Indoor Navigation: How It Works, and How to Choose a System

A system that determines where a person or a machine is inside a building and guides them from there, using something other than satellite positioning. The positioning part is the measurement. The navigation part is the routing you build on top of it, and it is only as good as the position underneath.

It depends entirely on the job. Guiding someone to a room needs metres. Landing an overlay on the correct unit among identical units needs centimetres and orientation as well as position. Decide which one you are buying before you compare vendors, because the two categories are priced and installed very differently.

Not for a camera-based system. A visual positioning system works from a prebuilt map of the space, so the only requirement is a capture of the building, and most industrial sites have been scanned at least once already. Beacon and ultra-wideband systems do need hardware: devices mounted, positions surveyed, and batteries replaced for the life of the deployment.

Positioning is determining where something is. Navigation is planning and guiding a route once you know. Vendors use the terms interchangeably, which hides the important part: a navigation product is only as accurate as the positioning method underneath it, so ask about that method rather than about the app.

Yes. Localization can run on the device itself with no network call, which keeps imagery local and gives the lowest latency. For sites where no third-party cloud call is permitted, the same software runs self-hosted on your own hardware, and in air-gapped deployments with no outbound connection at all.

If the building has already been scanned, the map is a processing job rather than a project. If it has not, a walkthrough covers roughly 1,000 sq ft per 60 seconds of capture, and larger sites are recorded in several passes and stitched into one continuous map. Compare that against a beacon rollout, where the install and the survey are the schedule.