What You Actually Get From a Construction Drone Survey

Diagram of stacked mapping deliverable layers

If you have never commissioned a drone survey before, the quote you receive can be hard to evaluate. Two vendors quote the same site, one is half the price, and the line items read like a foreign language. This is a plain explanation of what actually lands in your inbox after a flight, and what each output is good for.

The orthomosaic

An orthomosaic is the output most people picture when they think of drone mapping: a single, seamless image of the whole site viewed straight down. What separates it from an aerial photograph is that it has been orthorectified, meaning the distortion caused by camera lens and terrain relief has been corrected so that scale is consistent across the image.

That correction is what makes it measurable. You can pull a distance between two points anywhere on an orthomosaic and trust it, which you cannot do on a raw photo. It usually ships as a GeoTIFF, which carries its coordinate information inside the file, so it drops into your GIS or CAD environment already positioned.

Practical uses: verifying as-built conditions against plan, documenting site conditions on a given date, measuring areas for takeoff, and giving everyone in a meeting the same view of the site.

Surface and terrain models

These two get confused constantly, and the difference matters when you are paying for one.

A digital surface model, or DSM, records the elevation of whatever the drone saw from above. That includes stockpiles, equipment, trailers, trees, and buildings. A digital terrain model, or DTM, is the bare earth with those objects removed through classification and filtering.

If you want to know the shape of the ground for grading or drainage, you want the DTM. If you want to measure a stockpile or check clearances, the DSM is what holds that information. Filtering to bare earth is interpretive work, and on a heavily vegetated or congested site it is where quality varies most between providers.

The point cloud

A point cloud is the raw three-dimensional measurement behind everything else: millions of individual points, each with a coordinate and usually a color sampled from the imagery. Photogrammetric point clouds are generated by matching features across overlapping images and solving for their positions in space.

Most clients never open the point cloud directly, but your engineer or surveyor might, because it is the densest and least processed form of the data. It typically ships as LAS or LAZ. If a deliverable is ever disputed, the point cloud is what you go back to.

Contours and linework

Contours are derived from the terrain model, drawn at whatever interval suits the work. One foot is common for detailed grading; larger intervals suit broader planning.

The thing worth understanding is that contours are a derived product, not a measurement. Their quality is entirely inherited from the terrain model underneath. Smooth, attractive contours generated from a poorly filtered surface will look convincing and be wrong. Exported as DXF or SHP, they drop straight into Civil 3D or equivalent.

Volume and cut/fill reporting

This is where drone data most often pays for itself outright. Volumes are calculated by comparing a measured surface against a reference: a base plane under a stockpile, a previous flight, or a design surface.

Comparing against a design surface tells you how much material still needs to move. Comparing against a previous flight tells you how much moved since last time, which is the basis for progress payments and for catching a quantity dispute while it is still small. A stockpile inventory that takes a crew most of a day with a rover takes one flight.

Reports usually come as PDF for the record and CSV so your team can work with the numbers.

What to ask before you sign

Three questions separate a serious quote from an optimistic one.

What accuracy will you achieve, and how will you demonstrate it? A number in a proposal is a promise. A number in a delivered QC report, checked against control points that were not used in processing, is evidence.

What ground sample distance are you flying at? This determines how much detail the imagery resolves, and it is set by flight altitude and sensor. It is a fair question and the answer should be immediate.

What formats will I receive? Data you cannot open is not a deliverable. Confirm the file types match what your engineering software actually reads before the flight, not after.

The short version

The orthomosaic is what you look at. The terrain model is what you measure the ground with. The point cloud is the underlying evidence. Contours and volumes are the products your team acts on. A quote that does not distinguish between them is worth asking questions about.