Enhancing Transportation Insights: A Scalable OpenStreetMap Harmonization and Synthesis Framework

Transportation infrastructure is crucial in determining accessibility and mobility metrics, which help us to optimize transportation systems. However, current representations of the local environment are less standardized, making it hard to extract relevant data. To address this, we propose a framework that enables geographic and categorical homogenization of environmental data and recombination with additional sociodemographic data. Ensuring semantic consistency, our framework leverages the styling rules of OSM Carto representation to extract land use and transportation infrastructure data worldwide. We evaluated the framework's performance in Munich, Germany, by refining WorldPops' distribution via OSM building data from our framework to synthetically derive a more accurate population distribution for Munich's districts. Our benchmark against official administrative ground truth shows a significant improvement compared to the baseline distributions.