[Talk-us] tagging rumble-strip-separated road shoulders

[Paul Johnson]

On Fri, Feb 3, 2017 at 12:04 PM, Volker Schmidt <voschix at gmail.com> wrote:

> Hi Paul
> Thanks for the quick reaction - I knew you would reply.
>
>>
>> Well, the shoulder wouldn't count as a lane (a bicycle lane would,
>> however).  Not quite sure how to tag the bicycle use shoulder case (though
>> I am aware that it is extremely common in the US), but if it were an
>> actual, bicycle-only lane instead, assuming three lanes on a side with the
>> right lane being bicycle only, and signage indicating that bicycles must
>> use the bicycle lane:
>>
>> lanes=3
>> oneway=yes
>> cycleway=lane
>> motor_vehicle:lanes=yes|yes|no
>> bicycle:lanes=no|no|designated
>>
>
> This does not reflect the real situation, which is the presence of an
> emergency stop lane for motor vehicles, which may be used by bicycles. It
> is different from a dedicated cycle lane, which is what your tagging
> describes.  I have seen and ysed a number of bicycle lanes in the US that
> "deserve" your tagging.
> The shoulder=yes tag is being used a lot in combination with bicycle=yes
>

That works.  I'm not super-familiar with the shoulder key in general since
it's not *that* common in the grand scheme in North America, but I like
it.  Maybe for a typical expressway with wide outfield shoulder...

lanes=2
oneway=yes
shoulder=yes
bicycle=yes
bicycle:lanes=no|no

This one's a little bit of a headscratcher mostly due to shoulder tagging
being relatively uncommon on my continent, though I would generally expect
highway=(motorway|trunk) bicycle=yes to imply that anyway.  (I'm not
rehashing the argument that motorway and trunk should imply bicycle=yes
anyway since, mile for mile, that *is* the most common access combination
in the US).

http://overpass-turbo.eu/map.html?Q=%2F*%0AThis%20has%
> 20been%20generated%20by%20the%20overpass-turbo%20wizard.%
> 0AThe%20original%20search%20was%3A%0A%E2%80%9Cshoulder%
> 3Dyes%20and%20bicycle%3Dyes%E2%80%9D%0A*%2F%0A%5Bout%
> 3Ajson%5D%5Btimeout%3A25%5D%3B%0A%2F%2F%20gather%
> 20results%0A(%0A%20%20%2F%2F%20query%20part%20for%3A%20%E2%
> 80%9Cshoulder%3Dyes%20and%20bicycle%3Dyes%E2%80%9D%0A%
> 20%20node%5B%22shoulder%22%3D%22yes%22%5D%5B%22bicycle%22%
> 3D%22yes%22%5D(33.17434155100208%2C-116.773681640625%2C36.
> 98939086733937%2C-112.48901367187499)%3B%0A%20%20way%5B%22shoulder%22%3D%
> 22yes%22%5D%5B%22bicycle%22%3D%22yes%22%5D(33.17434155100208%2C-116.
> 773681640625%2C36.98939086733937%2C-112.48901367187499)%3B%0A%20%
> 20relation%5B%22shoulder%22%3D%22yes%22%5D%5B%22bicycle%
> 22%3D%22yes%22%5D(33.17434155100208%2C-116.773681640625%2C36.
> 98939086733937%2C-112.48901367187499)%3B%0A)%3B%0A%
> 2F%2F%20print%20results%0Aout%20body%3B%0A%3E%3B%0Aout%20skel%20qt%3B
>

Wishlist item for overpass-turbo...shorter URLs.  Seriously link gore
there.


> (2) how to tag longitudinal rumble strips,
>>> (situated between motorized-traffic lanes and shoulders, example: [1],
>>> not the "sleeping policeman" type that goes across the road nd which is
>>> normally tagged as traffic_calming=rumble_strip on a node of the highway)
>>> To tag their presence is important because they represent an augmented
>>> protection of cyclists on the shoulder from cars invading the shoulder by
>>> mistake, i.e. reduced risk of being killed.
>>> To note that I have encountered rumble-strip-separated shoulders also on
>>> roads below the rank of trunk or motorway
>>>
>>
>> I'm not quite sure how necessary it is at this point.  The kind of
>> example you have provided is being phased out in favor of strips with gaps
>> in them so bicyclists can get on and off the shoulder without dealing with
>> the rumbles, or eliminated on roads with a narrow hard shoulder.  In both
>> cases, for the reason for that is that not taking bicycles into account
>> does more harm than good.  This is true even on routes that are normally
>> closed to bicyclists except when police tell you to use it anyway (like
>> Interstate 70 in Kansas).
>>
> My point maybe was not clear enough: I would like to tag the longitudinal
> rumble strips, independently of whether they are continuous or interrupted,
> because I want to be able to classify route sections according to their
> level of bicycle safety. If I have a stretch of motorway with a shoulder,
> this same stretch is safer for bicycle use when a rumble strip is present
> than when it is not, because it reduces the risk of a motor vehicle
> invading the shoulder.
>

No, you were perfectly clear on that.  While longitudinal shoulder rumble
strips (SRS) are known to help reduce the number of run-off-road incidents
for motorists.  I was coming at it from the perspective that SRS
actually create
a nuisance hazard for bicyclists
<http://safety.fhwa.dot.gov/roadway_dept/research/exec_summary.htm>, which
is also why states are moving to friendlier designs that aren't quite as
deep and wide (older designs cross the *entire* shoulder, for example;
these can be found on parts of I 5 in Washington and older concrete parts
of the Kansas Turnpike where it runs diagonally NE/SW for example, and are
of such an amplitude that it can be difficult to steer a motor vehicle on,
to say nothing of a bicycle's experience on those).  The radical reduction
in number of bad SRS designs and further reduction in SRS designs that lack
bicycle gaps in the last decade (and especially since the minimum standards
for them were tightened in 2009) along with the perspective that SRS tend
to be a hazard for cyclists is where I was assuming you were coming from in
terms of mapping them.

In either case, I'm not sure there's a tag for SRS features yet.  Maybe
shoulder:rumble_strip=* ?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-us/attachments/20170204/e146e860/attachment.html>