How our scan measurements compared with a tape measure

A remote measurement is only useful if you know how far to trust it. So during development we took captures of real objects, measured them in the measurement workspace, and compared the result with a tape measure. The tape value was used only to check the result, never as an input. Nobody typed in a reference length; the scale came from the capture alone.

The design target is about ±1.5 cm over spans of around 2 m (roughly ±⅝ in over 6½ ft). The checks below are individual cases, not a statistical study, and we report them as they are.

• A 50.0 cm item (about 19.7 in) measured 49.4 cm (about 19.4 in) on the scan. That is 1.2 % short, about a quarter of an inch.

• Several measurements on our staging setup came within about 2 mm (under ⅛ in) of the tape.

• Captures of a TV screen came in between 1 % and 4.6 % short. A TV is a useful test object because it is a flat, sharp-edged rectangle with a known size.

So the results ranged from very close to clearly off. On a span of about 6½ ft, 1.2 % is close to an inch and 4.6 % is more than 3½ in, which is outside the target. The interesting part is why.

When a scan length differs from the tape, it usually comes from one of two places, and they behave differently.

Scale error

The capture determines the scale of the whole scene. If that scale is slightly off, every length in the scan is off by the same percentage. A 1 % scale error is about 0.2 in on a 20 in item and about 0.8 in on an 80 in door. The TV results, all short by a percentage, look like scale error. This is why the sweep in the guided capture matters, and why a slow, steady sweep is worth asking for.

Marking error

This is about where you select the endpoints. A point away from the intended edge changes the measured distance. Use clear edges, zoom in where needed and complete the photo checks. Do not assume that every point-selection error changes the result by the same fixed amount.

Splitting errors this way is practical. If several lengths in one scan are all off by about the same percentage, look at the capture. If one length is off and the others are fine, look at the points.

Each measurement includes an estimated uncertainty, shown with a ± value. Review the selected points and complete any photo checks before using a result in a preliminary estimate. Low-confidence scan results are not charged. The displayed uncertainty is not a guaranteed maximum error; a small value alone does not prove accuracy.

• Use scan measurements for preliminary estimates: deciding standard vs custom, pricing a range, planning the visit.

• Treat any number close to a size boundary with care, and note it as an open item.

• Ask for a good capture; it reduces scale error. See capture tips.

• Verify final dimensions on site before ordering or installing. See pre-survey vs final measurement.

A tape in a skilled hand is still the tool for ordering. A scan is the tool for not driving out to every inquiry. Used that way, the two work well together.

Try a scan of an object whose dimensions you can check with a tape measure. Similar percentage differences across several measurements may indicate a scale issue; review the capture before drawing a conclusion. If one result differs, inspect its endpoints and photo checks. Compare more than one dimension, and confirm final sizes on site.

Plan the whole inquiry with the remote site survey workflow and preliminary estimate checklist. For measurement methods and their limits, read measuring from a photo.