[Tagging] Reliable OSM routing

Richard Welty rwelty at averillpark.net
Sun Feb 9 20:30:07 UTC 2014


On 2/9/14 3:11 PM, Tod Fitch wrote:
>
> Some very good points here. . .
>
>
> On Feb 9, 2014, at 10:47 AM, André Pirard wrote:
>
>> Hi,
>>
>> I left this unsent for some time, so it may duplicate what has already been said.
>> But it's uttermost important.
>> <snip>
>>
>> There doesn't seem to be routing quality assurance tools, not even Osmose detecting detectable mistakes, and I've corrected quite a number of bad tagging mistakes that I found with Osmand.
> Osmand is definitely a quality assurance tool for me. I have taken to using its routing facility at every opportunity. When it comes up with a route that makes no sense to me I go back to look at the OSM tagging to look for clues as to why the route was selected. Usually it is bad or missing tagging. In the western U.S. there is still a lot of old Tiger imported data that has not been corrected or added to. Sometimes it is incorrect topology, but often it is simply the lack of speed information as well as the location of stop signs and traffic signals that cause the routing issues.
i agree 100%. i use Osmand a lot, and regularly check to see
how it is routing in places that i know well. it has on more than
one occasion alerted me to errors in the map, like a mapper
inadvertently reversing the sense of oneway on a dual carriageway.

> Since I'm not willing to setup for generating my own OBF files it takes a couple of weeks for me to see if the fixes I add make any difference.
i will at some point need to learn how to generate files for Osmand.
another thing on a very long list of things i need to learn to do.
>> Many people appear confused with the restriction rules.
>> There should be a wiki page listing exactly what tags mappers and routers must use.
> That would be very helpful.
i think a tutorial "how to map for routers" alongside a reference page would
make a good pairing.
>> AFAICS, OSM routing is not reliable. It does make useful routes on the large scale but it makes many errors when coming down to the details and one may be booked for trusting it too much.
>>
> In the business sector I was in until recently, the phrase "eating your own dog food" seems to apply. That is, use your own products in your production environment to be really sure they are ready for market. I strongly suggest that anyone doing tagging on highways use the OSM based routing tools and when the results don't match expectations find and fix the tagging problem.
>
routing in OSM is at a very early stage. best way to move it forward is
for mappers to start using it in order to smoke out the problems.
i'm very glad that i've tried to maintain some sort of OSM based
routing capability on my dashboard for the past couple of years,
initially using garmins & mkgmap output, and now with Osmand.

richard

-- 
rwelty at averillpark.net
 Averill Park Networking - GIS & IT Consulting
 OpenStreetMap - PostgreSQL - Linux
 Java - Web Applications - Search


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 946 bytes
Desc: OpenPGP digital signature
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20140209/b949e647/attachment.pgp>


More information about the Tagging mailing list