I am working on a research project at University of Bremen, Germany, and we need accurate road width information for analysis. While OSM provides excellent road geometry and length data, width tags are often missing.
I have also noticed that sometimes we assume default width values (for example, setting a standard width of 5–6 m for residential roads without field measurements). Could you please clarify:
How reliable are these default width values?
Is there a recommended approach to distinguish between measured widths and estimated/default widths in OSM data?
Whether there is any existing dataset or extract that includes road width information.
Are there recommended tools or approaches to estimate width from existing attributes (e.g., number of lanes or highway type)?
We highly appreciate the efforts of the OSM community and would be glad to contribute back if possible.
Thank you very much for your time and support.
Kind regards,
Adita Sen
Student Assistant
University of Bremen, Germany
In the UK I don’t think that most highway=residential roads are 5 or 6m wide, but that might be different in your area of interest. It will obviously vary hugely around the world.
No, there’s only the value “width”, not how the mapper got that value.
Only yesterday I added a newly constructed road segment, width of the road measured by a smartphone app, width of the new cycleway measured by my shoes. Second one is a reliable value
Even with the same highway type there may be great differences in width:
Examples of highway=scondary in the same federal state of Germany
L 102 at Riesweiler
I did move the second one from tertiary to secondary some years ago but I hadn’t the courage to downgrade L 102 from secondary to tertiary.