Making OpenDRIVE HD map data easily accessible in GIS

Abstract

We provide an open-source software extension which enables translation of OpenDRIVE road network datasets into other common GIS formats such as GeoPackage, Shapefile, GeoJSON, KML or spatial databases. This makes OpenDRIVE easily accessible through common GIS applications. Within the domains of automotive and transportation, highly detailed road network datasets (HD maps) emerged as a core component for development, testing, function validation and also for later production use. Applications such as autonomous driving, driving simulation and traffic simulation often rely on special engineering data formats, of which OpenDRIVE evolved as an open industry standard. This domain-specific data model bundles mathematical, continuous track geometry modelling with all necessary topological links and semantic information from traffic-regulating infrastructure (signs and traffic lights). OpenDRIVE's complexity makes data acquisition a sophisticated task, often financed by the automotive industry and conducted by third-party mobile mapping providers, outside of reach for public cadastral authorities. Recently, public institutions show high interest in such data, especially in the context of urban planning and road infrastructure maintenance. But, tool support for OpenDRIVE is mostly limited to expensive commercial software and - even worse - lacks integration to common Geographic Information Systems (GIS) which are widely used by authorities and researchers. Our free software contribution extends the common Geospatial Data Abstraction Library (GDAL) and transforms OpenDRIVE road elements into OGC Simple Features which can now be loaded and processed ad hoc by all commercial and free GIS tools! By bringing the domains of automotive engineering and GIS closer together, we hope to stimulate interdisciplinary knowledge transfer and the creation of an interconnected research community. With our open software extension, we empower public authorities and research institutions with easier access to highly-detailed road data, which might otherwise be limited to just industrial players. Based on our experience with extending GDAL, other domain-specific data formats such as railML, RoadXML and the NDS Open Lane Model can be made accessible for the greater audience of GIS users as well.