I ride an adult tricycle in Ilhéus, Bahia, Brazil, and started hand-mapping my daily routes in OSM to test an idea: a Walk-Score-style accessibility score for trikes, which have different failure modes than bikes (no curb climbing, grade sensitivity, wind exposure as a directional constraint).
So far: a route=bicycle/network=lcn relation covering my daily routes, tagged with kerb=, incline=, segregated=, surface= where I could determine them. Rendered live via CyclOSM: trikescore.com
This overlaps a lot with wheelchair/mobility accessibility mapping — same tags, similar constraints. Curious if anyone here has done similar work, or has thoughts on where this schema breaks down at scale. No scoring algorithm yet — just validating that the mapping approach holds up.