Classification of roads in the DR Congo

[ian727]

Hello community,

I’ve been thinking of this issue for a long time already, but the personal experience in the region during the past months confirmed my thoughts:

I’m primarily mapping in the DR Congo, which is due to my interest and studies in development politics and Geography. In May, I started a 3-month internship in the country which luckily allowed me to see various parts of the Congo.

The first issue is the following: How to exactly classify roads? I’ll describe a good example: Through the provincial town of Gungu goes the route RP230 (Route Provinciale prioritaire 230). The whole RP230 is not paved while the condition of the compacted dirt varies for every segment. Non 4x4 vehicles cannot pass. The road number was given to this road by the ‘Office des routes’, the national ministry for roads, somewhen around 1997. Back then, the road went right through the city, approaching the Kwilu river where there used to be a car ferry, then continuing eastwards. In 2006, a bridge was constructed quite some kilometers north, changing the path of the RP230: Not only in the city (the route is by-passing Gungu now north) but also east of the Kwilu river (vehicles use a new segment, the old road some kilometers south is only used to access the villages there). As it is typical in the Congo, this change is not found in any official documents of the Office des routes, leading to a first conflict.

But there is even more: West of Gungu, the RP230 is completely destroyed by erosion and impassable. Cars (4x4) take a detour, going over some fields, bypassing the destroyed section. Trucks can (as of now) not pass, taking a huge detour via another road. It is unclear if and when the RP230 will be rehabilitated.

Now for the first issue: What is to map as RP230? The theoretic course as given by the Office des routes and that is found on other maps? For the former course through the village, that wouldn’t make too much sense as the ferry does not exist anymore and nobody ever uses the old course. As of now, I classified the new road as RP230. For the course east of the Kwilu river, it is a bit more difficult, as the old course still exists and is in an acceptable condition because it is still used by vehicles that want to reach villages along it. However, its purpose changed as vehicles that actually want to follow the RP230 do not use this segment anymore. There, I left the classification as RP230.

For the second issue: I honestly don’t know how to deal with it. For now, I did not modify the course of the RP230, although this section is impassable. I don’t want to classify the bypass as RP230 because this is really just a small path going over some fields.

Now that I have typed that much, let me as well approach a third issue: In the whole region, almost every village was once connected to the road network (dating back to colonial times). Nowadays, those roads are barely used at all. You can pass them with a well-equipped 4x4 jeep, but they are totally overgrown with grass and sometimes barely visible. From their look, I would not even classify them as “Unclassified Road”. For now, I map them as “Unmaintained track roads”, although, as I understood, this classification is mainly used for agricultural roads – and that is not the case here, those are actually the roads to access the villages.

If you read this until here, thank you very much. I would be happy to get some ideas for the issues, if you need any more info, I’ll gladly provide it! I also have a lot of pictures, but I think the description does quite well.

Kind regards,

ian727

[hadw]

You can’t even use the official route at all, unless it has been released under an OSM compatible licence, which seems unlikely.

In any case, the route has to be the de facto route. I would say the remaining question was whether there is sufficient evidence to actually associate the RP230 reference with the new route, or whether one should show breaks in RP230, with no reference on the new sections.

If the village tributary roads are sometimes used by motor vehicles, I believe they have to be at least unclassified and you would use surface, smoothness, and tracktype (not limited to tracks) to indicate their quality. I’d suggest it also reasonable to use trail_visibility, even though it doesn’t officially cover vehicular roads.

The recent Canadian forest roads thread may also be relevant.

[ian727]

@hadw

You can’t even use the official route at all, unless it has been released under an OSM compatible licence, which seems unlikely.

Not sure what you mean by that. I add lines (roads in this case) to OSM based on sattelite imagery and my own GPS tracks.

@hadw

In any case, the route has to be the de facto route. I would say the remaining question was whether there is sufficient evidence to actually associate the RP230 reference with the new route, or whether one should show breaks in RP230, with no reference on the new sections.

Pretty much, yes.

@hadw

If the village tributary roads are sometimes used by motor vehicles, I believe they have to be at least unclassified and you would use surface, smoothness, and tracktype (not limited to tracks) to indicate their quality. I’d suggest it also reasonable to use trail_visibility, even though it doesn’t officially cover vehicular roads.

The usage really differs and also varies a lot. However, if you want to reach the villages, you’ll have to use exactly those roads. What I want to point out is that they are not usable for normal cars, only for at least 4x4 vehicles (we went there with a Toyota Land Cruiser, sometimes the road condition was so bad that you could barely go 10 km/h). However, the classification “Minor/Unclassified Road” renders the road regardless of any additional tags such as surface, smoothness, tracktype - right? By looking at the map, you should be able to tell that you cannot use this road with a normal car - but at the same time, you should be able to see that it is the main connecting road. Difficult, eh?

Thanks for your reply!

[muralito]

Hello ian727.

As OpenStreetMap is the geographic database, roads only usable for 4x4 vehicles should be tagged with 4wd_only=yes.

When you say

By looking at the map, you should be able to tell

keep in mind that “the map” is not what you see in openstreetmap.org, and that there are a lot of maps produced from OSM data, each one with their own requeirements and use cases, and showing what and how their producers think it has to be shown.

[alan_gr (Alan)]

@ian727

However, the classification “Minor/Unclassified Road” renders the road regardless of any additional tags such as surface, smoothness, tracktype - right?

Not necessarily - it depends on the renderer or application. For example, OsmAnd has some user-configurable options to render road surface and quality. And OpenAndromaps renders surface=paved differently from surface=unpaved (I’m not sure which other surface= values it distinguishes).

@ian727

I map them as “Unmaintained track roads”

I have seen this terminology used elsewhere on the forum and I’m not sure where it comes from. Does this simply mean highway=track? Or highway=track with some other tag to indicate it is unmaintained?

[ian727]

@alan_gr

I have seen this terminology used elsewhere on the forum and I’m not sure where it comes from. Does this simply mean highway=track? Or highway=track with some other tag to indicate it is unmaintained?

It is simply highway=track. The terminology comes from the OSM browser editor iD. You’re probably using josm for your edits, hence you never came across it! Thanks for your reply!

[hadw]

@ian727

@hadw

You can’t even use the official route at all, unless it has been released under an OSM compatible licence, which seems unlikely.

Not sure what you mean by that. I add lines (roads in this case) to OSM based on sattelite imagery and my own GPS tracks.

If you are only using satellite imagery, you cannot add names or references to any road! Quite a lot of your original article implied that there was more than satellite imagery being used as sources. You cannot tell the state of a road without surveying it on the ground.

However, I was particularly responding to:

As it is typical in the Congo, this change is not found in any official documents of the Office des routes, leading to a first conflict.

Unless they are released on OSM compatible terms, the official documents don’t exist as far as OSM is concerned. Any designation for the road is likely to have to be based on your experience travelling on it, and what local people tell you that the road is.

Others have responded to your rendering point, but basically there is a rule in OSM that you do not tag for the renderer, but rather accurately tag and assume that anyone needing the map for a particular purpose will choose or pick a renderer that interprets the tagging appropriately.

[ian727]

@hadw

If you are only using satellite imagery, you cannot add names or references to any road! Quite a lot of your original article implied that there was more than satellite imagery being used as sources. You cannot tell the state of a road without surveying it on the ground.

I was personally in the ‘territoire’ and drove on most of the roads myself. However, I prefer to map based on sattelite imagery because my GPS tracks are not that accurate.

@hadw

Unless they are released on OSM compatible terms, the official documents don’t exist as far as OSM is concerned. Any designation for the road is likely to have to be based on your experience travelling on it, and what local people tell you that the road is.

No. The Congolese people living there barely use the official designations for the roads, they simply call it (for example) “Route Kahemba” referring to a town which the road passes.

Check this page for the official designations: https://wiki.openstreetmap.org/wiki/WikiProject_Congo-Kinshasa/Routes

[hadw]

In that case you have two sources, which should both be listed. survey (or local_knowledge) for the coarse geometry, naming etc., and satellite/aerial for a slightly better geometry.

The name on the road should be “Route Kahemba”. Whether you can put a reference is likely to depend on whether a reasonable person, with more than a parochial interest, would consider the section part of the originally designated route.

It sounds to me as though the designations are mainly only historical, and OSM is not a historical map.