
A robot that cannot say where it is cannot act. Not roughly which room, but the exact position and orientation, in a frame something else can also read.




| Single-capture approach | Multi-capture approach | |
|---|---|---|
| How a site is covered | One scan, one map, one instrument | Many scans joined into one coordinate frame |
| Past one scan's range | Coverage stops | Add a map and it aligns against the set |
| Instrument choice | Whatever the pipeline accepts | Aerial for open areas, handheld for interiors, survey data where it exists |
| When an area changes | Re-capture the whole map | Re-scan the changed area into the same frame |
| Where the frame comes from | The vendor's pipeline | Your plant grid, georeferenced to WGS 84 |
Stated plainly: a refinery is never one scan. It is dozens, taken with different instruments over months, and the question is whether they compose into one frame or stay separate maps that each stop at their own edge.
Two related explainers: GPS-denied navigation covers what happens when the satellite fix degrades rather than disappears, and E57 to VPS covers how a scan you already hold becomes a localization map.