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

Brent Easton b.easton at exemail.com.au
Tue Jul 31 06:36:43 BST 2007

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'.


*********** 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
>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
>as can be seen here:
>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:
>> 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
>do need a solution that caters for route planners and navigation devices.
>What is the best solution?
>talk mailing list
>talk at openstreetmap.org
>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                       
University of Western Sydney                                   
Email: b.easton at uws.edu.au

More information about the talk mailing list