How Accurate Is a LiDAR Scan?

A smartphone as a measuring instrument invites justified scepticism — especially when the result is to be submitted to a bank. This page explains how the measurement works, what its accuracy depends on and where the limits of the method lie.

How LiDAR measures

LiDAR stands for "Light Detection and Ranging". The sensor emits infrared light pulses and measures how long they take to return. Time of flight and direction give a distance for each measurement point — repeated thousands of times, this builds a three-dimensional point cloud of the room.

Unlike photographing, then, an actual measurement is taken rather than being extrapolated from images. That is the decisive difference from methods based on photographs alone.

Accuracy in figures

Apple does not state a guaranteed measurement tolerance for the LiDAR sensor, referring generally to high-accuracy depth information. For its own Measure app, Apple recommends a capture distance of roughly 0.5 to 3 metres from clearly identifiable objects. Independent studies therefore provide the orientation:

Our assessment: under suitable capture conditions, accuracies in the region of ±1 cm are achievable with the LiDAR scanner in current iPhone Pro models. The accuracy actually attained depends on room size, surfaces, capture distance, scanning technique and the software used.

From our own project experience, the deviation in the resulting area, when capturing floor by floor or dwelling by dwelling, is below one per cent overall. This figure comes from our practice, not from the studies cited above.

The wording is deliberately "achievable" and not "guaranteed": capturing a whole room adds movement, position tracking and the merging of individual measurements to the pure sensor reading. Each of these stages can introduce further deviation. A blanket assurance of "centimetre-accurate" would therefore not be sound.

What affects the accuracy

The spread between a very good and an unusable scan is wide — and depends less on the device than on the capture:

• Scanning technique. Steady, slow tracing of all wall surfaces yields far better data than rapid panning.
• Completeness. Corners, recesses and areas behind furniture have to be captured, otherwise they are missing from the analysis.
• Surfaces. Glass, mirrors and highly reflective surfaces return the light pulses unfavourably and can cause erroneous readings.
• Furnishing. Cluttered rooms conceal precisely the wall edges that matter for the area calculation.
• Room size. Over longer distances small deviations can accumulate; very large rooms are more demanding than typical living rooms.
• Lighting. Very bright daylight can impair the infrared measurement.

This is why the mkm|Scan app guides the capture and flags gaps — but care while scanning remains the greatest lever for a good result.

Why a person still does the calculating

A scan is raw data, not a finished document. A point cloud does not reveal by itself where the load-bearing wall ends and the lining begins, whether a room qualifies as habitable under building law, or what proportion of a balcony is to be counted.

Yet it is precisely these decisions that determine the result of a living space calculation. At our office, a member of the team therefore processes the scan, produces the floor plan and checks the calculation before it leaves the building. On request, confirmation by a certified surveyor is added.

Which rules apply is explained on the page Living space calculation under the WoFlV.

When an on-site survey is the better choice

Both routes in detail: 3D scan via app and on-site survey.

Frequently asked questions