Knowing that a restroom exists is not enough.

Relief ATL starts with public data from a Georgia State University study of restroom accessibility in Atlanta. Researchers physically visited candidate locations instead of relying on map listings: and what they found is the reason this app exists.

The study

262: potential restroom locations audited
15: areas of Atlanta
Feb-Apr 2025: field work window
117: locations had an accessible restroom when visited

Of the 262 locations, 145 were inaccessible: no public restroom at all, "customers only" signage, locked doors, or occupied for more than ten minutes. That is the gap between a restroom appearing on a map and a person being able to use it.

Three different questions

Existence

A dataset says a restroom is here. That is all an ordinary map can tell you.

Verified accessibility

Somebody physically went there and confirmed the restroom could be used. The 117 GSU locations in Relief ATL all clear this bar: as of the 2025 field visits.

Current availability

Somebody confirmed recently that it is accessible right now. This is the only one that helps when you urgently need a restroom: and it is the one that goes stale fastest.

Static data goes stale

A restroom that was accessible in April 2025 can be locked today. Relief ATL layers recent anonymous community confirmations on top of the public datasets so you can see whether a restroom is likely to be available now. Every confirmation is stored as its own timestamped event in Tiger Data: nothing is ever overwritten, so each location accumulates a real availability history:

9:12 AM   ✅ Accessible
10:44 AM  ✅ Accessible
1:03 PM   ❌ Locked
1:29 PM   ❌ Locked
3:18 PM   ✅ Accessible

That history is what lets the app decide whether older information should still be trusted. A 2025 audit is good evidence: right up until two people report a locked door in the last twenty minutes.

How access confidence works

Each restroom gets a 0-100 score answering one question: how confident are we that someone can use this restroom right now? It is a deterministic calculation: no machine learning, no guessing: and you can open any restroom's detail view to see the exact factors that produced its number.

Confirmed accessible : someone reported successfully using it within the last few hours.
Likely available : physically audited as usable, but not confirmed lately.
Not recently verified : listed somewhere, never verified on the ground, or known to be restricted.
Reported unavailable : recent reports say people could not get in. This overrides older evidence.

Green is never awarded just because a restroom appears in a dataset. It has to be earned by recent evidence.

Where each restroom comes from

Georgia State University restroom audit

117

Locations where researchers found and used an accessible restroom between February and April 2025. Each record keeps its original audit fields: access restrictions, ADA stalls, soap, toilet paper, changing tables, posted hours and the auditors' own notes.

OpenStreetMap

Locations tagged amenity=toilets across metro Atlanta. These extend coverage beyond the 15 studied areas, but nobody has physically verified them: so they stay grey until a community report arrives. Where an OpenStreetMap entry overlaps an audited location, the physical audit wins and the map listing is kept as a cross-reference.

Community-derived information is never presented as official city data. Planned additions: MARTA, City of Atlanta parks and recreation, Atlanta public libraries, dedicated public facilities, Throne, and community-submitted locations.

Where the community reports come from

Every report in Relief ATL was submitted by a real person tapping I used it or Couldn't access after actually going there. There is no sample data, no generated history and no placeholder confirmations anywhere in the app.

That means a location starts out showing only what the public datasets can prove, and has to earn a green pin from someone who was physically there. Early on you will see a lot of amber and grey. That is the honest picture, and it is the point: a confirmation only means something if it was never given away for free.

Read the source

Public bathrooms as public goods: Assessing availability and accessibility in Atlanta, Georgia

PLOS Water · 24 June 2026

Read the paper

Full public dataset on OSF

207 individual restroom audits across the 117 accessible locations.

osf.io/fm9by

System

Restroom locations: 198
Report events: 1
Reports (24h): 0
Accessible (24h): 0
Problems (24h): 0
Event store: Tiger Data

No accounts, no sign-in, no tracking. Reports are anonymous: the app stores the restroom, the status and the timestamp, and nothing about you.

Back to the map