Crowdsourcing turning restrictions for OpenStreetMap

The abundance of GPS tracking data due to the emergence and popularity of smartphones has fuelled significant research around GPS trajectories and map-matching algorithms. Unfortunately, none of this previous research addresses the issue of identifying turning restrictions in the underlying road network graph. Our latest research endeavour remedies this, by proposing a novel, straightforward and e↵ective way to infer turning restrictions for OpenStreetMap data, by utilizing historic map-matching results from an existing fleet management service. Our experimental evaluation based on the results acquired for three European cities within an one-year period, proves the robustness and credibility of our method.