[Tagging] Mapping cycle tracks as separate geometries, but still know they are tracks?

[Jens Glad Balchen]

On 11.08.2022 16:04, Volker Schmidt wrote:
> Hi Jens,
>
> (I am a bicycle-centered mapper in Italy.)
>
>     I'm currently preparing to map a new cycle track that was recently
>     completed. We're lucky enough that someone has filmed this with a
>     drone: https://www.youtube.com/watch?v=5sevJH7AXxU
>
>
> A very forma question: can we use a youtube video as a source for OSM 
> data? I would not think so.

The luck in this case was that I was able to show it to you easily. I've 
already been there twice to survey it for mapping purposes. :)

>
>     The design manual for cycle tracks
>
> quote?

https://www.vegvesen.no/globalassets/fag/handboker/n100-n300-og-n302_nye-bestemmelser-for-envegsregulert-sykkelveg-med-fortau.pdf

>     states they must be elevated from the carriageway, and optionally
>     with a sidewalk that is even more elevated, and must be ramped
>     down to the carriageway level at every junction and be marked as a
>     cycle lane.
>
> Does this also hold for driveways joining the road?

No, see page 9 of the design manual.

>     Also, the design manual states that it must be placed behind bus
>     stops.
>
> How does the pedestrian (including wheelchair user) get to the 
> bus-stop-platform ?  She needs to cross the cycle track that that is 
> between the footway and the bus stop platform?

Yes, via marked crossing (see page 10).

>     What this video shows is going to be a typical setup in Norway in
>     the time to come, so I thought I'd prepare a mapping guide for OSM-NO.
>
> Good idea.
>
>      Cycle tracks can be mapped with the carriageway on a single line
>
> should be discouraged, in my view, if there is any kind of barrier 
> (kerb, strip of grass, ...)

I agree, but there are vocal differences in opinion on that subject, so 
I'll leave that for another day.

>
>     As a single line, a normal scenario will be this:
>
>     highway=* + cycleway:*=lane + sidewalk=* for the parts around
>     junctions
>
>     highway=* + cycleway:*=track + sidewalk=*  for the middle parts
>     highway=* + cycleway:left/right=track + sidewalk:left/right for
>     the non-bus stop-side andhighway=cycleway + sidewalk=* for the bus
>     stop-side
>
>
> Just to make sure I understand correctly
> This normal scenario is to tag motor-traffic-carriageway + two one-way 
> cycle-only ways + two pedestrian-only ways, all separated from each 
> other by kerbs?

This "normal scenario" is when the mapper has already chosen to map on a 
single line, and I am not questioning that mappers judgement in the 
matter. If the mapper has chosen the single-line approach, the scheme 
outlined here would be the normal way to map it based on my 
understanding of single-line mapping philosophy.

> I presume "Parts around junctions" means the cycle and pedestrian 
> crossings on the side road, as shown at 27" in the video.

I meant those stretches of the roadway that are constructed with cycle 
lanes (due to the requirement to ramp the tracks down from their normal 
elevation).

> This means you do not map the cycle-pedestrian crossing on the side road.

That is true. This is a general drawback of the single-line approach, I 
believe.

> The video shows only a cycle crossing (bicycle lane with dashed 
> lines), but where do the pedestrians cross the side road ???

I believe that particular junction is not yet completed. I noticed the 
same lack of connection.

>     Either scheme is fine, depending on what mappers want to do, so
>     that is NOT what this question is about :)
>
>     The downside of the second approach is we are not "allowed" to tag
>     the highway=cycleway as a track.
>
>
> What is the problem with this?. cycleway=track is only used in the 
> case the single-way approach, which is not applicable in that case.
> With the approach foot=designated / bicycle=designated people also use 
> highway=track, highway=cycleway instead of highway=path, if they are 
> happier that way.

The problem:
In the multi-line approach, we would tag the cycle track with 
highway=cycleway, because that is what it is legally signed as 
sf-20051007-1219-520-01.gif and classified as. But this is also how we 
tag a cycleway that is several meters away from the carriageway, or a 
cycleway that is completely on its own with no adjacent carriageway.

The cycle track is a much more dangerous place for children compared to 
a cycleway separated by several meters. As an example, a schoolgirl 
riding to school on her bike on a similar track collided with a friend 
and then fell onto the carriageway, where she was struck by a driver 
(who apparently did not see the need to slow down when passing a group 
of children). After this collision, the road authority built a concrete 
wall between the two: https://goo.gl/maps/adqyGwiqh6EwX9dc9

If there are several meters of separation, falling would not have those 
consequences.

So anyone looking at the OSM data would not ble able to know if a 
particular highway=cycleway is the safe one suitable for kids, or the 
not-so-safe one suitable for adults.

A cycle track and a cycle path would probably never appear side-by-side 
in a real scenario, so the question is knowing if the cycleway that is 
there, is one or the other.

>     The reason it would be valuable to tag highway=cycleway as a track
>     in this instance is that a track has requirements for separation
>     from other mode types that are more similar to a cycle lane than
>     to an actual bike path (at least in Norway). So you'll be
>     significantly safer on a bike path than you will be on a cycle
>     track, and marginally safer on a cycle track than in a cycle lane.
>     For that reason, it'd be nice to know if what we have is a track
>     or a path.
>
> Could you give pointers to photos of real situations, please,  for the 
> Norwegian cycle track and cycle paths you are referring to.

Here is an example of a road with a two-sided bike lane, and also a 
two-sided bike path:

https://kart.finn.no/?lng=5.73957&lat=58.89643&zoom=20&mapType=historicalm-Stavanger-2020%40c

If these bike lanes were instead built as a tracks, as in the drone 
video, the added safety is only a low kerb (10 cm), as opposed to three 
meters separation for the bike path. Bike paths would still be 
significantly safer, especially for kids. But both tracks and paths 
would be highway=cycleway.

>     How could we tag separately drawn cycleway tracks so that we know
>     they are tracks?
>
> This seems to boil down on how to map the Norwegian situation on the 
> existing tagging schemes in OSM, or even a translation problem. Should 
> be solvable, as keys and values are after all only arbitrary strings 
> in OSM. Any meaning of the strings is to help the mapper, but is not 
> essential to the way the data works.

One way of knowing implicitly is to tag the cycle track as 
highway=cycleway + oneway=yes (oneway=yes is a legal requirement any, 
due to sf-20051007-1219-526-1-01.gif). No separated bike path would have 
that combination. It's very implicit, though.

Cheers,

Jens
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20220811/d157e38e/attachment-0001.htm>