An OSM mapper has added start_date=* tags to roads throughout Kentucky by interpolating YEAR_BUILT_027 attributes from the National Bridge Inventory, a public domain dataset of road bridge inspection records throughout the United States. Perhaps we could import NBI as a starting point for more comprehensive road and river coverage. At a high level, the process could look something like this:
- Import NBI as
man_made=bridgepoints, each with astart_date=*(YEAR_BUILT_027) and temporary tags indicating the name of the road passing over the bridge (FACILITY_CARRIED_007) and the name of the feature passing under it (FEATURES_DESC_006A). - Set up a MapRoulette challenge or similar for replacing these
man_made=bridgepoints withman_made=bridgeareas based on aerial imagery. - Set up a Tasking Manager project per state for linking these bridges with roads based on the temporary name tags.
- Set up a Tasking Manager project per state for linking the features under the roads (such as streams and railroads) based on the temporary name tags.
The latter steps will be more challenging since a road can predate the bridges along it (for example due to reconstruction or a former ford). We could mitigate those issues by focusing on major highways at first.
Along these lines, I’ve been using the Caltrans bridge inventory to map road bridges and highways in the San Francisco Bay Area, adding the bridges to both OSM and OHM in tandem. Not every state has this kind of data easily accessible in the public domain, so NBI can be quite useful.