Open the panel and tap AR identify.
Camera starts. No setup. No corner-alignment dance.
AR Identify
Point at the panel. Read the answer.
Vision-based auto-calibration locks onto the real deadfront in milliseconds. Labels float over the live camera, anchored to each physical breaker as you move. No corner alignment. No QR marker. No fiddling.
Point at the panel.
Vision finds the deadfront edge. The reticle locks. Takes well under a second on iPhone 13 and up.
Labels appear on the live camera.
Each breaker carries its name as a floating tag. Move closer, move farther, tilt — the tags follow.
Pop a breaker without taking your eyes off the panel.
The label tells you what just shut off. No more "the one over the dryer, I think." No more flipping every switch.
How it's different
Other "AR" panel apps make you tap four corners.
We didn't. Breaker Map uses Apple's Vision framework to find the dead-front automatically — the same kind of object recognition that runs in Photos and Document Scanner. There's no setup ritual, no marker, no alignment grid. You point, you read.
-
Vision-based detection
The deadfront is a rectangle of metal with a known grid pattern. Vision finds it the way it finds receipts and business cards.
-
Anchored, not floating
Labels stay glued to physical breakers as you walk around. Move the phone, tags stay where they belong.
-
On-device, private
Detection, tracking, label rendering — all happen in the Neural Engine. Nothing leaves the phone.
-
Works in a dim utility room
Vision tolerates low light better than feature-point tracking. A flashlight tucked under your arm is enough.
Common questions
What devices does AR work on?
iPhone 13 and newer, iPad with M-series chip. The Vision framework calibration runs on the Neural Engine.
Does it need internet?
No. The AR detection and tracking are 100% on-device. The first time you enable AR your account is checked against the server (for the 7-day trial timer), but identifying any panel after that is offline.
Does my camera feed leave the phone?
Never. Camera frames are processed on-device by the Vision framework. We don't store, transmit, or analyze them off the phone.
How accurate is it?
In our testing, label placement is within a single breaker slot at any reasonable distance (1–4 feet). In bad lighting, the reticle holds longer before locking. Holding the phone steady for a second always nails it.