Several colleagues work together in a research project to calculate the rooftop PV potential in the Basque Country, and probably we will do it also for several regions of Spain.
GeoEuskadi (one contributor to OpenStreetMap) provides information about buildings in the Basque Country.
We obtain the results in vectorial files (shape files, but we can also convert information into csv files), so we can assign several relevant data to each roof, in a large area (every file can have information for a whole city).
We would like to share the results of this research, and upload them to OSM, so anybody can click on their building and see that data.
Is it possible to upload these files? What format should they have? Where will they be stored? Should we provide a server to store that data?
As OSM is a database of things that are actually there and not about abstract information like PV potential, I am not sure OSM is a good place for that data. If your aim is to have a visual representation of your data on a map, I’d suggest to use OSM as a background and add your shapes and associated data as a layer on top of that. It should be possible with Leaflet or Mapbox.
@RosaLA: I don’t think adding PV potential to OSM objects within OSM is the best way to proceed. Defined potential will greatly depend on the underlying methodology and types of information used. It is intrinsically non-verifiable by a normal mapper even if the methodology was available.
I think the best process would be to make available files associating a given building’s PV potential with a specific OSM element. In case elements get deleted or changed significantly you can also add something like the lat/lon of the building centroid as well, which would enable any mismatches to be caught & processed. Also a centroid would allow buildings not already in OSM to be matched at a later date.
The kind of information which would be appropriate in OSM are properties of the building: height, roof:height, roof:material, roof:shape etc. I imagine many of these might be used in your estimation algorithm (even if indirectly, if it uses machine learning). These properties may also help for different algorithms or for other purposes related to understanding building resilience in a changing climate.
OpenSolarMap was a project carried out a while ago in France to identify Solar PV potential, and involved several members of the OSM-FR community, including @cquest, who may be able to say what they did with their results. We haven’t yet used solar data from the UK as training data for this type of project, but we now have very substantial coverage suitable for this type of work.