How Drone Surveying Is Used in Mining
Drone surveying has become the default way an open pit or quarry measures itself, because it produces a complete surface of the site in an afternoon rather than a set of points over a week. What it does not do is remove the survey discipline around it: the flight decides resolution, the control decides accuracy, and the deliverable you choose decides whether the volume you compute is right. This guide covers how the workflow fits together on a mine site.
Why mines adopted it first
A pit changes shape continuously and has to be measured on a fixed reporting cycle. Traditional survey gives excellent accuracy at a limited number of points; a drone gives a complete surface, which is what a volume calculation actually needs.
Safety is the other half. Measuring a face, a highwall or an active dump from the air removes the need to put a person near it — the same reason repeat monitoring flights are now routine on walls being watched for movement.
The flight decides the resolution
Ground sample distance — the ground size of one image pixel — falls out of flying height, camera focal length and sensor. Lower flight means finer GSD and more images, more flight lines and more processing.
Overlap is the other lever. Photogrammetry needs each point of ground to appear in many images from different angles; thin overlap produces holes and noise exactly where the terrain is most complex, which on a mine site means the faces.
A pit is a hard case for a nadir-only flight plan: vertical walls are almost invisible looking straight down. Oblique imagery, or a flight plan that circles the pit, is what makes the faces reconstruct properly.
Control decides the accuracy
A photogrammetric block can be internally consistent and still sit a metre from where it belongs. That error is invisible in the model and appears only when the survey is compared with another one — at which point the whole difference looks like movement.
Ground control points tie the block to the site coordinate system. RTK and PPK drone systems reduce how many are needed but do not remove the need for check points, because a check point is the only thing that tells you the result is right rather than merely repeatable.
GSD and accuracy are not the same thing. A 2 cm GSD survey with poor control can be a metre out in absolute position, and its detail will look superb the whole time.
What comes out and what each part is for
A surface model, including everything standing on the ground — stockpiles, equipment, vegetation. This is the right surface for measuring a pile.
A bare-earth model, with those objects filtered out. This is the right surface for ground volumes and for design work.
An orthophoto, geometrically corrected so distances can be measured on it. On a mine site it is what makes a crest line, a toe or a stockpile boundary easy to trace by eye.
A point cloud, the underlying measurement from which the surfaces were made. It carries detail the gridded surfaces do not, and it is what you go back to when a surface looks wrong.
What the site does with it
Monthly volume and progress reporting: the difference between this survey and the last, restricted to the area in question.
Stockpile inventory: volume above a reference base, converted to tonnage with a material density.
Design checking: whether the benches, berms and floor levels being cut match the plan, and by how much they differ.
Blast planning: collar elevations and face geometry read from the surveyed bench rather than assumed, which is what makes hole depths and burden relief match the rock that is actually there.
The recurring mistakes
Measuring a ground volume on a surface model, so vegetation and equipment are counted as material. Measuring a stockpile on a bare-earth model, so the pile has been filtered away. Comparing two surveys that were controlled differently, so a systematic shift reads as movement.
None of these produce an obvious failure. They produce a plausible number, which is why the survey method belongs in the report next to the result.
Frequently asked questions
Is drone survey accurate enough for mine volume reporting?
Generally yes, with proper ground control and check points. The limiting factor is rarely the camera — it is the control that ties the block to the site system, and the surface type used for the calculation.
What GSD is needed for a mine survey?
It depends on what has to be resolved rather than on a fixed rule. Volume work tolerates a coarser GSD than detail mapping of faces or infrastructure. Remember that finer GSD means more images, longer flights and more processing — and that it does not by itself improve absolute accuracy.
Why do vertical faces come out badly?
Because a nadir flight plan barely sees them. Faces need oblique imagery or a flight path that looks into the pit, otherwise photogrammetry has too few views of the wall to reconstruct it.
Do I still need ground control with an RTK drone?
Fewer points, but you still need check points. RTK improves the positioning of the images; a check point is what independently confirms the finished model is where it should be.
STREAM opens the DEM, orthophoto and LAS/LAZ your photogrammetry pipeline produced and turns them into volumes and reports.
Native Windows · fully offline · free beta