[Talk-GB] Tagging UK Parallel crossings - Zebra for cycles

[James Derrick]

On 21/08/2022 14:06, Jacob Nevins wrote:
> [re-sent after subscribing to mailing list -- sorry if you see 2 copies]
> I fished out some docs about your scheme (assuming this matches what was
> actually built):
> <https://my.northtyneside.gov.uk/sites/default/files/web-page-galleries/Rake%20Ln%20-%20Billy%20Mill%20Ln%20rounadbout.jpg>
> <https://my.northtyneside.gov.uk/category/1564/rake-lane-billy-mill-lane-roundabout>

Crikey - £1.6m and the end of the cycleway still punts you straight into 
a sign post! :-)

https://www.mapillary.com/app/?pKey=5317417538312596

Thanks for the link - the diagram is as-built AFAICT, and emphasises the 
separation between footway and one-way cycleway.


> I think the closest to that in Cambridge (i.e. with unsignalled
> ped-and-cycle-priority crossings) is the Fendon Road roundabout:
> <https://www.openstreetmap.org/node/7773880586#map=20/52.17931/0.14892&layers=Y>
> <https://www.cyclestreets.net/location/148655/>
> <https://www.cambridgeshire.gov.uk/residents/travel-roads-and-parking/transport-projects/cycling-pedestrian-improvements/fendon-road-and-queen-edith-s-way-roundabout>
> <https://therantyhighwayman.blogspot.com/2021/06/fabulous-fendon-road.html>

Useful examples - the coloured cycleway tarmac should be mandatory (red 
in Oxford, Green in Manchester) to emphasise the crossing. A bus driver 
friend reports a car driver got confused and stuck after turning down 
the narrow cycleway!

`crossing=uncontrolled` + `bicycle=yes`


> ...but whoever tagged it chose crossing=uncontrolled for the cycle
> crossing, which fails to fully capture the cycle-priority nature of the
> crossing. So now I have the same question about how to tag an
> unsignalled cycle-only crossing where cyclists have priority :)

My attempt was based on the JOSM preset for a Zebra- 
`crossing=uncontrolled` which I agree doesn't quite capture the MUST 
nature of either Zebra or Parallel. Might be down to some countries 
using stripes without priority.


> For these sorts of junction systems, it does seem usual to tag the foot
> and cycle crossings separately, e.g. these CYCLOPS (signalled)
> junctions:
> <https://www.openstreetmap.org/#map=19/53.46765/-2.25758>  (not me)
> <https://www.openstreetmap.org/#map=19/52.22185/0.11204>  (me)
> I think that's right, for the reasons you state. So if we are to show
> priority for cycle-only crossings, we do need a way to tag it.

Another possible reason for separation is the difference in "inbound" 
and "outbound" highways - at Rake Lane drivers pass Zebra then Parallel 
towards the junction, but the reverse leaving.  Sadly the attention of 
drivers seems towards the roundabout and some seem not to perceive the 
Zebra never mind the less familiar Parallel on the way out.

It's also interesting to see the difference in design - some place 
cycleway towards the centre (Rale Lane) others footway (Cambridge).


> crossing_ref=zebra does feel like a bit of a bodge for a cycle-only
> crossing in the UK, depending whether you take 'zebra' to denote the
> official definition in the Highway Code etc (which is explicitly about
> pedestrians), or the physical appearance. Might still be the best option
> currently?

It was a tad surprising to see the familiar "Belisha Beacons" at the 
parallel crossing, further suggesting the Zebra heritage.

Jon spotted `crossing_ref=tiger 
<https://wiki.openstreetmap.org/w/index.php?title=Tag:crossing_ref%3Dtiger&action=edit&redlink=1>` 
which I'd not heard of but looks good apart from the 'parallel' language 
used in the HWC.


> Looks like someone did try to come up with a scheme for explicitly
> tagging who has priority at a crossing, but it turned out to be
> complicated, and was abandoned:
> <https://lists.openstreetmap.org/pipermail/tagging/2020-December/057071.html>
> (see also followups)

`crossing:priority=yes` has some merit, but perhaps 
`crossing:priority=cycleway` removes all ambiguity without having to 
interpret HWC Rule H1 heirarchy of road users.


> Here for completeness is an isolated parallel crossing in Cambridge.
> <https://www.openstreetmap.org/node/9514771833/history#map=20/52.21754/0.10385&layers=Y>
> <https://www.cyclestreets.net/location/114080/>
> Apparently I chose a single node (which seems OK here) with
> crossing_ref=parallel; as others have noted, crossing_ref=tiger also
> exists, and is more popular in taginfo, both inside and outside the UK:
> <https://taginfo.openstreetmap.org/keys/crossing_ref#values>
> (although personally I still think 'parallel' makes more sense, both
> inside and outside the UK).

I'd agree about one node here - especially as the "main" cycleway also 
has to give way to the "crossing" cycleway.

For a combined node, sounds like `crossing_ref=tiger 
<https://wiki.openstreetmap.org/w/index.php?title=Tag:crossing_ref%3Dtiger&action=edit&redlink=1>` 
or `crossing_ref=parallel` which is more about UK-specific HWC language 
than substance.

For separated nodes, `crossing_ref=zebra` + `crossing_ref=parallel` perhaps?


> I've no idea what if anything routing engines such as CycleStreets do
> with all this. (I suspect this page is not fully up to date:
> <https://www.cyclestreets.net/help/journey/osmconversion/>)

Must agree -  "CycleStreets does not yet [Jan 2011] make use of this 
information." !


Thanks again for interesting detail,


James
-- 
James Derrick
     lists at jamesderrick.org, Cramlington, England
     I wouldn't be a volunteer if you paid me...
     https://www.openstreetmap.org/user/James%20Derrick
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-gb/attachments/20220821/095acb4e/attachment-0001.htm>