Door sensors as building data
Every automatic door carries a sensor. DoorData turns what those sensors see at the threshold into building intelligence: people flow, occupancy, security replay and machine safety. No cameras. No identities. Only the geometry of motion.
Intelligence
The same sensing, read seven ways. Six cards open the live prototype on that view; the seventh opens the factory scene. Everything is simulated.
People as anonymous sensor tracks at a sliding door, a machine cell and a delivery-robot lane, with three-second prediction.
open the flow lab →A badge-less follower within 1.8 s of a credentialed entry is flagged and replayed as two tracks and a door state. No images, no identities.
open the replay →The campus map: live crossings at twelve doors, queues and dispersal as heat, forecast for tomorrow, and the next best actions.
open the overview →The building drawn room by room, coloured by what the door counts say, with a clean-now recommendation from thirty days of usage.
open smart cleaning →Per-door cycles, close latency and hold-open time, trended into a service forecast before the door fails, plus the energy lost while it stands open.
open door health →D01 Main entrance opened up: sensor head, belt, motor, rollers, leaf edges and floor guide, each tinted by the cycles it has run since the last service, with a hold-to-stress-test what-if.
open the cutaway →Four machines with warning and safety fields on one floor: an operator, a loader, a setter and a technician walk into them, a muted AGV passes, and the stop log and availability follow.
open the factory scene →What a door can tell you
The sensors that open and protect a door already measure who approaches, who crosses and how the leaf moves. Read as data, each doorway becomes a witness that never records a face.
Directional crossings per door become live headcounts, room availability, popular times and capacity alerts.
A badge-less follower within 1.8 s of a credentialed entry is flagged, and the two tracks are replayed as anonymous sensor tracks for review.
Cycle counts, close latency and hold-open time predict service and put a number on the air that leaves through an open door.
The same sensing slows a machine cell when a track enters its warning zone and stops it in the safety zone, with every event logged. See the factory scene.
Three prototypes, all simulated
All three run entirely in the browser on generated data. Nothing you see was recorded from real people.
A control room for a campus of twelve doors: the flow lab with sensor-tracked people, security replay, overview map, rules and alerts, rooms and desks, energy, queues, portfolio and threshold evidence.
People as anonymous sensor returns at a sliding door and around a machine cell: activation square, safety strip, warning and safety zones, with machine stop alerts and cases.
Machine safety at scale: four machines with warning and safety fields, an operator, a loader, a setter and a technician who walk into them, a muted AGV, and the stop log and KPIs that result. Drag to orbit.