How to Measure Stockpile Volumes with a Drone Survey
A step-by-step guide to drone stockpile volume measurement: flight planning, ground control, photogrammetry or LiDAR, base surfaces, volume calculation and accuracy.
By Ahsan Mukhtar, CEO and Managing Director
Key takeaways
- The base surface you choose affects the volume more than the drone does.
- Ground control points or RTK/PPK positioning are what make a survey measurable, not just pretty.
- Repeat surveys with the same method and base so changes over time can be trusted.
Drone surveys have become the standard way to measure stockpiles at quarries, mines, ports and construction sites. One flight captures every pile on a site in minutes, without anyone climbing a stockpile with a GNSS pole. The volume itself is simple arithmetic. What makes it trustworthy is everything around it: how the flight is planned, how the survey is tied to the ground, and how the base of each pile is defined.
This guide walks through the whole workflow, from flight to report.
Try it yourself: our Drone 3D Viewer loads a classified drone point cloud and 3D mesh of a former limestone quarry in Kielce, Poland. You can measure distances, heights, areas and volumes in your browser.
1. Plan the flight for measurement, not for pictures
Photogrammetry rebuilds 3D shape from overlapping photos, so overlap matters more than image beauty.
- Overlap: plan high forward and side overlap so every point on a pile appears in many photos. Steep faces and tall piles need more than flat ground.
- Altitude: lower flights give a smaller ground sample distance (more detail per pixel) but take longer and produce more data. Pick the height that resolves the pile edges you need to measure.
- Oblique images: add a few angled passes around tall or steep piles, so their sides are reconstructed rather than guessed.
- Light: fly when shadows are short and consistent. Deep shadows and wet, reflective surfaces reconstruct poorly.
2. Tie the survey to the ground
Without control, a drone model can have the right shape but the wrong size, tilt or height. There are two common ways to fix it:
- Ground control points (GCPs): marked targets surveyed with GNSS and spread around and between the piles, with a few kept back as independent check points.
- RTK or PPK positioning: the drone records precise camera positions during the flight, which reduces the number of GCPs needed. Check points are still recommended.
Keep the coordinate system and vertical datum the same for every survey of the site. Comparing a survey in one datum with another in a different one produces volume changes that never happened.
3. Process the images into a surface
Processing software such as OpenDroneMap, Pix4D or Agisoft Metashape turns the photos into:
- a dense point cloud, millions of 3D points with colour
- a digital surface model (DSM), a height grid of everything on the site
- an orthomosaic, a map-accurate image used to outline the piles
Read the processing report. It shows how well the photos aligned and how far the model sits from the check points. If the check point errors are large, fix them before measuring anything.
Drone LiDAR is an alternative to photogrammetry. It measures the ground directly and sees through vegetation, which matters on overgrown sites. On bare stockpiles, photogrammetry usually gives a dense, detailed surface at lower cost.
4. Clean and classify the points
Before measuring, remove what is not stockpile: machines, vehicles, conveyors, people and vegetation. Point cloud classification separates ground, vegetation, buildings and other classes, either with rules or with machine learning. Our guide to LiDAR point cloud classification covers the methods, and the same ideas apply to photogrammetric point clouds.
5. Define the base of each pile
This is the step that decides the answer. The volume of a pile is the space between its top surface and its base, and the base is never measured directly, because the pile sits on it. Common choices are:
| Base surface | How it works | Good for |
|---|---|---|
| Lowest point | A flat plane at the lowest point of the pile outline | Piles on level, hard ground |
| Triangulated toe | A surface stretched between points around the pile edge | Piles on sloping or uneven ground |
| Best-fit plane | A plane fitted through the points around the edge | Roughly level ground with small bumps |
| Earlier survey | The ground surface surveyed before the pile was placed | Bays and pads surveyed when empty |
Two people measuring the same pile with different bases can disagree by more than the drone's error. Choose the method once per pile type, write it down and use it every time.
6. Calculate the volume
With a top surface and a base, the calculation is straightforward. The area is divided into small grid cells; for each cell, the height of the surface above the base is multiplied by the cell area, and the results are added up. Where the surface dips below the base, that volume is reported separately as cut, so you see both fill and cut.
To report tonnage rather than cubic metres, multiply by the material's bulk density. Density varies with moisture and compaction, so it should come from the site's own tests rather than a textbook value.
7. Report and compare over time
A useful stockpile report shows, for each pile:
- the outline on the orthomosaic
- the base method and the volume, cut and fill
- the change since the last survey
- the survey date, coordinate system and check point accuracy
Monthly or weekly surveys turn single numbers into a trend: what came in, what went out, and whether the inventory in the accounts matches what is on the ground.
Common pitfalls
- Changing the base method between surveys, which creates false gains and losses
- No independent check points, so nobody knows how accurate the survey is
- Measuring piles with machines parked on them
- Using a generic density to convert volume to tonnage
- Mixing coordinate systems or vertical datums between surveys
Sharing the results
Survey outputs are large, and most of the people who need them do not run GIS or photogrammetry software. Streaming the point cloud and mesh as 3D Tiles lets anyone open the survey in a web browser, measure it and see the same numbers as the surveyor. That is the approach behind our Drone 3D Viewer, and our Site in Context demo shows the same idea with point clouds, city models and BIM together.
Summary
Accurate drone stockpile volumes come from good overlap, solid ground control, clean points and, above all, a consistent base surface. Measure the same way every time and the trend becomes as valuable as any single number. If you want a processing pipeline or a web viewer for your own surveys, our GIS development team builds both.
References and further reading
Working on something similar? Jugrafiya builds GIS products, spatial data pipelines and the web and mobile software around them.
Talk to an engineer