Birdi’s Stockpile model finds and outlines the stockpiles on your orthomosaic automatically. You don’t need to draw each pile by hand. It’s built for quarries, mine sites, recycling yards and construction sites where you need a fast, repeatable stockpile count.
The model is trained on real stockpile data. It looks at your orthomosaic’s colour and also at the elevation and slope from your DSM. That means it can tell a pile apart from flat ground of a similar colour, or from a shadow.
Trained on more than 15,000 real stockpiles
On sites it had never seen before, it found about 81% of stockpiles, and 86% of the outlines it drew were stockpiles
Results are added to your map as volume annotations, ready to calculate
Available on: Growth and Ultimate plans
What’s new compared with the previous stockpile model
The new model replaces our earlier stockpile detection. It finds more piles and outlines them more accurately, especially on large sites with lots of material.
Before (Sam 3) | After (Birdi's New Model) |
Before = image with fewer outlines (old model)
Example 1 - 2D view 
| After = image with more outlines (new model)
Example 1 - 2D view  |
Example 2 - 2D view 
| Example 2 - 2D view  |
Example 2 - 3D view  | Example 2 - 3D view  |
Example 3 - 3D view 
| Example 3 - 3D view  |
Tip: The 3D views show the difference best. Elevation helps the model trace each pile’s edge where it meets the ground.
When to use the Stockpile model
Running a stockpile count or month-end inventory survey
You need volumes or tonnage for many piles on one site
Your site has piles of similar-coloured material, or piles that sit close together
You want consistent stockpile outlines from survey to survey
Before you start
You’ll need:
A processed orthomosaic of your site
A DSM from the same survey. The model uses it to read the shape of each pile, so it won’t run without one.
A Growth or Ultimate plan
Tip: The Stockpile model only runs on a full orthomosaic. It isn’t available in Map Canvas AI Detect or Media AI Detect, because those don’t have elevation data.
Step-by-step: Detecting stockpiles
1. Open your map
Open the map with the survey you want to analyse.
2. Open AI Detect
Click AI Detect, then select Tag orthophoto.
3. Choose your orthophoto
Select the orthophoto you want to analyse from the dropdown.
4. Select the Stockpile model
Choose Birdi Stockpile model from the model list.
5. Select your DSM
Select the DSM for the same survey.
Tip: Always pick the DSM that matches your orthophoto. If you use a DSM from a different flight, the elevation won’t line up with the imagery and results will be poor.
6. Run the detection
Click Detect objects.
A map notification shows progress.
You can close the panel, or even the map. Detection keeps running on the server.
When it finishes, a new folder of volume annotations appears in your Layers panel.
7. Review your results
Before you calculate anything, check the outlines:
Delete anything that isn’t a stockpile: select it, right-click and choose Delete.
Edit an outline: select it and press Enter to adjust its points.
Split merged piles: if two touching piles come out as one outline, edit or redraw them as two.
Add missed piles: draw any the model didn’t pick up.
8. Calculate volumes
Detected stockpiles come in as volume annotations, but volumes aren’t calculated automatically. To get them:
Hold Shift and select the stockpiles you want to calculate.
Choose your Volume DEM and Base method in the right-hand panel.
Click Calculate.
For the full steps, see How to turn your stockpile AI Detect results into a volumetric report.
9. Add tonnage and export
With the piles still selected, choose a Material to report tonnage alongside cubic metres. Then open Table View to review and export your results.
What the model handles well, and where to check
The model works best on open, well-defined piles on active sites. Check your results more closely when:
Piles touch. Neighbouring piles sometimes come out as one outline. Split them before you calculate.
You’re on a recycling or mixed-material site. Irregular, low or patchy piles are harder to detect, so expect to add a few by hand.
Material is under tarps or in covered bunkers. Covered grain bunkers aren’t detected yet. Draw these manually.
Piles are at the edge of your survey. Partly captured piles may be outlined only in part.
Best practice tips
Fly with good overlap and consistent lighting. A clean orthomosaic and DSM give the best outlines.
Use the orthophoto and DSM from the same survey.
Always review outlines before you calculate volumes or share reports.
Rename the results folder (for example “Stockpiles – Oct 2026”) to keep repeat surveys organised.
Name your piles by material or location so your exported table makes sense to others.