From quarterly guesswork to monthly truth: auditing aggregate inventory with drones

An inventory audit is the moment the finance ledger and the physical yard are forced to agree. Propeller Aero's piece on auditing aggregate inventory frames it as a financial safeguard rather than a survey chore, and that framing is the useful one: the point is to stop small measurement errors from compounding into a write-off nobody saw coming.

The problem with quarterly

When each audit means a surveyor on site for a day or two, most producers settle on a quarterly cycle — and run on estimates for the three months in between. Propeller's observation is blunt: a pile can move by thousands of tons in a quarter, so a quarterly audit tends to find the problem after it has already become expensive. Their illustration puts a 10,000-ton discrepancy across several sites at $60,000–$70,000 of inventory that exists on paper but not on the ground.

What changes when the survey takes an hour

Drone surveys reset the cost side of the equation. Propeller describes flights of under an hour with results inside 24 hours, which is what makes a monthly audit affordable without hiring anyone. The cadence shift is the real product: the audit stops being a periodic compliance event and becomes a routine operational checkpoint, like reading a meter. Propeller recommends monthly as the default and weekly for high-throughput or high-value operations, and reports 97–99% accuracy when ground control is used — enough for defensible financial reporting.

Running a monthly program that auditors accept

• Same flight plan, same ground control, same base surface every month — consistency is what makes the month-over-month delta meaningful.
• Reconcile each month's measured volume against scale tickets in and out; variances that appear early are small and explainable.
• Keep the dated ortho and volume report together in one archive. The image is the evidence; the number is the summary.
• Agree material density factors per product with finance once, document them, and don't let them drift.
• Flag any pile whose variance exceeds a threshold (say 3%) for a closer look before quarter-end, not after.