OpenStreetMap logo OpenStreetMap

I’ve recently been taking on more tagging for wheelchair accessibility in my neighbourhood. It’s been one of those “you don’t see it until you think to look for it” kinds of things. I’m not disabled myself, so I never thought to consider the heights of steps (until I tripped on them.) Now that I look for them, I must freak out other people around me as they see me stare at my foot to see whether a step is ankle high or not.

There were a couple things I came to think about while mapping that I could not seem to find answers for in the wiki. A lot of the time, I just answer the question with the framework of “I’ve seen places do these things, so I will treat them as the standard and anything that falls under that bar is limited or not accessible.” I would be curious for others thoughts on them.

  • Do establishments need powered doors to be accessible?

Here in Regina (and a lot of Canada, these were extremely common back in Ontario), doors have buttons that you can press to swing the door open mechanically. They have a wheelchair icon on them most of the time, so they are clearly marked as an accessibility feature.

While common, they are not ubiquitous. In my mapping practice, if a place does not have at least one powered door, I label it as limited wheelchair accessibility at maximum. My logic is that even if a door could be physically be opened manually by a patron in a wheelchair (as in, they can open the door and have enough space on the platform outside to maneuver while doing it – I have seen places where the platform is elevated and small enough that you would fall off trying to do so), they shouldn’t have to do that while the option of powered doors exists.

There are also problems where these exterior doors use door closers as a fire prevention feature – but those closers are set up improperly and close too fast. So if you’re trying to open a door manually while trying to enter with a wheelchair, you have a door slamming into you and pushing you into the frame (which I have seen happen a few times.)

  • Do we evaluate accessibility at the POI feature level, or in context?

I am running into an issue where I am mapping out the accessibility of a local mall. The individual stores are accessible on their own: they have wide aisles, there are no steps, and everything is at an appropriate height. But the mall itself has two levels.

For the longest time, I assumed there would be an elevator or some way you could access the second level in a wheelchair. But after actively looking, I couldn’t find one. Technically, there are pedways leading into the mall from other buildings and those buildings have elevators. But I feel like “to get to the second level, exit the building and enter through another building after using their elevator” is not a reasonable accessibility plan.

Still though, the stores themselves on the second level are accessible. So do we label them accessible because the features are accessible, or do we label them unaccessible because there is no reasonable way for a wheelchair user to get to them?

My take has been that they are not accessible, even if the stores themselves are. If someone in a wheelchair entered the mall and saw a map that showed the stores as accessible, then they would assume that they could actually access them. If they learned that they are accessible after they leave the building and get back inside through a maze, I would say it is reasonable for them to feel like they got lied to. It is not useful information to say “after you manage to find a way in there, you’ll have an easy time of shopping inside.”

Location: Capital Crossing, Regina, Saskatchewan, Canada
Email icon Bluesky Icon Facebook Icon LinkedIn Icon Mastodon Icon Telegram Icon X Icon

Discussion

Comment from crazycolbster on 26 May 2025 at 03:21

I’ve had similar questions when mapping out the sidewalk networks of the endless suburban sprawl covering the US. Technically, pedestrians are allowed to cross at any intersection, but T-intersections are usually designed to be accessible in the normal direction, but not in the parallel branch. I’ll sometimes put the crossing at a slight angle so it lands on a driveway instead of a curb with grass, but that feels “hacky” at best.

Comment from Malle_Yeno on 26 May 2025 at 15:16

I’ve seen that sort of situation too! Our local map hasn’t gotten to mapping sidewalks as separate from road features at large (though that is something that I have been strongly considering as of late – it would make mapping differences in sidewalk surface quality versus road surface quality more straightforward). Guess I’ll have to be on the lookout for t intersections when I get started with it.

Comment from Glassman on 1 June 2025 at 21:04

It is a shame that the mall doesn’t have elevators for people with mobility issues.

I’ve been mapping sidewalks as separate ways for some time. I got started after working with the University of Washington Taskar Center Accessible Technology. The Taskar center even has a great map tool, AccessMap.app. It likely because the Taskar Center is located in Seattle that they created such a great map tool. Seattle is a very hilly city. So AccessMap.app will even direct wheelchair users to elevators during opening hours.

If you get Regina sidewalks map, let the Taskar Center know so they can add your city to their map.

OSM US is also has a Pedestrian Working Group. We hope to have a sidewalk map done in the near future.

Comment from Pieter Vander Vennet on 9 June 2025 at 17:01

In my opinion, adding wheelchair=yes/no/limited is not very useful. It gives some indication, but actually measuring is more valuable - especially cause there is a huge difference in the range of abilities of wheelchair users (e.g. a young, strong person in a wheelchair who has the arm power to get over e.g. 10cm high kerbs vs an old person in an electric scooter, for which even 5cm is an insurmountable problem).

With https://mapcomplete.org/onwheels, I’ve tackled this problem by adding the individual (indoor) doors; it is up to the data consumer to figure out afterwards what doors are suitable to them and which are not.

Comment from Malle_Yeno on 9 June 2025 at 17:41

@Glassman Will do! That’s a cool project. Regina isn’t particularly hilly (that’s the prairies for ya haha) so there probably are fewer elevation challenges for us. But I still think there’s value in separately mapped sidewalks so that surface quality can be reported separately for a road versus sidewalk. That is a challenge for us because our soil here makes features shift rapidly, so you can see a difference in maintenance between roads and sidewalks.

@Pieter Vander Vennet That’s an interesting point. I have been doing more wheelchair:description:en on my features to describe what’s happening. Would be interesting to see this question pop up on SCEE so I can do descriptions in the field (so far I either save it for when im home and go off memory, or use show tags and do it live). That’s an interesting solution. I didn’t consider using indoor mapping and treating the doors themselves as nodes with accessibility features. I might have to think about how I’d go about making an indoor map of the mall I’m thinking of!

Comment from UW Amy Bordenave on 9 June 2025 at 19:01

Amy here, from TCAT @ UW and the OpenSidewalks project:

Thanks for the shout-out, @Glassman!

@Malle_Yeno, let me know if there’s any way I can be of assistance!

The #sidewalks channel in the OSM US Slack is a great place to ask any questions that come up and to get feedback from the community.

Happy mapping!

Comment from Pieter Vander Vennet on 10 June 2025 at 15:47

@Malle_Yeno,

Actually, I see accessibility as a routing problem. If one maps all the indoor corridors, the elevators, the stairs, the doors, … with their respective properties, a user can ask a (still to be created) routeplanner to give them an accessible route. This way, they can get a route (or “no route possible”) for their specific usecase; could also be reused for moms with strollers, people with bicycles in e.g. train stations, …

Comment from Pieter Vander Vennet on 10 June 2025 at 15:47

I’ve also did quite some research and tagging work on accessible wheelchair toilets: osm.wiki/Wheelchair_accessible_toilets

Comment from Glassman on 10 June 2025 at 16:13

@Pieter Vander Vennet Check out AccessMap.app. It routes via elevators inside of buildings. For example, in downtown Seattle, where the streets are steep, as well as at light rail stations.

Comment from Malle_Yeno on 11 June 2025 at 20:25

@UW Amy Bordenave Thanks for the offer! I will reinstall slack and join up with y’all!

@Pieter Vander Vennet As soon as I posted my response, I was like “wait how would a description help an application with deciding what door to use” and noticed that I misunderstood where you were going with that haha

I like the approach of viewing accessibility as a routing problem. I had viewed it more from a desireability lens (ex. making the decision of destination based on available accommodation.) So I had approached tagging based on what information would help someone make that decision with as much information as possible. But viewing it from a routing pov makes more sense because it allows a user to pick an option that best suits their needs, regardless of what their destination is. After all, maybe someone really needs to get to a shop on the second floor and feels it’s worth it to go through a different building – and in that case, just saying “sorry second floor doesnt meet accessibility requirements” isn’t a helpful answer to them. The user can make the call based on what route is calculated as meeting their needs, instead of making the call on the data end. I think that makes sense.

I’ve been mapping the indoors of a local library to get my feet wet on indoor mapping and using some accessibility tags (not gonna try the mall for now cuz it’s much too big for my comfort at the moment). Once I have a complete version of the building, I might post a diary about it and ask your thoughts on it!

Log in to leave a comment