For a long time the standard way to size a stockpile was a walking wheel, a tape, and a formula. Someone paced the base, estimated the height, treated the pile as a tidy cone, and did the arithmetic. It's fast, and on a perfectly shaped pile it's not unreasonable. But real piles aren't cones, and they sit on uneven ground. Every shortcut the formula takes is an error baked into the answer, and because the next person paces it differently, two estimates of the same pile rarely agree. When the number drives inventory, billing, or a year-end count, that spread is a problem.
Measuring from the air removes the guessing. Instead of approximating the pile as a shape, the drone captures the pile as it stands, thousands of points across its real surface, and the volume is calculated from that. Nothing is assumed about the form. The math runs against the geometry the camera saw.
The workflow, start to finish
It begins with a flight, not a photo. The drone flies a nadir mapping grid: straight-down passes laid out in a back-and-forth pattern at high overlap, so every part of the pile is seen from several angles. That overlap is what lets the software reconstruct depth. We fly in even light where we can, because hard shadows and blown-out highlights confuse the matching; a flat, overcast sky or consistent sun gives the cleanest result.
Position comes from an RTK workflow, which ties each image to a precise real-world coordinate rather than a rough consumer GPS fix. That is what makes the numbers reliable and easy to repeat next time: every flight sits in the same coordinate frame, so this month's pile and last month's pile are measured on the same map.
Back at the desk, those images are processed in-house into a 3D surface, a dense model of the pile and the ground around it. From that surface, the volume is calculated against a defined base surface: pick where the pile meets the ground, and the model returns the cubic metres sitting above it. The base can be the toe of the pile, a flat pad, or a surface captured before the material was ever placed. Define it consistently and you've defined exactly what's being counted.
Why repeatability is the whole point
A single volume is useful. A volume you can trust against the last one is what runs a yard. The real value shows up when you can compare this month's number to last month's. When the same site is flown the same way (same grid, same overlap, same RTK frame, the same base surface), the numbers line up. The change you see month to month is real material moving, not one person pacing it differently than the last. Compared like for like, the trend tells you what's coming in, what's going out, and what's on hand.
Recurring inventory is the real use
This is the kind of work that often makes sense as an ongoing program rather than a one-off, and I can set it up either way. A monthly, quarterly, or year-end capture turns a pile into a tracked line item. This suits gravel pits and aggregate yards reconciling what's sold against what's on the ground, ready-mix and material suppliers keeping a running yard count, and landfills tracking how cells fill over time. Each pass is captured the same way, so the record stays clean and comparable, period after period.
The honest boundary
Here's the part we say up front. These are inventory and management volumes, measured from the 3D model the capture builds, accurate and repeatable for planning, tracking, and reconciling what's on hand. We never present them as a legal or certified measurement; where a job needs that, it's a regulated professional's call, and our data supports that work.