[Talk-GB] National Cycle Network discrepancies between Sustrans and OSM data

Mon Feb 27 20:14:10 UTC 2023

Thanks Dave,

I've added the OSM map as an optional background (see bottom icon on 
right of map for options).

Your second point about "entrance paths" is probably one where there are 
different opinions. Within the Sustrans dataset some of the paths (e.g. 
connections from the local housing estate to the NCN) are included in 
the Sustrans NCN data, some aren't. Some are labelled as "links", some 
as "NCN". I'll be encouraging Sustrans to follow a consistent approach - 
I don't really care whether the paths are included or not so long as 
they are consistent - and would then argue that OSM should probably 
follow the Sustrans approach (whatever it is).

However, until Sustrans have a consistent approach there don't seem to 
be any issues in OSM mappers following their own decisions for these 
linking / entrance paths.

Ian

On 27/02/2023 18:10, Dave F wrote:
> This is useful. Thanks
>
> The two main discrepancies in my locale are
>
> 1. One way routes, which appear to be all Sustrans' lack of detail.
> 2. Entrance paths onto old railway lines/towpaths. Although 
> technically within Sustrans' maintenance  remit, I never considered 
> them to be a part of the linear route, so won't be tagging them as NCNs.
>
> Could you add the standard OSM-Carto layer as an option, maybe even 
> black & white, as I'm finding the cycle layer can obscure some of 
> Sustrans' data, such as links.
>
> DaveF
>
> On 25/02/2023 20:45, Ian Dent wrote:
>>
>> I’ve been on a campaign to get the accuracy of the Sustrans National 
>> Cycle Network improved for quite a while now and have made some 
>> progress in engaging with Sustrans.
>>
>> I’ve found that the NCN plotted on OpenStreetMap differs in a number 
>> of places from that in the Sustrans mapping.
>>
>> I’ve created a coarse visualisation of the differences – see 
>> https://dent.org.uk/sustrans/sustrans-ncn-inconsistences/ - which, by 
>> focusing on the thick red and blue areas shows which areas to 
>> investigate in detail. Note this is a snapshot at Feb 23 and won’t 
>> reflect any changes in OSM or Sustrans data until I update it. Also 
>> note it is intended to show where to focus and some of the smaller 
>> focus areas may be errors in how I’ve done the mapping.
>>
>> I know the Derbyshire NCN fairly well and have found that most of the 
>> discrepancies are due to errors in the Sustrans data rather than OSM 
>> and suspect this may be the case across the country.
>>
>> Sustrans have gone through a process of removing some of their routes 
>> and “reclassifying” others (i.e. not being responsible for them in 
>> the future). I’ve excluded these removed and reclassified sections 
>> from the map and the comparison with the OSM even though a lot of the 
>> routes still appear in OSM as they still have signage on the ground. 
>> I’m taking advice on what to do about removed and reclassified routes 
>> and will post a separate note on this subject later.
>>
>> I’m keen that Sustrans internally review the discrepancies and make 
>> changes to their mapping when it is in error. I’ve had some success 
>> in discussions but it is a slow process! There is discussion about 
>> involving the Sustrans volunteer force who are very geographically 
>> spread and will have local knowledge.
>>
>> I’m also keen to understand how best to communicate the discrepancy 
>> areas to people within the OSM community with local knowledge and who 
>> can, where the OSM is in error, make the necessary OSM changes.
>>
>> Discrepancies are for various reasons including:
>>
>>  *
>>
>>     mismatches on how the route is mapped (the map shows differences
>>     of 10 metres or more).
>>
>>  *
>>
>>     Routes that Sustrans have classed as regional but OSM as national
>>     (or vice versa). e.g. NCN 30 south of Lowestoft, NCN 568 on The
>>     Wirral
>>
>>  *
>>
>>     New Sustrans routes that haven’t yet reached OSM – e.g. NCN 28
>>     east of Plymouth
>>
>>  *
>>
>>     Routes that Sustrans classes as “links” but OSM has as NCN – e.g.
>>     NCN 28 near Dartmouth. I think these are generally Sustrans errors.
>>
>>  *
>>
>>     Routes that OSM have included as NCN but which are not part of
>>     the Sustrans network. e.g. NCN 627 north of Sheffield, NCN 422
>>     near Wokingham
>>
>>  *
>>
>>     Some ferries are included as routes in OSM – e.g. NCN 2 near Plymouth
>>
>> Sustrans data for the NCN can be found at 
>> https://data-sustrans-uk.opendata.arcgis.com/ (Open Government 
>> Licence). Note that a lot of this isn’t accurate so don’t take it as 
>> gospel – on the ground survey is needed.
>>
>> I’ll continue to work with Sustrans to try and get their errors 
>> improved. I’d appreciate advice on how to get the OSM community to 
>> consider the discrepancies and fix any OSM errors found (local 
>> knowledge needed).
>>
>> Thanks to those who’ve already discussed this with me – particularly 
>> the recent East Midlands OSM meeting.
>>
>> Ian
>>
>>
>> _______________________________________________
>> Talk-GB mailing list
>> Talk-GB at openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/talk-gb
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-gb/attachments/20230227/e50ff502/attachment.htm>