# Volume on a pile grown together with another pile

I have a question for you skilled guys.
I have a request from a potential customer. They are having two piles that have grown together. The right one is of no interest to them, but the left one is. They are asking me if I can measure the volume on it.

As mentioned the piles have grown together. See image

Questions:
Can “split” the the piles as depicted (A-B) and get a volume with a reasonable accuracy?
Which of the available volume calculation methods (In DroneDeploy) is preferrable in this case?
How much error should I expect?

/Roger

That’s the way to do it. The Base plane will depend on the slope of the base across the piles. I almost always use triangulated. If there is more than 5% then definitely use triangulated. You can switch between triangulated and linear just to see the difference in numbers and split the difference if you like, but it shouldn’t be much.

Just don’t place a vertex in the “joined” area in order to interpolate between the vertices on either side.

2 Likes

@Dave: What happens if I do place a vertex in the joined area? There is one spot where I might have to do so, because the area makes a soft curve.

@MichaelL: The methods available for measuring volume are still confusing me. I normally use Triangulated with lots of “points” added around the pile, but I haven’t yet found a paper explaining/comparing the three methods available. That I understand, I must mention…
I’m also using RHINO using a Python script I got from from @SolarBarn that analyzes the .obj file

Gee how i appreciate you guys helping me out!

It will skew the base plane. Best to see it for yourself in the 3D view. The base plane will be a light gray line underneath the cut\fill colors. View with the extra point and then remove it and watch the base plane change. Same thing between triangulated and linear.

I’ve read the pages in your link before, but not really getting the “base plane” explanation. I normally double check the calculations using RHINO, and Triangulated works normally perfekt. I have never been in a situation where “Lowest point” would be useful.
But I’ll do the capture one of these days (tomorrow if the light is sufficient) and then I can try it out,

“3D view”? Are you referring to the section where the volume msmt is actually done?

Linear averages the elevation of the point digitized as the toe of the pile, but maintains slope. Traingulated is the most popular method in 3D content and is used by pretty much every other program on Earth to model. It creates those triangles by the nearest 3 neighbors. Lowest point sets a plane at the elevation of the lowest point you digitized. These are good when stockpiles are up against wall, in materials bins or on slopes that need to be benched.

If you click on the model tab any annotations that you have created will be visible. If you click on one of the stockpile areas a visual of the top surface of the pile will project above the pile. It shows red/blue (cut/fill) and just underneath that will show the base plane. One tip is that you want as little blue as possible in a stockpile and this is a good way to tweak it.

Michael covered it. I run into this quite a bit with one client’s yard. The only thing you can do is explain to the client that less than ideal conditions on the ground equates to less than ideal estimation results.

I’ve wondered why none of the processor vendors (that I know of) have added the ability to place a vertex and specify it is for x/y but ignore it’s z during base averaging. It would be mathematically pretty easy to do (for a computer) and would help tighten up these kinds of issues.

I think it just comes down to practice, but projecting a surface to a point with no known values beyond it can cause more harm than good. You would have to sample outside the perimeter to prorate the surface at that point otherwise you’re projecting a flat plane from the last set of triangles which may or may not be accurate. To be the 1% accurate that people think they need on stockpiles and in my opinion don’t would require DTM’ing the other stockpile which no processing software I have seen does well except for true point cloud editors. They either leave remnants or smooth the surface too much which could lead to just as much if not more error than just point exterior points in the right place and cutting across. I see all to often where people digitize way to many points. All you have to do is capture the general shape and pull out any significant blue areas in the model viewer.

I pretty much agree. I once had a conversation with a guy that has been managing aggregates for decades. He told me that in general, the process we are using with drones, when done as best as conditions will allow, are so far ahead of old methods that the errors are not really a consideration in most cases. But I still try and educate the operators that if their piles are clean, they will get the most accurate estimates and then leave it to them.

1 Like

Totally true. One of the first things I had to do was hold the Superintendents accountable for their guys keeping the stockpiles separated. Especially when there are different materials right next to each other.

and
This is a 3D of piles of ash.
I made a “Volume” annotation on the pile, and this is what I get.
The whole area under the white “carpet” is blue-ish. “As little blue as possible” you mentioned. How is this acheved?
I cannot se any “cut” nor “fill” colors. Confused…

1 Like

This was supposed to be fixed. @Jamespipe?

Nothing broken here - the white geometric view is what we show on lower tier accounts - Cut/Fill visualization is only available to enterprise customers.

@KRWaara you do not appear to have a paid DroneDeploy account. I assume those images are from a public link?

1 Like

At the moment I’m using a 14-day test account since I’m reluctant to renew my ordinary account due to the price tag and the forthcoming winter.
The images posted are copied from beeing logged in. I don’t know if this is what you call “Public link”…

Ah that makes sense! Ok yes so the business trial is also missing the cut/fill tools and visualisation.

1 Like

Now this was disappointing imho…

R

A quick one for you, guys, but toally impossible to understand for me:

I have this volume annotation

This gives me some 19883.1 qu meters.

If I move the upper left corner like this:

the volume jumps to 27 000 cu meters.

The height curve over the affected area does not (at least in my eyes) explain this jump.

Can you guys see any obvious reason to this? I find it disturbing. Neither do I know what is happening nor can I leave this to the customer knowing that there is a 7 000 cu meters difference if i cough when adding the vertexes just like that.

And yes: The straight line is in the middle of a slope grown together with another slope as I explained earlier. No vertex there as you recommended me.

Roger

Roger,

If you can send me a link to the .obj file for this site, then I could do a detailed look at this issue in Rhino.

Regards,
Terry.