[Talk-ca] Highway recoding

Ken Wuschke chandler.vancouver at gmail.com
Thu Jan 28 18:31:59 UTC 2016


Morning Daniel,

First, I really appreciated your comprehensive email. It provided me with a
lot of background on this topic.

As to my level of employment within the Ministry of Transportation and
Highways, I was on the operations end and dealt directly with the public.
Through this I quickly learnt and understood how the public perceives the
roading infrastructure in British Columbia. A perspective which can be
quite different from the engineering point of view. Hence I focus on
functionality of roads versus the design profile of a road. For the most
part the public does not care nor understand the difference between a
highway's cross section other than they can generally travel faster when
there is four lanes versus two lanes. Their focus is more about getting
places directly and quickly.

But I am going to step back from defining trunk routes in Canada for a
moment and focus on coming to OSM as a new person. Quite frankly the OSM
wiki is not clear nor well connected and it offers a lot of room for
improvement so that newcomers can learn what is the consensus based on
years of discussion. In the case of Canadian trunk status the wiki leaves a
lot to be desired.

1. On the highest level page *Tag:highway=trunk*
http://wiki.openstreetmap.org/wiki/Tag:highway%3Dtrunk Canada is not
mentioned in the table called *International equivalence*. Nor is there a
link to be found to get a person to the Canadian definitions for OSM. When
comparing through the countries that are listed in the table it ranges from
design specs of a road profile (i.e. Vietnam) to it is based solely on does
the road in question have a route number (i.e. Hong Kong). For the Canadian
OSM volunteer this leaves a huge room for interpretation. Therefore,
regardless of what the OSM community consensus is for Canadian trunk route
standards by not having a definition on this page nor a link to a
definition the new OSM volunteer is floundering in the dark.

To make it easier, I feel the *International equivalence *table be
abandoned and simply a link to the *Highway:International equivalence* page
- http://wiki.openstreetmap.org/wiki/Highway:International_equivalence. As
OSM is mainly volunteers there is far less work in maintaining one table
then maintaining several tables.


2. On the *Canadian tagging guidelines* page
http://wiki.openstreetmap.org/wiki/Canadian_tagging_guidelines#Trunk has a
link to an unnamed document of the NHS inventory, however clicking on that
link simply takes the OSM volunteer to a blank page. Further there is no
mention that some provinces on the OSM wiki have further tagging
guidelines. I only learnt this through Paul Norman's email earlier in this
thread.

3. On the *Canada:British Columbia* guidelines page
http://wiki.openstreetmap.org/wiki/Canada:British_Columbia#Highways_and_provincial_roads
trunk roads are defined differently from the Canadian standard. Instead of
using the National Highway System as defined by the Council of Ministers
defining a trunk road is more about whether or not it is a divided highway.
In addition, this page does not refer to the BC Ministry of Transportation
and Infrastructure's document *Overview of B.C. Highway Functional
Classification.*

As you can see there is a lot of confusion not just with the definition but
with how to access the information and how the guideline changes from a
provincial to a national guideline page on the OSM wiki. If you can make
suggestions on how to improve this situation I would welcome them.

To be frank this level of confusion just encourages me to stop contributing
to OSM. I can imagine there are more people out there like me. Life is too
short for this type of frustration, let alone on a volunteer level.

Getting back to the definition. I found this document by the BC Ministry of
Transportation and Infrastructure of great interest - *Overview of B.C.
Highway Functional Classification* /
https://www.th.gov.bc.ca/publications/planning/Provincial%20Highways/BC_Numbered_Hwy_Functional_Classes.pdf
- In it discusses five levels of functionality class. While it does not
discuss trunk route, it does provide primary, secondary, major, minor, and
local roads. But then it tosses in something that could be subscribed to as
a trunk road:

*The “National Highway System” (NHS) is a subset of primary highways which
have been deemed to be of national importance, and therefore which have
higher expectations placed on them regarding mobility, reliability,
geometric standards and condition. *


This is very close to the OSM definition on
http://wiki.openstreetmap.org/wiki/Canadian_tagging_guidelines#Trunk -

*A highway=trunk is a roadway that has limited access and is part of the
national highway system, as defined by the Council of Ministers, an
intergovernmental agency with representatives from each province and
territory. Maintenance of these highways is under provincial jurisdiction.*


*The surface=* does not need to be paved, nor is it assumed to be paved. As
in many regions of Canada (far north in the Yukon / Northwest Territories,
there is no need for it to be paved, as there is not a high volume of
traffic. But it still has that national level of importance. This is the
core cross-Canada road network, and it should be viewable at low zoom
levels.*

The problem is that the OSM definition puts in the design profile as a
requirement.

Then when I go the Council of Ministers Responsible for Transportation and
Highway Safety document *Canada’s National Highway System Annual Report
2014* a very clear definition is found:

*The [National Highway System] comprises three categories of routes, each
of which are defined by specific criteria that can be used to assess route
eligibility:*


*Core Routes*

*- Key interprovincial and international corridor routes (including links
to intermodal facilities and **important border crossings)*


*Feeder Routes*

*- Key linkages to the Core Routes from population and economic centres
(including links to **intermodal facilities and important border crossings)*


*Northern and Remote Routes*

*- Key linkages to Core and Feeder routes that provide the primary means of
access to northern **and remote areas, economic activities and resources. *

When I read this is clear that the Council of Ministers are defining a
trunk routes in Canada, while they do not use the word. So again I lean
towards functionality of a route rather than the routes specific profile
for applying the tag of trunk to it.

So a suggestion to a definition for trunk routes in Canada could be a
simple as:

*A highway=trunk is a roadway that is a part of the National Highway System
as defined by the Council of Ministers Responsible for Transportation and
Highway Safety and is found in annually updated document called **Canada’s
National Highway System Annual Report.*

This eliminates all the interpretations done by individual OSM
contributors. Further I would recommend that provincial and territorial
variances of the *highway=trunk *definition be removed to avoid confusion.
i.e.
http://wiki.openstreetmap.org/wiki/Canada:British_Columbia#Highways_and_provincial_roads

Cheers,
Ken

On Thu, Jan 28, 2016 at 7:40 AM, Begin Daniel <jfd553 at hotmail.com> wrote:

> Bonjour Ken,
>
> The text below is a bit long but it will give you a lot of context around
> current topic J
>
>
>
> First, I also used to work for a government agency (Natural Resources
> Canada)!
>
>
>
> I was in charge of the National Road Network (NRN) project when it started
> in 2001 – mapping the whole Canadian road network using GPS technology. I
> held negotiations with provincial agencies about its content, and to insure
> they would keep the NRN updated after we had created the first version.
> Since then, provinces/territories are keeping the resulting product updated
> and make it available as a GeoGratis (formerly GeoBase) product: National
> Road Network NRN (Google it!)
>
>
>
> There is no *trunk* in the NRN product since most Canadian *Mapping*
> agencies do not use it. They describe road segments based on their
> infrastructure. Maps represent objects based on what can be seen on the
> ground (data capture from aerial photography, field work with GPS…). Well,
> very similar to what OSM contributors are doing ;-)
>
>
>
> Most Canadian *Transportation* agencies do not use it either – the pdf
> document cited in the wiki uses core and feeder roads instead. Their
> classification relies on the importance of a road and the traffic it must
> support, in order to invest in appropriate infrastructures. The link
> between OSM *trunk* definition and the *core* roads described in the pdf
> was made in the wiki about 10 years ago, without much discussion with the
> community!
>
>
>
> *About OSM Road Network*: Most of the OSM road network in Canada is an
> import of the NRN. NRCan made it available in OSM format through the Canvec
> product (reference: 1, 2) – to help the community in creating the OSM
> database. The resulting road classification is then the result of
> discussions held between NRCan and the Canadian OSM community in 2009, few
> years after the definition of trunk was written in the Canadian tagging
> guidelines.
>
>
>
> That is why I consider that the 2009 definitions (reference 2) superseded
> the tagging guidelines. In 2009 classification, a trunk is a NRN’s Freeway
> (with access restrictions) that has no separated ways for most of its
> length. That is the origin of my initial proposal.
>
>
>
> I understand pretty well your point of view considering your background
> (transport related) but it seems to me so different from the consensus
> gained in the past years which is more mapping related. There is obviously
> a strong correlation between both classifications since the investments
> made on an important road segment (transport consideration) will eventually
> have an impact on its infrastructure (mapping considerations).
>
>
>
> However, I would not like to see more than a year of discussion with the
> community being reverted (and all the current tagging) because of the
> existence of contradictory trunk definitions. Furthermore, because there is
> a strong correlation between the NRN product and current OSM
> classification, many contributors are maintaining OSM road network updated
> by using the NRN as a reference.  There is even a WMS layer (GeoBase Roads)
> that provide the NRN as a reference when editing in JOSM and the tagging
> conversion is easy because of that strong correlation with OSM (reference
> 2).
>
>
>
> As I wrote in a previous email, I am just looking for definitions that
> keep aligned with OSM, the consensus obtained in 2009, and that minimise
> interpretation gaps between tags. Do you have any new ideas/proposal
> considering the above?
>
>
>
> Best regards,
>
> Daniel
>
>
>
> References:
>
> 1-      https://wiki.openstreetmap.org/wiki/CanVec
>
> 2-      https://wiki.openstreetmap.org/wiki/CanVec:_Transportation_(TR)
>
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-ca/attachments/20160128/c2d673ab/attachment-0001.html>


More information about the Talk-ca mailing list