We fly weekly stockpiles at our quarry. Using M3E RTK with DroneDeploy. Comparing our drone volumes to scale ticket records and we're usually within 2-3% which management is happy with.
But I'm curious what others are seeing. Are you comparing to truck scales? Survey grade total station volumes? What's realistic accuracy for stockpile work?
Sign in to join the discussion
Sign in2-3% vs scale tickets is pretty standard and honestly as good as you'll get. The variables are:
For internal inventory tracking, 2-3% is excellent. For payment disputes, I wouldn't rely on either method alone.
One tip - be religious about your base plane methodology. Pick a method and stick with it. Don't let the software auto-detect one week and manually draw the next. Consistency matters more than perfection for tracking changes over time.
We've done some testing against total station volumes (grid points, breaklines, the old fashioned way). Drone vs TS was within 1% on a clean conical pile. But those controlled conditions aren't realistic.
Real piles with loaders working, irregular shapes, debris - 2-3% is good. 5% isn't unusual if conditions are messy.
Worth noting that your GSD matters a lot for stockpiles. Lower and slower = more detail on the pile surface. I see people flying at 300ft and wondering why their volumes are inconsistent. For stockpiles I fly at 150-200ft AGL. Yes it takes longer but accuracy is better.
Good point on the base plane. We had exactly that problem when I first started - volumes jumping around because I wasn't consistent with toe definition. Now I have saved templates for each pile location.