[Tagging] Classifying roads from Trunk to Tertiary and Unclassified

Kevin Kenny kevin.b.kenny at gmail.com
Tue Aug 13 16:14:14 UTC 2019

On Tue, Aug 13, 2019 at 7:35 AM Paul Allen <pla16021 at gmail.com> wrote:
> Hard agree.  Even though it's starting to look like I live in the only country
> in the world with a national classification system that is logical and
> internally consistent (and even we have a few rare exceptional cases). :)

The US certainly doesn't, because of the way we practice federalism.
All fifty states have their own classification systems, and the
classifications have more to do with how funding is doled out than the
actual importance or traffic volume of the road. It is more political
than technical. (Thanks to Martijn van Exel for pointing that out in
last night's Zoom call!)

Back in the days when you would grab a road map at a gas station in
the US, the map-makers used their own hierarchy which was usually
based on the attributes, "toll charged", "grade separated", "number of
lands" and "surface". They typically aggregated these into about half
a dozen different renderings, and had a legend on the map explaining
the symbology used.

I suspect that as routers get more sophisticated, the classification
will become progressively less important, because attributes such as
these will be taken into account more.

Until then, we'll we'll always have some arguments, at every level of
the hierarchy.  Recent arguments: "Is a six-lane dual-carriageway with
a 120 km/h speed limit a motorway if it's hgv-no?" "Is a road built to
full motorway standards actually a motorway if its service is suburban
rather than interurban?" "Can a surface street that parallels a
motorway ever be a trunk, or even primary?" "Can there be tertiary
roads in a state that doesn't have county highways?" "Is there
actually a difference between 'residential' and 'unclassified' in
rural areas?" "Is the last segment of a motorway actually a motorway
when it ends at a grade crossing? (What if the grade crossing is tens
of km from the last elevated crossing?)"

Of such questions are edit wars made, and edit wars make the
classification even less useful.  Out in the boonies, I've encountered
roads where other mappers have argued to me that the classification
ought to be "tertiary", "unclassified", "residential", "service" or
"track" - because different attributes of the (admittedly poor) road
were important to different mappers. (Numbered county highway; the
principal route, at least in summer, between two villages;
non-hard-surfaced; a low-clearance automobile would have a bad day in
inclement weather and a road bike would have a bad time any time;
there are at least a few homesteads along it; and the primary reason
for travelling in that particular area at all is forestry.)

Alas, we can't do what Google Maps does, and aggregate the private
information of everyone carrying a cell phone to measure current
traffic speeds. That appears to be how Google's router makes its

More information about the Tagging mailing list