PowerGEM is a power generation and distribution analysis company based in the United States. Over the past few years, we have integrated spatial concepts into our software with OpenStreetMap being a key component of that strategy.
One of the types of data that we use is power substations (OSM tag: power=substation). While OSM contains thousands of substations there are still more to be added. One source of substation data is from Homeland Infrastructure Foundation-Level Data (HIFLD). However, this data was taken offline in 2022 and has not been publicly updated since. Copies of the latest vintage abound on the internet with one being here. This effort is about integrating the substation data from HIFLD into OSM.
I haven’t looked at the HIFLD substation data, but my experience with other HIFLD datasets is that they have sometimes-substantial position errors. How are you planning to deal with this?
There are more details on the Wiki but we created a custom application with both OSM tiles and Bing Ariel Imagery as base layers with the HIFLD substation on top. If there is a substation in OSM we move on to the next HIFLD substation. If there isn’t a substation in the Bing ariel layer we skip it and move on to the next HIFLD substation. If there is a substation indicated in the Bing imagery then we draw the polygon that represents the substation.
I think it’s fair to say that we are using HIFLD substations to “hint” where a substation could be so we pan and zoom to that location and if we see a substation in Bing imagery but not in OSM we draw it by hand and upload it to OSM.
Hi @PowerMapper77
It’s great to see data users like PoweGEM contributing back to OpenStreetMap. Mapping the missing substation of HIFLD using the MapYourGrid tool in JOSM is relatively straightforward. I have mapped hundreds of substations in this way in the US. If you would like training on using the JOSM+MapYourGrid tools, please let me know. We provide two hours of free training and have experience of mapping in the US. The planned substation layer of OurGridFuture also provides useful information for maintaining the data beyond HIFLD: https://ourgridfuture.org/
While the changeset give a rough source there is still no mention of the wiki page nor is there any indication that the account used is working for a company.
It has been explained that these things are required when editing in such a manner. Please update the user profile and add a link to either the forum discussion or the wiki page to each edit. Thank you
@PowerMapper77 just because I’m curious: why are there so many “remove duplicate” changesets when the workflow you described should avoid producing duplicates in the first place?
Firstly, The tool we have displays a split screen with the OSM tiles on the left and Bing Aerial Imagery on the right. The HIFLD substation is displayed as a point on both sides. The person doing the tracing was instructed to look at both sides and if a substation exists on the left, to not draw it. However, they failed to check the left-side map so if the satellite image displayed a substation they would draw it, possibly causing a duplicated substation in OSM after upload.
Secondly, when we were told to stop, we were also told that our changesets were reverted / deleted. We thought this applied to ALL of our substation uploads. When we resumed work last week, we updated our geojson files to remove the substation name, and re-uploaded. Unfortunately, we did not realize that not all of our changsets were reverted, causing duplicates.
We are now going through everything we have uploaded and deleting those duplicates.