[Accessibility] Mapping complex road junctions with pedestrian crossings

[Barbeau, Sean]

Lulu-Ann,
Thanks, we've been advocating sidewalk-as-a-separate way to anyone that will listen, including the OSM listserve.  We'd definitely be interested in presenting at the next "State of the Map." We had scheduling conflicts this past year which prevented us from attending.

I agree with your observation, in reality there aren't any perfectly symmetrical junctions, which is why we recommend sidewalk-as-a-separate-way.  It seems that some intersections in dense urban areas (e.g., downtown NY) can be modeled very simplistically as perfectly symmetrical, but typically that's looking at a single attribute like presence of pedestrian signals at the crossings.  So, you could label a 4 way crossing intersection with a single "pedestrian signal" tag, which would mean that all 4 crossings have pedestrian signals.  In our opinion, however, once you get down to high detail and many attributes, sidewalk-as-a-separate-way and explicit coding is preferred, otherwise the exceptions list gets large and complex.

Sean

-----Original Message-----
From: Lulu-Ann at gmx.de [mailto:Lulu-Ann at gmx.de] 
Sent: Tuesday, February 14, 2012 6:25 AM
To: Accessibility; Barbeau, Sean
Subject: Re: [Accessibility] Mapping complex road junctions with pedestrian crossings

Hello list members,
hello Sean,

thank you for your information and recommendation.

Actually probably the persons on this list are already aware of the fact, that sideways need to be tagged as separate ways. :-)

You are kindly and urgently asked to inform those persons on talk at openstreetmap.org , who don't believe us yet!
Maybe you want to present your results on the next "State of the Map"?

I guess you are wrong on one assumption:
There are no perfectly symmetrical junctions.
If you find one, anyone will invent a new tag in the future that makes a reason to be aware of one object that is not symmetrical.
So to reduce re-tagging it is sensible to tag even junctions, that seem to be symmetrical, with separate sideways.

This was just a fast reply, I will read your links later.

Best regards
Lulu-Ann



-------- Original-Nachricht --------
> Datum: Mon, 13 Feb 2012 10:44:02 -0500
> Von: "Barbeau, Sean" <barbeau at cutr.usf.edu>
> An: "accessibility at openstreetmap.org" 
> <accessibility at openstreetmap.org>
> Betreff: Re: [Accessibility] Mapping complex road junctions with 
> pedestrian crossings

> Hi all,
> We've looked at general accessibility issues in OpenTripPlanner (OTP) 
> and OpenStreetMap as part of a research project at USF 
> (http://www.locationaware.usf.edu/ongoing-research/projects/open-trans
> it-data/), and we would strongly suggest that micro-mapping of 
> intersections in OpenStreetMap be done according to the "Sidewalk as a separate way"
> (http://wiki.openstreetmap.org/wiki/Proposed_features/Sidewalk_as_sepa
> rate_way) convention that Flaimo mentions.  This convention seems like 
> the best method to fully support detailed accessibility routing for 
> intersections that aren't perfectly symmetrical.
> 
> There are two distinct categories for intersection mapping:  implicit, 
> in which it is assumed that an intersection is perfectly symmetrical 
> (e.g., curb cuts/tactile strips are universal to all sidewalk/road) 
> unless exceptions are stated, and explicit, in which the detail for 
> the entire intersection is coded.  For dense, well-established 
> pedestrian environments like New York or Washington, D.C., the 
> implicit code might be preferred due to less effort by OSM users when 
> coding data.  However, for suburban/rural areas in which pedestrian 
> infrastructure isn't as dense and sidewalks aren't universal the 
> explicit coding seems to be the preferred method.  There is still an 
> open area of work for how routing is handled in transitions between 
> implicitly and explicitly coded areas, and to my knowledge 
> OpenTripPlanner hasn't tackled this yet.  We've coded OSM data 
> according to explicit "Sidewalk as a separate way" method and tested 
> it in OTP on the USF campus in our OTP deployment (http://opentripplanner.usf.edu/), and this method works best for our suburban campus.
> 
> More details about "Sidewalk as a separate way" OSM coding in context 
> of OpenTripPlanner is available in a presentation here 
> (http://goo.gl/1xltp) starting on slide 6, and in a final report here 
> (http://goo.gl/k3Uiy) starting on page 47.
> 
> I should mention that our research team that worked on this is 
> sighted, and we would welcome any additional insight into this problem 
> from anyone with visual impairments, or anyone with differing opinions.
> 
> Thanks,
> Sean
> 
> Sean Barbeau
> Research Associate
> Center for Urban Transportation Research University of South Florida
>
--
Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de