Case Study: Drone vs Ground Survey on a Live Aggregate Yard
Aggregate producers traditionally rely on ground-survey crews for monthly stockpile inventory. UAV Imaging ran a side-by-side comparison on a working central Alberta aggregate yard to document, in concrete numbers, how a drone survey performs against the established ground-crew baseline.
Site profile
- Yard type: Working sand and gravel operation, central Alberta
- Surveyed piles: 6 active material types (washed sand, pea gravel, 20mm crush, road base, oversize, fines)
- Total area: Approximately 6.5 hectares of active stockpile footprint
- Combined estimated volume: Roughly 110,000 cubic metres
- Conditions: Clear sky, light wind, dry surface, mid-summer
Method 1: ground survey crew (baseline)
- 2-person Alberta Land Surveyor crew, GNSS rover + total station
- Shot density: roughly 250-350 points per pile depending on shape
- On-site time: 7 hours 20 minutes
- Office processing time: 1.5 days
- Deliverable: PDF volume report plus basic contour plot per pile
Method 2: drone photogrammetry survey
- DJI Matrice 300 RTK with H20T camera (20 MP wide), RTK GNSS base
- Image overlap: 80% forward, 70% side
- Ground sample distance: 1.8 cm per pixel
- On-site time (mobilization through demob): 1 hour 40 minutes
- Office processing time (Pix4Dmatic + QA): 1.5 days
- Deliverable: PDF volume report, orthomosaic, 3D point cloud (~140 million points), contour CAD, digital surface model
Results — same piles, two methods
| Metric | Ground crew | Drone | Delta |
|---|---|---|---|
| Total volume reported (m³) | 109,420 | 110,090 | +0.6% |
| Largest individual pile delta | — | — | 1.7% |
| On-site time | 7 h 20 min | 1 h 40 min | -77% |
| Surveyor cost | ~$4,100 | ~$1,300 | -68% |
| Operations disruption | Periodic loader pause near crew | None | — |
| Deliverable richness | PDF + contour | PDF + ortho + 3D + DSM + CAD | — |
What the numbers mean in practice
- Accuracy is at parity. 0.6% total-volume delta on six independent piles is within the noise of either method. The drone's 1.7% maximum single-pile delta corresponds to a steep-faced fines pile where the ground crew's 280 shot points underrepresented an irregular concave face that photogrammetry captured fully.
- Cost is the headline. 68% savings on a single quarterly survey. On a 12-yard regional aggregate operator with quarterly cycles, that is roughly $135,000 per year of survey cost back.
- Operations stay live. Drone flights run above loader and truck activity. No crew on the piles, no equipment shutdown, no scheduling around survey windows.
- Deliverables are richer. The point cloud and orthomosaic give the producer ongoing assets for pit-plan updates, regulator submissions, and seasonal change detection — not just a one-time volume number.
Where the drone did not win
Two piles were partially covered by tarps during the survey. The drone captured surface geometry only — the tarped material had to be estimated by the producer from delivery records. A ground crew with a probe can sample under the tarp, although in practice both methods relied on the producer's records for the tarped fraction.
Repeatability
UAV Imaging re-flew the same yard 28 days later as a repeat-survey check. Stockpile-by-stockpile deltas matched the producer's known material movements (deliveries in, sales out) within 1.1% across all six piles. Repeat-flight RTK ground-control re-occupation keeps absolute accuracy locked between cycles.
What this case study supports
For most open-air aggregate operations in Alberta, the drone survey is the rational default. It pays for itself inside the first survey cycle and produces deliverables a ground crew cannot. The ground crew still wins in tarped, covered or fully indoor storage where surface geometry alone is insufficient.