I am cleaning up more roads I use on OSM with better information, and there's a particular stretch of highway that is used by all routing systems out there because it's the most direct route. However, it's a very high-volume road during cycling season and the shoulder all but disappears for long stretches. This wouldn't seem to matter but most of the traffic there is traveling at high speed AND towing large trailers (boats or campers). It's so bad I refused to cycle it on a tour, and to give you an idea of my cycling risk tolerance, I used to commute in and out of the South side of Chicago, Illinois, entirely on surface streets!
The problem here is mainly that the shoulder disappears. There may be alternate, lower-traffic, (longer) and more scenic options.
I'm wondering how c.t guesses traffic volume (if at all) and weighs that against shoulder tags (not that common, so maybe never?) so that I can sort of maximize my OSM editing for the best benefit of cyclists using these kinds of tools.
Of course the best solution would be the state just paves the rest of this highway shoulder! (Odd they haven't. They have wonderful, 3m wide shoulders on much lower volume roads near us. Those are cycling paradise.)
Comments
cycle.travel does take account of the shoulder= tag but not currently the shoulder:width tag. Maybe it should! I’ll look into what the current state of the data is. For traffic volume it uses government open data where available, and fortunately it’s quite decent in the US.
I’d be interested to know what road it is so I could take a closer look.