How to test a room scan app's accuracy yourself

What you need

• A laser distance meter. Write down its make, model and the accuracy printed in its manual. It is your reference, so it has to be better than what you are testing.
• A steel tape for doors, windows and anything short.
• A stopwatch, the one on your phone is fine, if you also want to compare time.
• The app on a LiDAR device: iPhone 12 Pro or a newer Pro model, or an iPad Pro from 2020 or later for PlaceAnew.
• A sheet with one row per measurement: room, item, tape value, app value.

Pick rooms that can fail

An easy rectangular bedroom proves little. Include the rooms where scans are known to struggle:

• A small bathroom, tight to walk and full of fixtures.
• A kitchen, where cabinets hide the walls.
• A long room or hallway. In one study, error grew with length, and features beyond about 13 feet (4 m) could be off by more than 25 cm in a single fast pass.^[1]
• A room with a large mirror or glass wall. Apple says full-height mirrors and glass are a challenge for the LiDAR sensor.^[2]
• An ordinary bedroom as the baseline.

Measure first, then scan

1. Measure each wall corner to corner at about chest height, between the wall faces, above the baseboard.
2. Measure the ceiling height in the middle of the room.
3. Measure each door and window opening: width and height of the opening, jamb to jamb, the same way every time.
4. Write every value down before you scan, so the app's numbers cannot steer you.
5. Scan the rooms the way you would on a real job, at a normal walking pace. Keep the first scan of each room. If a scan fails or you have to restart, write that down too.
6. Read the app's values for the same walls, heights and openings from its plan or export.

Do the math

For each pair, the error is the app value minus the tape value. Keep the sign: a plus means the app read long. Then take the absolute error, drop the sign, and look at three numbers across everything you measured: the median, the worst, and how many errors are larger than what your work can absorb.

An example with made-up numbers. Absolute errors 1, 1, 1, 2 and 4 cm: the median is 1 cm, the worst 4 cm.

Item	Laser meter	App	Error
Bedroom wall A	3.60 m	3.62 m	+2 cm
Bedroom wall B	2.95 m	2.94 m	−1 cm
Hallway	4.80 m	4.76 m	−4 cm
Ceiling height	2.44 m	2.43 m	−1 cm
Door width	0.91 m	0.92 m	+1 cm

Percent error helps compare short and long items: 4 cm on a 4.80 m hallway is 0.8%. For room area, multiply the measured length by the width of a rectangular room and compare it with the app's area for that room.

Most laser meters show feet and inches or meters. Pick one and use it for every value.

Read the results fairly

• Judge by the median and the worst, not the best. One perfect bedroom says nothing about the hallway.
• Count failed scans. A scan that had to be redone is part of the result; dropping it flatters the app.
• Measure the same thing. If you measure above the baseboard and the app measures wall to wall at the floor, the difference is method, not error. Compare inside faces: RoomPlan-based apps often set wall thickness to a default,^[3] so a wall's thickness is not a fair test of the scan.
• Count what the app owns up to. PlaceAnew draws a wall it could not see dashed and marks anything it estimated "estimate", so a hidden wall is never passed off as measured.
• Decide what matters for your work. Layout, design and quotes are where a scan saves the most time; cabinets and stone are ordered to fractions of an inch. See Is LiDAR accurate enough for remodeling?

What published tests found

Abdel-Majeed and others, HBRC Journal 2024; Ersoz and Bosché, ISARC 2026; PlaceAnew office check, 2026.^[1][3][4]

Test	Reference	Result
iPad Pro LiDAR with a scanning app, one hall	Trimble laser scanner	Within 1-2 cm up to about 3 m; more than 25 cm possible beyond 4 m in a fast pass
RoomPlan app on iPad Pro, a house and two office floors	Leica laser scanner, modelled by hand	Wall area within 0.1% to 1.9%; room area 4.9% to 6.3% small; every room found
PlaceAnew on iPhone Pro, our own office	A hand-measured floor plan	All 7 wall lengths that compare one to one within 2.5 inches, median under 1 inch

Your own rooms will tell you more than either study, because they are your rooms, your device and your way of walking.

Time it too

Use the same person and the same rooms for both methods. Start the clock at the first measurement and stop it when the plan is drawn, so the hours spent redrawing a hand sketch are counted. For published times, read Measuring a house by hand vs scanning.

Questions people ask

How many rooms should I test?

At least five, of different kinds. More rooms give a steadier median, and the long and awkward rooms are the ones that teach you most.

Why the median and not the average?

One bad wall pulls an average a long way. The median shows the typical error; the worst value shows the risk. Report both.

Can I test with a tape instead of a laser meter?

Yes, if you are careful: keep it straight and level, and have someone hold the far end on long walls. A laser meter is quicker and easier to repeat.

Is this how PlaceAnew tests itself?

It is the same method we wrote for our own measurement study. Our first result, from our own office, is in How accurate is PlaceAnew? If you run it with PlaceAnew, we would like to see your sheet: contact us.