Where people actually went
Recorded walks.
A person walks the library carrying a phone. The phone records its own path and broadcasts an identity frame. The radios around the floor hear that frame and, from signal strength alone, work out where the person is, the way you can guess which room a voice comes from. Both paths are drawn on the surveyed plan, and the distance between them is the reading. Spoiler: the voice trick is harder than it sounds.
One walk has been placed on a surveyed floor. The fleet heard it.
| Walk | Floor | Length | Duration | Fit residual | Fleet | Typical miss | 9 in 10 within |
|---|---|---|---|---|---|---|---|
| 2026-08-28 | fiit-ground-0 | 302 m | 25 min | 0.26 m | heard it on 9 radios | 2.48 m | 7.96 m |
The recorded path is the phone's own tracking, fitted onto this plan on the surveyed radios the walker stopped at. It is a fit with a measured residual, not ground truth, and the residual is printed with every walk.
Every miss above was computed with a propagation model fitted on that same walk, so these metres are the best the room allows rather than what the method would do on a floor it had never seen. Read the size of the gap, not its decimal.