[Tagging] Feature Proposal - RFC - Pedestrian lane

Tue Nov 12 22:53:26 UTC 2019

You make a very good point! A road can have a pedestrian lane, shoulder,
both, or neither, so it wouldn't make any sense for a pedestrian lane to be
a type of shoulder. The widths do vary quite a bit as well, regionally.

> You mean a situation like this?:

> https://wiki.openstreetmap.org/wiki/File:Sidewalk_and_crossing.svg

One very similar to that, yes! I think I normally wouldn't add
sidewalk=both to any length of the highway=residential. Is that a typical
thing to do? I would assume that meant the highway=residential street had
its own short piece of sidewalk, when it actually doesn't.

The challenge I'm describing is in reliably associating the crosswalk with
the pedestrian paths. After all, the crosswalk is a node on a different
street way. I know that I could do it 99.x% of the time, but it will
require using some graph traversal approaches that most people aren't
familiar with. Plus, those cases where I couldn't reliably determine it
could be very important. I suspect this is one of the reasons I haven't
found anyone using these data in concert (sidewalk=both + highway=crossing)
to do pedestrian routing.

Mapping for a router isn't the end-all-be-all of this kind of data, of
course, but it is one thing that would be hard to do with this tagging
schema. I'd be interested to know if there are other data consumer plans
for the data, since use does dictate what the schema looks like. Making
streets be ways was a conscious choice informed by routing, for example!

> I would simply connect the sidewalk way with the road where the
sidewalk ends (and map a barrier=kerb + kerb=* node) and add
pedestrian_lane=* to the road starting from where the pedestrian lane
begins.

So there would be a segment of footway=sidewalk that is not actually on a
sidewalk? I've been unsure about what to do in similar situations, like how
to connect footways to roads without implying there's literally a footway
on top of the road. Probably worth its own, separate discussion (it was
discussed previously, but without conclusion), so I won't elaborate.

> There is a pedestrian lane along the the south-eastern part of the road
Reichenbachstrasse. On the opposite side there are public steps as well as
many (currently unmapped) driveways and private footpaths. Mapping the
pedestrian lane as a separate way would either make it disconnected from
the steps, driveways and footpaths on the opposite side of the road or you
would need to add many highway=footway connections from the pedestrian lane
to the steps, driveways and footpaths, which would make the map very
confusing.

> Therefore i strongly advise against mapping pedestrian lanes as separate
ways.

> By the way, the same problem occurs with sidewalks mapped as separate
ways.

Yes, it's a trade-off: the actual pedestrian path's primary connections and
attributes vs. its association with the street. Neither are actually
perfect options, which is why I'm suggesting the possibility of redundant
tagging. Ideally, we'd come up with a universal strategy for relating these
ways together, but I don't want to monopolize this proposal!

> I'm not a programmer and therefore don't have concrete plans to use this
data, but i imagine (and hope) that pedestrian routers could use this data
to prioritise roads with pedestrian lanes and to tell blind people on which
side of the road they should walk.

Maybe it would be helpful to set up a meeting with some organizations that
serve the visually impaired along with programmers that build routing
software. We (Taskar Center) might be able to help with that sort of
meeting, and it would be even better to have organizations from different
cultures and geographies involved as well. As-is, I think the challenge of
reliably associating paths with crosswalks is a big one for mapping for
routing for the blind.

> Thanks. The content of this page seems to be identical to this PDF
document by the FHWA i mentioned in some of my earlier messages:

Oh, you're right! Sorry I missed that!

Thanks again for raising this tagging question and proposal!

Best,

Nick

On Mon, Nov 11, 2019 at 2:02 PM Markus <selfishseahorse at gmail.com> wrote:

> Hi Nick,
>
> Please excuse my late reply. :(
>
> On Wed, 6 Nov 2019 at 00:53, Nick Bolten <nbolten at gmail.com> wrote:
> >
> > ## Similarities to shoulders and an opportunity to figure out how to tag
> them.
> >
> > Would it be fair to say that the only differences between this feature
> and a shoulder are (A) it has paint designating where pedestrians should go
> and (B) it has some right-of-way implications? Because it's often the only
> pedestrian option in rural areas near me, I'd appreciate having a way to
> tag shoulders and then enhancing them with a subtag. e.g., something like
> shoulder=left/right/both + shoulder:right=pedestrian_lane.
>
> Another difference is the width: in Switzerland, pedestrian lanes are
> about 1.5 m wide and shoulders about 4.5 m. But in my opinion their
> different purpose is reason enough to use different tags. Besides,
> cycle lanes already have a separate tag. (Even though cyclists are
> also allowed to use shoulders in the USA afaik.) And finally,
> shoulder:right=pedestrian_lane doesn't make much sense semantically.
>
> > ## Challenges of mapping pedestrian paths as street attributes
> >
> > As proposed, this tag would apply to streets. I understand the appeal -
> it's a minimal change from current maps and the feature is basically just
> paint on a street - but I think there are also some potential risks to
> describing the pedestrian path this way that would be valuable to discuss.
> Examples:
> >
> > (1) Intersections, particularly ones with marked crossings.
> sidewalk=left/right/no/both has difficulties with this as well. Put
> yourself in the shoes of someone trying to analyze the paths a pedestrian
> could take using this tag to determine that there is a path using
> pedestrian lanes and a crosswalk. There is a street way (way 1) with
> pedestrian_lane=right that continues through an intersection. There is a
> crosswalk tagged as highway=crossing, crossing=uncontrolled on another way
> that shares a node with another street way (way 2). How do you proceed and
> associate these path data so that you can reliably say that a pedestrian
> path exists that uses that crosswalk? I believe it will require some fairly
> nerdy graph analysis I think it could be a significant hurdle for using
> this data.
>
> You mean a situation like this?:
>
> https://wiki.openstreetmap.org/wiki/File:Sidewalk_and_crossing.svg
>
> I add a sidewalk=both tag to the road section up to the crosswalk,
> then sidewalk=no to the rest of the road that doesn't have a sidewalk.
> This may look a bit strange in this example, but usually the sidewalks
> are more curved at crossroads, like for example here:
>
> https://www.openstreetmap.org/way/744453045
>
> (The "Stadt Bern 10cm/25cm (2012)" imagery has the highest resolution
> at this place.)
>
> I suggest the same mapping for pedestrian lanes.
>
> > (2) Transitions to other pedestrian paths, such as sidewalks. Pedestrian
> lanes are sometimes used as a means to have a "temporary" sidewalk-like
> feature, pending some future construction of actual sidewalks. There will
> be sidewalks that are half-built, then transition into a pedestrian lane.
> How do we tag that situation, given a separately-mapped sidewalk?
>
> I would simply connect the sidewalk way with the road where the
> sidewalk ends (and map a barrier=kerb + kerb=* node) and add
> pedestrian_lane=* to the road starting from where the pedestrian lane
> begins.
>
> > With the above issues in mind, what would you think about allowing
> highway=footway, footway=pedestrian_lane as a possibly redundant tagging
> option?
>
> Consider this example:
>
> https://www.openstreetmap.org/#map=19/46.99149/7.45448
>
> There is a pedestrian lane along the the south-eastern part of the
> road Reichenbachstrasse. On the opposite side there are public steps
> as well as many (currently unmapped) driveways and private footpaths.
> Mapping the pedestrian lane as a separate way would either make it
> disconnected from the steps, driveways and footpaths on the opposite
> side of the road or you would need to add many highway=footway
> connections from the pedestrian lane to the steps, driveways and
> footpaths, which would make the map very confusing.
>
> Therefore i strongly advise against mapping pedestrian lanes as separate
> ways.
>
> By the way, the same problem occurs with sidewalks mapped as separate ways.
>
> > ## Usefulness / data consumption
> >
> > Knowing where pedestrian lanes are would be very useful, in my opinion,
> but the devil is always in the details. Do you have any examples of how
> this data could be consumed downstream? Not saying there always has to be a
> data consumer, but the exercise could reveal advantages between different
> approaches.
>
> I'm not a programmer and therefore don't have concrete plans to use
> this data, but i imagine (and hope) that pedestrian routers could use
> this data to prioritise roads with pedestrian lanes and to tell blind
> people on which side of the road they should walk.
>
> > ## Other sources
> >
> > A potentially helpful resource during these international comparisons:
> https://www.fhwa.dot.gov/environment/bicycle_pedestrian/publications/small_towns/page05.cfm.
> The FHWA defines standards in the United States.
>
> Thanks. The content of this page seems to be identical to this PDF
> document by the FHWA i mentioned in some of my earlier messages:
>
>
> https://www.fhwa.dot.gov/environment/bicycle_pedestrian/publications/small_towns/fhwahep17024_lg.pdf
>
> Best regards
>
> Markus
>
> _______________________________________________
> Tagging mailing list
> Tagging at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20191112/6fc014f6/attachment-0001.html>