Japan Road Surface Classification from OSM and AI Predictions

Approximately 2.3432 million kilometers of roads in Japan are mapped, with AI-derived estimates classifying 17.1957% as paved and 1.2527% as unpaved. The dataset, created by HeiGIT and last updated in March 2026, augments OpenStreetMap data with deep learning predictions from Mapillary imagery to address missing surface information. It includes attributes for highway type, surface classification, and urban-rural context.

Use Cases

• Transportation planning based on paved versus unpaved road classifications.
• Infrastructure analysis based on highway type, surface, and smoothness attributes.
• Climate emissions assessment based on road surface data and urban classification.
• GIS applications for routing and logistics based on geospatial road network data.

Strengths

• Covers approximately 2.3432 million kilometers of roads in Japan.
• Provides AI-derived surface classifications for paved and unpaved roads.
• Augments OSM data with an additional 0.0405 million km of surface information from Mapillary.

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.

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.

Time Range

Mean timestamp of predictions (DL_mean_timestamp) is included, but overall temporal coverage is not specified.

Freshness

Last updated 2026-03-02 17:05:02.537976; freshness should be verified.

Geography

Japan