As a result of OpenStreetMap’s (OSM) openness and wide availability, there is increasing interest in using OSM street network data in routing applications. But due to the heterogeneous nature of Volunteered Geographic Information (VGI) in general and OSM in particular, there is no universally valid answer to questions about the quality of these data sources. In this paper we address the lack of systematic analyses of the quality of the OSM street network for vehicle routing and the effects of using OSM rather than proprietary street networks in vehicle routing applications. We propose a method to evaluate the quality of street networks for vehicle routing purposes which compares relevant street network features as well as computed route lengths and geometries using the Hausdorff distance. The results of our case study comparing OSM and the official Austrian reference graph in the city of Vienna show close agreement of one-way street and turn restriction information. Comparisons of 99,000 route pairs with an average length of 6,812 m show promising results for vehicle routing applications with OSM, especially for route length computation where we found median absolute length differences of 1.0 %.
CITATION STYLE
Graser, A., Straub, M., & Dragaschnig, M. (2015). Is OSM Good Enough for Vehicle Routing? A Study Comparing Street Networks in Vienna. In Lecture Notes in Geoinformation and Cartography (pp. 3–17). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-319-11879-6_1
Mendeley helps you to discover research relevant for your work.