Structuring Road Information in Open Data: A Nested Wikidata – OSM – BD TOPO (IGN) Architecture Co-produced by Territorial Authorities

Room: Talks II - Amphi Bienvenüe (Bienvenüe)

Friday, 11:15
Duration: 20 minutes (plus Q&A)

Language: en


Back to schedule
  • Jean-Louis ZIMMERMANN

Road infrastructure data held by French territorial authorities is rich, authoritative and largely invisible to the open data ecosystem. This talk presents the journey of an active OSM and Wikidata contributor who also works in road information management at municipal, intermunicipal and departmental levels in Southern France. This dual role has led to the design of a coherent, nested open data architecture for road intelligence. At the core lies a data model in which Wikidata acts as the geometric abstraction layer enabling convergence of non-official data sources and multi-database identifiers — its design encouraging exhaustiveness through sourced statements and a completeness gauge that makes quality progress both visible and rewarding. OpenStreetMap holds the geometric and operational details, while BD TOPO (IGN) serves as the authoritative reference dataset to which both OSM geometries and Wikidata items are systematically cross-referenced and linked. This nesting creates a powerful convergence: the State’s national reference, locally produced data from territorial authorities, and their near real-time mirroring through OSM geometry and Wikidata semantics. Panoramax street-level imagery feeds both OSM mapping tasks and change detection workflows — including early experiments with AI-assisted approaches applied to infrastructure monitoring. The presentation covers modelling choices at the Wikidata–OSM interface, automation pipelines that reduce friction between internal road management tools and open data outputs, and practical benefits for road users across GPS and navigation applications. This talk targets contributors interested in Wikidata–OSM integration, open government data processes, and the practical use of street-level imagery for infrastructure maintenance. It will also openly address methodological limits, work organization practices, and current gaps in tools within and around the OpenStreetMap ecosystem — including possible ways forward, and an honest look at tensions between the pace of field experimentation and the deliberative nature of community consensus.