[Talk-GB] Conflation of TfL data into OSM (was: Is TfL data allowed on OSM?)

Andy Townsend ajt1047 at gmail.com
Sun Jul 3 10:25:00 UTC 2022


Thanks Rob.

As a bit of context, on 30/6/2022 Jez wrote:

> OSMUK were approached by TfL at the beginning where we put forward 
> cycling OSMers like CycleStreets and cycle.travel 
> <http://cycle.travel>. CycleStreets did the initial consulting, with a 
> fair amount of community discussion and lots of effort, and have now 
> handed over the baton to Sweco and GHD.  I do hope that we can work 
> with them rather than against.
I then asked "links" and Rob replied with the email below.

https://wiki.openstreetmap.org/wiki/TfL_Cycling_Infrastructure_Database#Work_to_Date 
and 
https://wiki.openstreetmap.org/wiki/TfL_Cycling_Infrastructure_Database#Current_Work 
seem to be a good summary of what's happened so far and answer the "were 
approached by" and "cyclestreets did consulting" parts of the question 
that I was asking for links for.

The third part was asking for information about "a fair amount of 
community discussion and lots of effort".  I don't doubt that 
https://wiki.openstreetmap.org/wiki/TfL_Cycling_Infrastructure_Database#Data_Merge_Workflow 
involved "lots of effort" but I don't see any evidence of "community 
discussion" at all.

Worryingly I also don't see anything in 
https://wiki.openstreetmap.org/wiki/TfL_Cycling_Infrastructure_Database 
that addresses what happens when TfL's data suggests one thing and OSM's 
suggests another. In some cases TfL data will be correct and in some 
cases OSM will -  I'd expect that an on-the-ground survey (or access to 
photos from one) would be needed to resolve.

https://wiki.openstreetmap.org/wiki/TfL_Cycling_Infrastructure_Database#Process_Workflow 
also contains "we will regularly check for comments [on OSM 
changesets]".  That simply didn't happen effectively, as can be seen at 
https://resultmaps.neis-one.org/osm-discussion-comments?uid=15976978 .  
Many of the issues there appear seem to be because  TfL's data is out of 
date.  I'd have expected the conflation process to explicitly say what 
should happen in this case.

Had https://wiki.osmfoundation.org/wiki/Organised_Editing_Guidelines 
been followed, an entry for Sweco being present at 
https://wiki.openstreetmap.org/wiki/Organised_Editing/Activities (but 
note *1), https://www.openstreetmap.org/user/AyushS183 and any other 
editors had a profile that pointed to 
https://wiki.openstreetmap.org/wiki/Organised_Editing/Activities/TfL_Cycling_Infrastructure_Database 
and been suitably trained (both also as required by 
https://wiki.osmfoundation.org/wiki/Organised_Editing_Guidelines ) none 
of this would have been an issue and the DWG wouldn't have had 
complaints about the quality of the editing and the other aspects of the 
project.

While we're waiting for the people running this project to reply here, 
it would be extremely useful to know *whether the data that has been 
added for this project so far is of a good enough quality* to add to 
OSM. So far we're only aware of edits by 
https://www.openstreetmap.org/user/AyushS183/history .  The lack of 
documentation means that we can't assume that there aren't other 
contributors, but the ones that we do know about have a changeset source 
of "tfl cycle database".

There are 554 changesets that we're aware of so far and the overpass 
queries at *2 will help identify local changes,

Best Regards,

Andy Townsend, on behalf of OSM's Data Working Group


*1 confusingly something for TfL with a start date of 2020-06 and an end 
date of 2020-07 does exist there, but it does not mention Sweco at all


*2 All features including relations:

http://overpass-turbo.eu/?q=LyoKVGhpcyBoYcSGYmVlbiBnxI1lcmF0ZWQgYnkgdGhlIG92xJJwxIlzLXR1cmJvIHdpemFyZC7EgsSdxJ9yaWdpbmFsIHNlxLBjaMSsxIk6CsOiwoDCnHNob3AgxLogbG91Z2hicsWTxZXFiMKdCiovClvFk3Q6anNvbl1bdGltZcWiOjI1XTsKLy8gZmV0xYIgxLDFgCDFiMKcxZLFlMWWxZjGhMWbxJvEq8S_xYHFg8S6Cnt7xJBvY29kZUFyxYA6xoPFlcSqxobFlX19LT4uxotyxYLGmMWAxbTFtsSPxJTEnXIgxplzdWx0cwooCiAgxqxxdcSSxJrEo3J0xbhvcjrGgMWJxYvFjcWbxrogbndyKMaQYsSqeMahKSh1xL_HhyJBecebaFMxODMiKcW0xrvFtndheVsix4twIl0oxb5hxqXFgManaMapYceox47FtsaZbMSUacWnx6_Hscezx7XGmce3xqbGqMiKx77HvsascMS3bseExrLGtMa2CsWixJjGlXnFtD7FtMiac2tlxL1xdDs&c=BJnwm9b5HT

nodes and ways:

http://overpass-turbo.eu/?q=LyoKVGhpcyBoYcSGYmVlbiBnxI1lcmF0ZWQgYnkgdGhlIG92xJJwxIlzLXR1cmJvIHdpemFyZC7EgsSdxJ9yaWdpbmFsIHNlxLBjaMSsxIk6CsOiwoDCnHNob3AgxLogbG91Z2hicsWTxZXFiMKdCiovClvFk3Q6anNvbl1bdGltZcWiOjI1XTsKLy8gZmV0xYIgxLDFgCDFiMKcxZLFlMWWxZjGhMWbxJvEq8S_xYHFg8S6Cnt7xJBvY29kZUFyxYA6xoPFlcSqxobFlX19LT4uxotyxYLGmMWAxbTFtsSPxJTEnXIgxplzdWx0cwooCiAgxqxxdcSSxJrEo3J0xbhvcjrGgMWJxYvFjcWbxrogbsaVZSjGkGLEqnjGoSkodcS_x4ciQXnHm2hTMTgzIinFtMa7d2F5x5N7x5Vvx5d9x5nHm8SSOseex6DFi8ejx6XHp8epxrwvxplsxJRpxadbIseLcCJdKMW-YcalxYDGp2jGqWHHqArIlcascMS3bseExrLGtMa2CsWixJjGlXnFtD7FtMihc2tlxL1xdDs&c=BJnwcOyk9R

The "nodes and ways" query is likely to be the more useful of the two I 
expect.


On 02/07/2022 19:47, Rob Nickerson wrote:
> Hi Andy,
>
> You asked for links:
>
> Overview:
> https://wiki.openstreetmap.org/wiki/TfL_Cycling_Infrastructure_Database
>
> Licence details:
> https://lists.openstreetmap.org/pipermail/talk-gb/2019-August/023356.html
>
> OSM UK's involvement:
> "TfL Cycle Infrastructure. They accepted our proposal to ask the 
> Talent Directory to help with them, this has been sent."
> https://osmuk.org/minutes/board-meeting-2019-03-13/
>
> Other references on OSM UK website including meeting slides:
> https://www.google.com/search?q=tfl%20site%3Aosmuk.org
>
>
> Should you want any more links, please detail what your are seeking 
> and I'll dig out an appropriate link.
>
> Thanks
> Rob
>
> _______________________________________________
> Talk-GB mailing list
> Talk-GB at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-gb
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-gb/attachments/20220703/e31f4e6c/attachment-0001.htm>


More information about the Talk-GB mailing list