Assuming you meant accuracy here…
In my opinion, 99.9999% accuracy is not achievable in OpenStreetMap.
This allows only 1 in 1 000 000 objects to be wrong. Good luck with achieving that level of accuracy. For AEDs this level of accuracy would require at least hiring people to check each AED on hourly basis.
Then you shouldn’t advertise with OSM data to emergency services. Emergency services require a high degree of accuracy.
Thanks for pointing that out so clearly.
If I would be advertising OSM to emergency services I would definitely mention quality that can be expected from it. In fact if I would advertise any use of OSM data I would mention it
Few, if any, map data sources achieve 99.9999% accuracy
Emergency services want a high degree of accuracy, but sometimes lower is enough AND/OR best available is better than nothing
never ever
actual emergency services using OSM data at least for some purposes, despite being highly incomplete, as it was still useful are disagreeing with you
my favorite use case was for preparing flood evacuation plans as official building data was then updated at multiyear plan so latest release was mnissing
OSM had deeply incomplete building data but it was already useful to partially amend these gaps in data
Unfortunately, this misses the point entirely…
I admit that I am entirely missing your point then.
I have understood your claim as “emergency services use solely perfect dataset or almost perfect ones, and if those are unavailable there will not use best available one”
Optionally that they should not be allowed to use best available if it is not nearly perfect.
Either way I guess I should mute this thread.
Picking up on the 6 nines accuracy; that means that less than 1 defibrillator in Switzerland mapped wrong (currently we have 16 904 of them mapped).
1 wrong defibrillator of all of them would be an accuracy of 99.9941%…
Dear the_third_man, I think it’s great that you’re improving the quality of OSM. Keep up the good work.
But it seems to me that you haven’t quite grasped the aspects of crowdsourced data and how emergency services professionally use several data sources - including OSM data - yet.
And I find your tone in the messages in all AED related threads to be very negative - including the title of this thread.
The next OSM meetup near you is this coming Thursday at 6:30 p.m. (see Switzerland): If you have time, stop by so we can discuss this a bit more constructively.
Yours, Stefan
To offer another perspective:
I was visiting a few emergency control centers (Einsatzleitzentralen) in Switzerland and frequently deal with various emergency response organizations (Blaulichtorganisationen). OpenStreetMap is everywhere in these circles and often used as the primary map source. Specifically for AEDs there’s absolutely nothing comparable and the quality is constantly improving. In my opinion, defikarte.ch has played a major role in this development, as it has really pushed the available data. I also see a lot that defikarte.ch is recommended and used by many emergency responders, which speaks more for the quality of the data than against it.
And yes, OSM is not perfect and will never be, but at the moment it is the best solution overall and hopefully will get better and better. That’s what we’re all working on together every day.
Therefore, I can’t understand the criticism of @the_third_man
OSM isn’t the only imperfect map. No map is perfect (see the essays on the motif “The Map of the Empire on a Scale of 1:1” by Umberto Eco and Jorge Luis Borges).
Government-issued maps, in particular, are often outdated and therefore lack temporal accuracy. “Official” data does not guarantee “high” quality, and no one is liable for it (except in some cases of land registry). Prefer to say “fit-for-use quality” to “high quality.”
While there is heterogeneity in OSM, it is manageable and not necessarily indicative of a “quality deficiency.” Studies on citizen science and volunteered geographic information show that statistical aggregation methods and “wisdom of the crowds” algorithms can produce data of the same quality as professional sources in many areas.
The days when white men with theodolites roamed the jungle alone are over. Citizen science combines geomatics and precision with diversification and increased efficiency. The goal is not to replace the theodolite but to complement it.
https://www.openstreetmap.org/node/13909132501 here the aed got added on 04.06. via defikarte.
The missing building on osm was reported only on 09.07.
This page talks about a opening party on 27.06.
So this shows via the defikarte we have pretty much the best possible data for aed’s on osm. I don’t know how one could improve this…