Quantity surveyors use drone and laser scan data in three ways: to measure lengths, areas and heights from an orthomosaic or point cloud instead of walking the site with a tape and a wheel, to back progress claims and valuations with a capture dated to the claim period, and to share that evidence with the client and contractor so everyone is looking at the same picture. The data does not replace the standard method of measurement or the surveyor’s judgment. It replaces most of the site visits that checking used to need.
This post covers what each file type can give you, how accurate the numbers are, how to run captures so they hold up at a progress claim (a pay application, in US terms), and where Swyvl fits. One thing up front: Swyvl measures lengths, areas and heights, keeps the dated record and shares it. It does not calculate stockpile volumes or cut and fill.
What can a quantity surveyor measure from drone data?
A drone survey of a construction site usually produces an orthomosaic (a top-down, map-accurate image), a point cloud and one or two elevation models. A laser scan, from a terrestrial scanner or a handheld unit like an XGRIDS, produces a point cloud, often with 360 photos attached. Each answers a different kind of question.
| Deliverable | What you measure from it | Typical QS uses | Not suitable for |
|---|---|---|---|
| Orthomosaic | Plan lengths and plan areas | Hardstand and paving areas, slab footprints, site clearing, fence and curb lengths, trench runs, roof plan areas | Heights, sloped surface areas, anything under a roof or canopy |
| Point cloud (drone) | Lengths, areas and heights in 3D | Retaining wall heights, column and wall heights, as-built checks against the drawings | Fine internal detail, surfaces the camera could not see |
| Point cloud (laser scan) | Lengths, areas and heights at mm to cm level | Internal fit-out, services, facade and structural dimensions, verifying claimed work in place | Large open earthworks sites, which are slow to scan from the ground |
| DSM / DTM | Elevations and surface differences | Earthworks volumes, cut and fill, stockpile volumes | Measuring in a browser viewer; these need dedicated software |
Plan quantities come off the orthomosaic. Anything with a vertical component comes off a point cloud. Volumes come off elevation surfaces in dedicated software, and that is the one row of the table Swyvl does not cover.
How do you measure from an orthomosaic?
An orthomosaic is the fastest deliverable to measure from because it behaves like a scale drawing. Open it on a map, draw a line or a polygon, read the number. For external works that is most of the take-off: paving areas by surface type, curb lengths, fencing, line marking, the footprint of what has been poured.
Three habits keep orthomosaic measurements honest.
Measure plan areas and say so. An orthomosaic is a vertical projection. The area you trace around a pitched roof or a batter is the plan area, not the surface area. If the method of measurement wants the sloped surface, divide by the cosine of the slope or measure it in 3D. A 30 degree roof has about 15 percent more surface than its plan area.
Do not measure heights from it. There is no vertical information in an orthomosaic. Walls, columns, retaining structures and stockpile heights need the point cloud. Facades are also slightly smeared or leaning, so do not trace to an edge you cannot see from directly above.
Check the scale before you trust it. Swyvl’s measurement tools show a badge on every measurement saying whether the file is georeferenced, true scale or only indicative. A georeferenced orthomosaic from a controlled survey gives you a dimension you can act on. An image with no coordinate system gives you a sketch with a number on it.
The viewing side is covered in how to view an orthomosaic in a browser.
What can you measure from a point cloud or laser scan?
A point cloud gives you the third dimension. Click the base of a retaining wall and the top, and you have the height. Click along a wall and you have its length, with slope and bearing if you want them. Trace an area on a slab and you get the area, perimeter and edge lengths.
For a QS the practical uses are:
- Confirming heights and dimensions the contractor has claimed, without a second site visit.
- Checking as-built dimensions against the drawings when a variation is in dispute.
- Measuring internal work from a laser scan: wall lengths, ceiling heights, openings, services runs. A handheld scan of a fit-out captures in minutes what a tape takes an afternoon to do.
The main limit is that you can only measure what the sensor saw. A drone point cloud does not include the underside of a slab or the inside of a building, and a terrestrial scan has a shadow behind every object. Where the data is missing, the measurement is a guess.
Swyvl opens LAS, LAZ and E57 point clouds in the browser with the same four tools on every 3D viewer: point, distance, area and height. How to measure a point cloud in your browser covers the mechanics.
What about stockpile volumes and cut and fill?
This is the question every QS asks, so the answer needs to be plain.
Swyvl does not calculate volumes. Not stockpiles, not cut and fill, not earthworks balance. A volume needs a defined base surface and a comparison surface, and getting either wrong produces a confident number that is badly incorrect. That is a job for the software that made the data (Pix4D, Metashape, DJI Terra, DroneDeploy, Propeller and similar) or a civil package like Civil 3D or Trimble Business Center, where you set the base, choose the surfaces and see the residuals.
Those tools work from elevation models, and which model you use matters. A DSM includes the stockpile, the vegetation and the excavator parked next to it; a DTM is bare earth. Comparing the wrong pair gets the quantity wrong. DTM vs DSM vs DEM explains the difference.
What Swyvl does is keep the context around that number. You can pin a note to the stockpile with a traced outline and a height, attach the volume report, and leave it on the site at the date it was measured. The volume shape in a note is a marker, not a calculation. The cubic figure comes from the tool that computed it.
How accurate are measurements from drone and scan data?
A measurement is only as good as the capture behind it. This is the part to understand before any number goes into a claim.
Ground sample distance (GSD). This is the size of one pixel on the ground, usually 1 to 3 cm for a site flown at 60 to 120 m. You cannot measure more finely than the pixel, and in practice horizontal accuracy with good control lands at roughly one to three times the GSD, with vertical around two to three times. Those are rules of thumb. The processing report states the actual residuals, so read it.
Ground control and RTK/PPK. Without ground control points or an RTK/PPK-corrected drone, the model can be internally consistent but sit meters from its true position, and scale can drift across the site. For anything compared against a design surface, a previous capture or a survey, you want control. Ask the operator what they used and keep the answer with the data.
Laser scans. A terrestrial scanner is accurate to millimeters at the instrument, but the registered point cloud is only as good as the registration between setups. Handheld SLAM scanners trade some accuracy for speed.
Your click. In a browser viewer you are picking points, and the exact top edge of a wall is a centimeter or two of judgment either way. Zoom in, pick twice, and treat a single click as a check rather than a certificate.
The practical rule: use drone and scan measurements where the tolerance of the quantity is wider than the accuracy of the capture. Hardstand area to the nearest square meter is comfortable. A 2 mm tolerance on a bearing plate is not.
How do dated captures support progress claims?
A progress claim is an argument about what was in place at a cutoff date. The weaker the record, the longer the argument. A capture flown on or just before the cutoff turns most of it into a look at the data.
- Fly to the claim calendar. Capture within a day or two of each cutoff, with the same flight plan each time so the captures are comparable. Monthly flights are a small cost next to a disputed certificate.
- Keep the capture date on the capture. Not the upload date, not the date the report was written. In Swyvl each capture carries its own capture date, pre-filled from the photo or flight metadata and editable on upload, and captures sit on the site in date order.
- Measure the work in place. Trace the slab actually poured, the wall to the height it actually reached. Save the measurement with the file so the number, the view and the capture stay together.
- Note the exceptions. Where the claim and the data disagree, pin a note at the spot with a sketch and the question. Notes in Swyvl follow the place rather than the file: a query raised on the March capture shows on the April capture at the same spot, so you can see whether it was resolved.
- Keep every capture. A final account or a delay claim two years later is argued from the sequence of captures. The same sequence settles variations and site conditions: rock encountered, unexpected services, a redesign of the earthworks.
How do you share the evidence with clients and contractors?
A measurement is only useful if the people who need to agree with it can see it. A marked-up screenshot in a PDF is fine until someone asks “where exactly?”.
In Swyvl the project is a site, each flight or scan is a capture on it, and the whole thing is shared with one share link per job. The client, the contractor’s QS and the superintendent or owner’s representative open the link in a browser with nothing to install, see the captures in date order, and can take their own measurements on the same files. Their measurements are temporary and never change what you sent. Yours are saved as notes they can reply to after verifying their email with a code, with no account and no paid seat. The link can be password protected or limited to invited addresses, and you can see when it has been opened.
For a claim review, that changes the conversation. Instead of “we disagree with your paving area”, it becomes “open the April capture, measurement 4, and tell us where our outline is wrong”. The client can approve the delivery from the link or request changes with a comment, and the activity log records the uploads, shares and views.
A practical workflow
For a QS starting with drone or scan data on a project:
- Agree the capture plan with the operator: cadence tied to the claim cutoff, GSD, ground control or RTK, and which deliverables you need.
- Create one site per project and upload each capture against it with its capture date.
- Take off plan quantities on the orthomosaic and heights on the point cloud. Save the measurements that go into the claim.
- Run volumes in the processing or civil software, and attach the report to a note at the location.
- Send one share link to the client and contractor, and keep every capture on the record.
None of this requires a QS to become a drone pilot. It requires asking the operator the right questions, measuring from the right file type and keeping the dates straight.