Hi all, I’m undertaking a project to manually import electricity infrastructure from National Grid Electricity Distributions open data. Most details can be found on the wiki page here: NGED Import Project - OpenStreetMap Wiki. The project has already started but was paused until recently do to a licencing issue that’s now been resolved. When I started the project originally I only consulted the OSM discord so before I resume I would like to consult the forums as well.
1 Like
Hi kitsee and welcome! Your project sounds really interesting. Are you able to link to the previous discussion on Discord? (It would also be good to document it on the wiki page.) I don’t know how many members of the UK OSM community are active on Discord, so it would be really good to have a full discussion here before you continue. (I assume you’ve read Import/Guidelines - OpenStreetMap Wiki .)
As the person who is likely most familiar with how GB power infrastructure is mapped in OSM, I’m happy for this import to go ahead as described, with one minor issue which I will cover at the end of this post.
This is very much a manually-guided import, and it needs to be, as some parts of the source data are not terribly accurate in terms of location. I would argue that the proposed process has sufficient manual involvement that it shouldn’t be classed as an automated edit – note that “every node and way will then be manual inspected and adjusted with use of imagery”.
I spotted this import a few weeks ago and I have been in contact with @kitsee because unfortunately the licensing of the source dataset has recently changed to be non-open. @kitsee has been extremely helpful in the face of my licensing pedantry and we have resolved this. (I’ve been in contact with the DWG about this and they were also quite complimentary about the quality of the import!)
I’m also not in Discord and I asked @kitsee to post here before proceeding further with the import. However, I have reviewed the areas which have already been imported and they look great. In my opinion, having all power lines comprehensively mapped is a benefit to all map users, not just us infrastructure nerds.
My one objection is that I don’t think we should import 11 kV underground cables at this stage. In urban areas there are a lot of these, and they will clutter up the map, confuse novice editors, and be difficult to maintain. They are not surveyable and in my opinion they have limited use to data consumers. I’ve generally avoided adding 11 kV underground cables in the past for this reason.
3 Likes
Hi Both,
I’ve documented the original Discord convo in a link at the end of the wiki page however I realise now that the original conversation was no where near adequate. Due to my naivety and over estimation of the documentation work required I attempted skirt the import rules. Before resuming this project I am attempting to correct those mistakes here. I do apologize and I do hope you will forgive me for this.
@Russ I apricate the compliment, it always nice to have your work validated especially from DWG too.
Concerning the underground lines, I understand your argument but I must say as a completionist I’m also very reluctant especially around substations as a good percentage of overhead lines connect to substations by short lengths of underground lines. my dream for this project is to have enough data available so that a consumer would be able to trace their power from a location to the nearest generation sites following the lines. this would require for the most part an unbroken chain for the location to the generation sites.
In ID it appears the default rendering for power=cable is a almost invisible line which would alleviate some opportunity for confusion but if a novice user still selected one then if could still cause a issue. One of the worst locations I’ve encountered for this issue would be here: Open Infrastructure Map
OpenStreetMap
Not only do you have a substation with a larger than average number of 11kv connections but a good number of those connections follow the same path south for some time.