived from OpenStreetMap (OSM) and augmented with deep learning predictions based on Mapillary street-level imagery, combined with urban classification layers.

Time Range

The dataset reflects the state of OSM and Mapillary imagery at the time of processing; specific temporal coverage is not stated.

Freshness

Last updated 2026-03-02 17:23:34.263251; freshness should be verified.

Geography

Democratic Republic of the Congo.