[Talk-GB] TfL CID conflation script

[Martin - CycleStreets]

A few years ago, CycleStreets did a piece of work to advise TfL on how they 
could get their Cycling Infrastructure Database dataset into OSM. This 
resulted in a project documented at:
https://wiki.openstreetmap.org/wiki/TfL_Cycling_Infrastructure_Database

This resulted in a mapping of attributes, and a data conversion script 
which Richard Fairhurst worked on for us:
https://github.com/cyclestreets/tflcid-conversion/
(The script basically does as much automated attribute translation as can 
be done by machine, but the actual conflation then has to be done by a 
human.)

Some initial data was conflated in directly by TfL as a trial. More 
recently, Sweco and GHD have been commissioned by TfL to continue that. 
However, our attention has been drawn to some quality problems in the start 
of that continuing conflation work, raised last month on this list and I 
see on the imports list. Like everyone else here, we are keen to ensure 
that any kind of conflation work is high quality.

We have reviewed some of the changesets and given Sweco some advice on 
ensuring good quality conflation and changesets more generally, and I see 
they have posted an update also.
https://github.com/cyclestreets/tflcid-conversion/issues/40#issuecomment-1202820574
https://lists.openstreetmap.org/pipermail/talk-gb/2022-July/029150.html

In addition:

The mapping of attributes from TfL CID to OSM in the script had a few 
unresolved questions:
https://github.com/cyclestreets/tflcid-conversion/issues

I have now worked through those. Comments are of course welcome. Richard F 
will be implementing these in the data conversion script shortly.

There are a few issues which do not have a clear tag available in OSM to 
map into, and I'll raise those separately on this list for comments.


Martin,                     **  CycleStreets - For Cyclists, By Cyclists
Developer, CycleStreets     **  https://www.cyclestreets.org/