[OSM-talk] traffic signals at dual-carriageway and complicated intersections

[Brent Easton]

>I really like this suggestion.  However, what should be done when two
>intersecting roads are dual-carriageway and different highway types.  In
>this case the middle square of the intersection would have two sides of a
>particular type and the other two sides a different type.
>
>Of course, it is possible to have multiple ways over the same segments, and
>this would be the solution.  Was this what you had in mind?


Hmm, hadn't though of that possibility. 

Yes, you could create a new way over the top of the roads that carries the 'junction' tag. That should do the trick. It would not have a 'highway' tag, so would not render any roads, but would just generate the traffic lights rendered image. You would only need to use this in the circumstance you outlined, where different parts of the junction need to be rendered differently. 

Auto-routing/description software would need to be aware of this, but it shouldn't be a problem.


>On 7/31/07, Brent Easton <b.easton at exemail.com.au> wrote:
>>
>>
>> I think it has been mentioned before, but I think the only real answer is
>> that the middle square of the intersection becomes it's own way with a
>> 'junction=traffic_lights' tag.
>>
>> The renderers can than display just one symbol at the c.o.g. of the
>> junction way. This reflects the reality that there is only one junction
>with
>> one 'set' of lights, not 4 seperate junctions, each with their own set of
>> lights.
>>
>> This also works where a divided road is crossed by a single way, the
>short
>> line connecting the two ways would be created as it's own way with
>> 'junction=traffic_signals'.
>>
>> Brent.
>>
>>
>>
>> *********** REPLY SEPARATOR  ***********
>>
>> On 31/07/2007 at 10:41 AM Lachlan Rogers  wrote:
>>
>> >I've been wondering about the best way to map complicated intersections
>> >with
>> >traffic signals. A common example of where my dilemma arises is when
>> there
>> >are traffic signals at any intersection involving a dual-carriageway
>> road.
>> >
>> >I've been tagging each node involved in the intersecting ways, since
>this
>> >seems the best method for providing route-planners with accurate
>> >information.  However, the rendering typically involves overlapping
>> >symbols,
>> >as can be seen here:
>> >
>>
>http://informationfreeway.org/index.php?lat=-35.27578979216632&lon=149.12939396721774&zoom=17&layers=B000F000
>> >
>> >The other aspect of this that has bothered me is that navigation systems
>> >would probably see duplicate sets of lights and give incorrect
>> instructions
>> >such as "turn left after the next two sets of lights" when it is really
>> >referring to a single intersection.
>> >
>> >In another thread recently, Cameron Patrick mentioned
>> >
>> >> In my experimental maps of Australia, I've used a transparent orange
>> >> circle on the intersection.  Look along e.g. Stirling Highway in the
>> >> middle of this map:
>> >>
>> >
>>
>http://osm.largestprime.net/map/?lat=-3738202.21988&lon=12890789.70953&zoom=15&layers=B0
>> >
>> >> What are people doing for traffic lights on junctions of dual
>> >> carriageways?  I've been placing a node in the middle of the
>> >> intersection, not sitting on any of ways involved, which renders
>nicely
>> >> but lacks semantic information for route planners.
>> >
>> >I like the way his traffic signal intersections are rendered, but I
>think
>> >we
>> >do need a solution that caters for route planners and navigation
>devices.
>> >
>> >What is the best solution?
>> >
>> >_______________________________________________
>> >talk mailing list
>> >talk at openstreetmap.org
>> >http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
>> >
>> >No virus found in this incoming message.
>> >Checked by AVG Free Edition.
>> >Version: 7.5.476 / Virus Database: 269.11.0/927 - Release Date:
>> 30/07/2007 5:02 PM
>>
>>
>> ____________________________________________________________
>> Brent Easton
>> Analyst/Programmer
>> University of Western Sydney
>> Email: b.easton at uws.edu.au
>>
>>
>> _______________________________________________
>> talk mailing list
>> talk at openstreetmap.org
>> http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
>>
>
>No virus found in this incoming message.
>Checked by AVG Free Edition. 
>Version: 7.5.476 / Virus Database: 269.11.0/929 - Release Date: 31/07/2007 5:26 PM


____________________________________________________________
Brent Easton                       
Analyst/Programmer                               
University of Western Sydney                                   
Email: b.easton at uws.edu.au