[Tagging] Feature Proposal - RFC - Hazards
Brian M. Sperlongano
zelonewolf at gmail.com
Sat Dec 5 19:03:09 UTC 2020
I want to address the points that were raised on crossings.
As we already have highway=crossing, I have resisted adding new hazard=*
values for crossing hazards, as that is properly the domain of the
highway=crossing tag. For golf cart crossing, there is already an
established tag combination highway=crossing + golf_cart=yes. This was
already described on the wiki page for golf_cart, and I edited the wiki
page for highway=crossing to match this.
Considering this, I've updated the text on hazard=bicycle to reflect that
it is specifically about the hazard of cyclists in the roadway, and not for
bicycle crossings, which also has its own existing tagging.
This leaves hazard=school_crossing as the sole remaining "crossing"
hazard. There are only a few usages of it, and I am leaning strongly
towards removing from the proposal entirely as it seems that a school
crossing belongs more properly within the scope of highway=crossing.
Lastly, the specific case of hazard=low_flying_aircraft is not a crossing
hazard, as the hazard is a distraction to motorists or perhaps jetwash
rather than collision, as with a crossing. The only actual "airplane
crossing" that I am aware of is the case of the Gibraltar airport which has
a road that crosses the airport runway. However, this is an exceptionally
rare case, and in any case, it too would belong under highway=crossing.
On Sat, Dec 5, 2020 at 12:43 PM Volker Schmidt <voschix at gmail.com> wrote:
> Hi,
> I have been following this proposal with interest. I often have tried to
> tag hazards, and not found a good ways of doing it.
> We are now compiling a long list of hazards, including golf players
> crossing the road, but I see some basic aspects which are not being
> addressed (unless I missed something):
>
> I would like to see signposted hazards completely separately tagged from
> hazards that the mapper perceives in a place, but which are not signed.
>
> Signed hazards should be mapped.
>
> - on nodes, if the extension of the hazard is point-like (example:
> dangerous railway crossing)
> - on ways, if the hazard exists along a highway (example: animals
> crossing zones)
> - (possibly) on areas, if the hazard is present in an area (example:
> landslides)
>
> In the case of signed hazards, I see two alternative ways of tagging the
> signing:
>
> - (only for nodes and ways highway segments) by adding source:xxx=sign
> like we do with speed limits
> - by mapping the relative signs as nodes
>
> Insertion of signposted hazards do not require any assessment of the
> presence of the hazard by the mapper.
>
> Signposted hazards are most often signalling dangers for vehicle drivers.
> Let's take the sign for hazard=cyclists (crossing), which warns clearly the
> vehicle drivers on the carriageway, that there could be cyclists crossing.
> There is normally no such warning on the crossing cyclists' path.
> There are exceptions of hazard warnings for both parties like a "cyclists
> sharing the road" sign, but that's the only one that comes to mind.
>
> Another aspect that should be defined: Are writings or pictograms on the
> road surface equivalent to vertical traffic signs?
>
>
> A completely different story are unsigned hazards with no signs on the
> ground, i.e. situations perceived as a hazard by the mapper.
> These are the tricky ones. I map cycling infrastructure, hence my examples
> come from that perspective.
> Examples:
>
> - foot-cycle crosswalks where there is a sign-posted speed limit of
> 30km/h, but where 90% of the cars pass with speeds far exceeding that value
> and making the place really dangerous
> - a two-way cycle path that is parallel to a main road and crosses a
> side road with a foot.bicycle crosswalk - car drivers entering the side
> road regularly overlook cyclists which ride in the same direction as they
> drive (to my knowledge the major cause of cyclists being killed in many
> countries. These in most cases in my part of the world have no danger
> signs.
> - And now consider the same situation with a row of trees between the
> cycle path and the main carriage way.
> - In my part of the world authorities put all kinds of bollards,
> arches, chicanes on cycleways (supposedly for the safety of cyclists, but
> in reality to keep car drivers from parking there). Many of these are grey
> metal objects that become invisible at night even if you have a good cycle
> light, as they have no reflective markers on them.
>
> The problem here is that the tagging will be based on my perceived version
> of ground truth. If I am a cyclist, I may be good at spotting hazards for
> cyclists. If I am a horse rider I will be good at mapping hazards for horse
> riders.
>
> Then we have also the asymmetric situations: e.g. car drivers are warned
> by a sign that there will be cyclists crossing, but the (bigger) hazard of
> cars hitting the cyclists on the same crossing is not signposted for
> cyclists.
>
> Volker
>
>
> On Sat, 5 Dec 2020 at 17:05, ael via Tagging <tagging at openstreetmap.org>
> wrote:
>
>> On Fri, Dec 04, 2020 at 09:48:27PM +0000, Paul Allen wrote:
>> > On Fri, 4 Dec 2020 at 19:56, Martin Koppenhoefer <
>> dieterdreist at gmail.com>
>> > wrote:
>> >
>> > Up until around ten years ago, a minor road went past the end of the
>> > runway at what passes for an airport. The planes could be so low on
>> > approach to the runway that there were traffic signals to prevent
>> > vehicles crossing the path of an aircraft. There were also signs
>> > warning of low-flying aircraft, which I referred to as "Give way
>> > to aircraft."
>>
>> Also at much larger airports. Brize Norton
>> (https://en.wikipedia.org/wiki/RAF_Brize_Norton), for example.
>>
>> https://www.openstreetmap.org/node/190194553 for one of the traffic
>> lights.
>>
>> ael
>>
>>
>> _______________________________________________
>> Tagging mailing list
>> Tagging at openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/tagging
>>
> _______________________________________________
> 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/20201205/c77ab2cb/attachment.htm>
More information about the Tagging
mailing list