I have a dog park with tools like
how can we tag it?
Asked ChatGpt that suggested
leisure=dog_park
dog_exercise=agility
more specifically, for individual features, you can use:
leisure=dog_exercise
dog_exercise:equipment=bridge
Alternative values for dog_exercise:equipment include:
ramphooptunneljumpweave_polesplatformDid you only ask Chatgpt or check the wiki and taginfo as well?
Do you like the AI proposed tagging? Why / why not?
You are not looking for a dog park, but for amenity=animal_training.
There’s currently no tagging for the individual exercises, but your suggested :equipment namespace makes sense, because animal_trainnig=* is already taken for the type of animals being trained. But if you want to end up with a top-level-usable tag, like playground=*, then you will want to put some work into fleshing this out, as it will probably have to work for all kinds of animal training.
Asked ChatGpt that suggested
leisure=dog_park dog_exercise=agilitymore specifically, for individual features, you can use:
chatgpt and the like are typically not proposing good tags, they look ok but are often not the ones we use. For example dog_exercise has 0 uses, it is a completely new tag: https://taginfo.openstreetmap.org/search?q=dog_exercise
this would only be a good suggestion if nobody ever had tagged such a property.
But both =dog_park and =animal_training can have the same low-difficulty equipment? One for playing. One for training. A =dog_park doesn’t suddenly become an =animal_training when you put these inside.
The *:equipment= suffix is unsuitable when the feature is considered. If it’s man_made=animal_training_equipment , it would simply add animal_training_equipment=
I think that you can train dog in dog parks with training equipment.
It is still a dog park if dog owners can go there and both train their own dog and let the dogs play without the equipment. We could add animal_training=yes to such dog_parks.
amenity=animal training is usually more organized training with a group of dogs. Usually require some kind of club membership and not open to the public.
Many animal training pitches are incorrectly tagged as dog_park. That means that when you are travelling with a dog, you check the map and find a dog park that only takes a 10 km detour, only to find out a locked gate and a sign saying it is for the local German Shepherd club that meet Tu,Th at 17:00.
This is just a normal dog_park
Did you only ask Chatgpt or check the wiki and taginfo as well?
Guess why I asked this question ;-) I found no good information source plus that I am not an expert on dog_parks…
Do you like the AI proposed tagging? Why / why not?
For me its ok… as said I am not a dog_park specilist…
I see no one else tagging agility on dog parks…
Why? Looks like OSM is a good place to store informatlon like that…
This is what I would have guessed 
I just wasnt sure.
Next time, if you dont find anything on taginfo and wiki, write this in your post. This is valuable information. It also shows you tried to solve it on yout own but it looks like it is unsolvable and needs discussion. Otherwise your post could be interpreted as “I am lazy and just post something out of chatgpt to the forum and want others to do the brainwork instead of me.”
Just entering ‘agility’ in Taginfo gives
dog:discipline:agility and several others Search results | OpenStreetMap Taginfo
Often in combination with
amenity = animal_training
animal = school
animal_training = dog
In contrast the suggested dog_exercise is not used at all.
Otherwise your post could be interpreted as “I am lazy and just post something out of chatgpt to the forum and want others to do the brainwork instead of me.”
I sometimes feel that our OSM discussions focus too much on tags, and too little on use cases — the why behind the data. As I personally lack experience with dogs, that’s exactly why I think such a discussion is valuable — to understand the real-world situations people map for, not just which tags exist.
If we better understand the users and their needs, we can tag more meaningfully and make OSM more useful for real people.
Recently, when talking with people about dog parks, I noticed how differently people use them:
dog_exercise=agilitydog_exercise:equipment=bridge, ...=tunnel, etc.fence=*, dog_size=small, that we have more asreas with a fence next to each othersurface=*, bench=yes, path=gravel or paved.dog_park and playground might matter.surface=* or maybe even seasonal notes.dogs=leashed, dogs=designated, fee=yes/no, opening_hours=*.By thinking from these user perspectives first, we can have a richer discussion about what data OSM should capture — and then decide what tags best describe it.
OSM is strongest when we tag for mappers and for users, not just for the database.
Frankly I do not understand why you focus on a topic which you do not understand or of which you do not have sufficient experience and knowledge what it is about. My suggestion would be to map things where you are aware how they work, how they are used, what are the characteristics and differences, etc. The solution IMHO, if you are interested in the topic but do not have the knowledge yet, would be to research it and then look in OpenStreetMap which tags there are, and come up with a concept what to change or add (if any).
The list of user groups that you sketch as potentially interested in this data, can be a good start where to go and look for information, but ultimately it will be up to a specific map provider to decide how to cater for them (for whom to cater) and which information to present. Some of it may already be implicit in the data (e.g. a dog park with a playground nearby does not have to be tagged explicitly, it is already spatially contained in the database) and just waiting for the right query to be extracted from the db.
Thanks for your comment, but I don’t agree that one needs to be an expert to contribute here — it’s not rocket science. OpenStreetMap grows through people exploring, testing, and improving data by mapping real situations.
My approach is to start from use cases — understanding how different people actually use a place — before jumping into tagging details. That’s often the most effective way to ensure our data serves real users, not just the database.
Take a dog park, for example: it’s a clearly defined area with a fence, a specific fence type, and at least one gate. From there, we can build on practical aspects — benches, shade, surfaces, or proximity to a playground — all things that reflect real-world needs.
Constructive feedback is always welcome, but discouraging contributions because someone is still learning misses the collaborative spirit of OSM. And if no one has yet documented the agility aspect of a dog park, that’s more an indication of an opportunity to improve the tagging, not a reason to stop discussing or experimenting with it.
yes, and on the other hand, it bears the risk that you cater just for a specific usecase, and produce data that is less universally usable. All approaches have pros and cons. Traditionally, we try to be generic in our tagging, hoping to enable many different use cases, even not yet developped ones…
Maybe what we need is some generic tagging of agility
please add
dog_exercise=agilitydog_exercise:equipment=bridge, ...=tunnel, etc.fence=*, dog_size=small, that we have more asreas with a fence next to each othersurface=*, bench=yes, path=gravel or paved.dog_park and playground might matter.surface=* or maybe even seasonal notes.dogs=leashed, dogs=designated, fee=yes/no, opening_hours=*.Not all dog parks have a fence. And some have a fence, but no gate, just an opening.
related discussion " Mapping equipment of dog parks"