Cut and Fill Volumes from Drone Surveys: Surfaces, Methods and Payment Checks
By Kush Agarwal, President and CEO, YelloSKYE. Published .
Executive Summary
- The Question: How do you turn a drone survey into cut and fill quantities that a designer can plan with and a commercial team can pay against?
- The Answer: Cut and fill is the height difference between two surfaces, summed over an agreed area. The quantity is only as good as the choice of those two surfaces, the method used to sum them, the state of the material (bank, loose or compacted) and the control behind the survey.
- The YelloSKYE Angle: Survey IQ builds terrain models from drone surveys and the DEX platform compares them with the design grading plan and with earlier flights, so each period's earthwork quantity can be opened and checked by both sides.
Cut and fill volumes measure how much earth has to be dug out of a site and how much has to be placed back to reach the design levels. With a drone, you build a bare-earth model of the ground, compare it with a second surface such as the design formation or an earlier survey, and the software reports the volume above (cut) and below (fill) that surface across the area you define.
At YelloSKYE this sits inside Survey IQ, alongside topography, stockpiles and change detection. This guide deals with earthworks on a site: which surfaces to compare, how the arithmetic works, how to convert between dug, hauled and compacted material, and how to use the result to check a contractor's quantity. Measuring loose piles of material is a separate job, covered in our stockpile volume measurement guide.
The surfaces every earthwork calculation needs
Every cut and fill figure is a comparison between two surfaces. Getting the right pair matters more than any setting in the software.
| Surface | What it represents | Where it comes from |
|---|---|---|
| Original ground | The site before any work | A pre-construction drone or ground topographic survey |
| Current ground | The site on the day of the flight | The latest drone survey, cleaned to bare earth |
| Design finished level | The top of the completed works | The designer's grading model |
| Design formation | The level earthworks must reach before pavement, slab or topsoil layers go on | Design finished level minus the build-up thickness |
| Previous survey | The ground at the end of the last measurement period | The earlier drone survey in the series |
Pick the pair to suit the question
Original ground against design formation gives the total earthwork for the job, which is what an estimator needs. Current ground against design formation gives the work still to do. Current ground against the previous survey gives what was moved in the period, which is the number that matters for a payment claim. Mixing these up is the most common reason a figure looks wrong.
Formation, not finished level
Earthwork contractors usually stop at formation. A car park designed at a given finished level might carry a few hundred millimetres of sub-base and asphalt on top, and a building platform carries its slab. Comparing current ground with finished levels makes fill look short across the whole area by the thickness of that build-up. Ask the designer for a formation surface, or subtract the layer thicknesses by zone before you calculate.
The baseline cannot be recreated later
Original ground disappears as soon as the first dozer arrives. A topographic survey flown before site clearance is the reference every later quantity depends on, and it is far cheaper than arguing about an assumed baseline at the end of the job. Our drone topographic survey guide explains how to specify that first survey so it holds up.
How the volume is actually calculated
Software uses one of three broad methods. All of them can produce reliable answers; they differ in how they treat the space between measured points.
| Method | How it works | Strengths | Watch for |
|---|---|---|---|
| Grid (cell) method | Lays a regular grid over both surfaces, takes the height difference in each cell and multiplies by cell area | Simple, transparent, easy to break down by zone | Cells too large smooth over sharp edges such as ditches and batters |
| Surface-to-surface (TIN) | Builds triangles on both surfaces and computes the exact solid between them | Follows breaklines and steep faces closely | Needs clean surfaces; spikes from vegetation inflate the result |
| Cross-sections (average end area) | Cuts sections at fixed spacing along a road or channel and averages the areas between them | Familiar on linear jobs; matches many contract documents | Misses changes between sections; spacing drives the error |
A dense drone surface suits the grid and TIN methods, because there is height data almost everywhere. Cross-sections remain useful on roads and channels where the contract asks for them.
A worked grid example
The table below shows six cells of a 10 m by 10 m grid on a building platform with a design formation level of 51.90 m. Each cell's volume is its area (100 m²) multiplied by the height difference.
| Cell | Current ground (m) | Design formation (m) | Difference (m) | Cut (m³) | Fill (m³) |
|---|---|---|---|---|---|
| A1 | 52.40 | 51.90 | +0.50 | 50 | 0 |
| A2 | 52.10 | 51.90 | +0.20 | 20 | 0 |
| A3 | 51.70 | 51.90 | −0.20 | 0 | 20 |
| B1 | 51.95 | 51.90 | +0.05 | 5 | 0 |
| B2 | 51.60 | 51.90 | −0.30 | 0 | 30 |
| B3 | 51.30 | 51.90 | −0.60 | 0 | 60 |
| Total | 75 | 110 |
On this patch the platform needs 35 m³ more fill than it yields in cut. A real drone model repeats the same arithmetic over millions of much smaller cells, but the logic is identical, and being able to explain it this simply is useful when a figure is challenged.
Bank, loose and compacted: the conversion that changes the answer
The volume calculated from two surfaces describes space, not material. The same soil occupies different volumes at different stages:
- Bank volume: undisturbed in the ground. A cut measured against original ground is a bank volume.
- Loose volume: after excavation, when the soil has broken up and bulked. This is what fills a truck.
- Compacted volume: after placing and rolling. A fill measured between two surfaces is a compacted volume.
Swell (or bulking) describes the increase from bank to loose; shrinkage describes the reduction from bank to compacted. Both depend on the soil type, its moisture and the compaction specification, so the reliable figures come from site testing or the geotechnical report rather than a generic table.
Worked example: is the site in balance?
Suppose the drone survey and design give a total of 18,500 m³ of cut (bank) and 14,200 m³ of fill (compacted). The factors below are illustrative only; use those from your own testing.
| Step | Calculation | Result |
|---|---|---|
| Cut available, in compacted terms (shrinkage factor 0.90) | 18,500 × 0.90 | 16,650 m³ compacted |
| Fill required | From the drone and design surfaces | 14,200 m³ compacted |
| Surplus after filling | 16,650 − 14,200 | 2,450 m³ compacted |
| Surplus back in bank terms | 2,450 ÷ 0.90 | about 2,720 m³ bank |
| Surplus as hauled (swell factor 1.25) | 2,720 × 1.25 | about 3,400 m³ loose |
| Truck loads at 10 m³ loose each | 3,400 ÷ 10 | about 340 loads off site |
On the raw numbers, the cut exceeds the fill by 4,300 m³ and someone might plan for that much export. After conversion the real surplus is closer to 2,720 m³ in the ground, and the haulage plan needs to be priced on the loose figure. On a tight site the difference decides whether you pay to import, export or neither.
Material that cannot be reused
Not every cubic metre of cut is suitable fill. Topsoil stripped at the start, soft or organic material and rock that needs breaking all have to be separated in the quantities. Measure the topsoil strip as its own layer with a survey after stripping, and record areas of unsuitable material on the map as they are found, so the balance is not built on material that will never go back into the works.
Verifying material moved for payment
Most earthwork disputes are not about the drone or the arithmetic. They are about which surfaces, which boundary and which state of material the claim used. Agreeing these at the start of the contract turns monthly measurement into a routine check.
What to agree before the first claim
| Item | What to fix in writing | Why it matters |
|---|---|---|
| Measurement surfaces | Baseline survey and design formation, by name and date | Both sides calculate from the same start and end |
| Measurement boundary | The polygon of each work package or zone | Stops claims for work outside the package |
| Pay lines | Whether quantities are paid to design lines or as measured | Over-excavation beyond design is not automatically paid |
| Material state | Bank for cut, compacted for fill, unless stated otherwise | Avoids paying swell as if it were extra work |
| Survey dates | A flight just before each claim cut-off | Period quantities line up with the billing cycle |
| Accuracy check | Checkpoint results reported with each survey | Turns accuracy into evidence rather than assertion |
A monthly routine that works
- Fly the site as close as practical to the claim cut-off, with the same control and flight plan as last time.
- Clean the surface: remove parked machines, temporary stockpiles and vegetation before calculating.
- Compare the new surface with the previous survey inside each agreed zone to get the period's cut and fill.
- Compare the new surface with design formation to see the remaining quantity and catch over-excavation.
- Share the colour cut and fill map and the zone totals with the contractor before the claim is assessed.
- Investigate only the zones where claim and survey differ by more than the agreed tolerance.
Temporary stockpiles on the platform deserve care. A pile of cut waiting to be placed is neither cut that has left nor fill that has been built, so clip it out of the earthwork surface and measure it on its own.
What it looks like in practice
On the highway project in our earthwork volumetric analytics use case, drone surveys produced a terrain model every few days and DEX compared it automatically with the design grading plan. The commercial team moved from monthly arguments to zero billing disputes, survey turnaround dropped from two weeks to a few hours, and site managers used the same data to balance cut and fill on site and cut paid external hauling. Where earthwork is one part of a wider valuation, aerial captures timed to each billing cycle support the rest of the claim, as described in our contractor billing disputes use case.
Controlling accuracy
A cut and fill figure carries the vertical error of both surfaces across the whole area, so small systematic errors grow quickly. A consistent 3 cm height bias over a 5 ha platform shifts the result by around 1,500 m³, which could be several days of haulage.
Three things keep that under control:
- Positioning and control. RTK or PPK on the aircraft, surveyed ground control and checkpoints that were not used in processing. Our guide to ground control points for drone surveys covers layout and numbers. With RTK-GNSS drones and ground control, Survey IQ works to sub-5 cm accuracy, and its earthwork volumes typically land within 1–3 per cent of a conventional ground survey.
- A clean bare-earth surface. Grass, scrub, machines and site cabins all sit above the ground and read as cut. Classify and remove them before the comparison, not after a number looks odd.
- Repeatability. Same flight height, pattern, control and processing settings every time, so the difference between two surveys reflects the ground and not the method.
Ground survey still has a role: confirming levels on hard surfaces, checking formation at a handful of points before sign-off, and surveying areas the drone cannot see. A mixed approach is normal on earthworks.
Before you design or pay on drone cut and fill
Drone cut and fill is also useful before any machine arrives. The terrain-based design validation use case shows a layout tested against drone terrain to confirm earthwork assumptions before design is locked, which is when a better platform level is cheapest to choose. For earthworks inside a quarry or mine, where benches and haul roads add their own questions, see our drone mining survey guide.
Checklist for drone cut and fill
- Fly a baseline topographic survey before clearance, and keep it as the agreed original ground.
- Get a design formation surface, not only finished levels, and confirm layer thicknesses by zone.
- Name the comparison surfaces, the zone boundaries and the survey dates in the contract.
- State whether each quantity is bank, loose or compacted, and use tested swell and shrinkage factors.
- Measure topsoil strip and unsuitable material separately from general cut.
- Use RTK or PPK, ground control and independent checkpoints, and report the checkpoint results.
- Remove vegetation, machines and temporary stockpiles from the surface before calculating.
- Fly just before each claim cut-off, using the same flight plan and processing settings.
- Share the cut and fill map with the contractor, and investigate only the zones outside tolerance.
Frequently asked questions
How are cut and fill volumes calculated from a drone survey?
The drone survey produces a bare-earth terrain model of the site. Software compares it with a second surface, usually the design formation or an earlier survey, and adds up the height difference over the whole area. Where the ground is above the comparison surface the material counts as cut, and where it is below it counts as fill.
What is the difference between bank, loose and compacted volume?
Bank volume is the material in its natural state in the ground. Once dug, it breaks up and takes more space in a truck, which is loose volume. When it is placed and rolled as fill it is squeezed into less space than loose and often less than bank, which is compacted volume. Earthwork quantities only compare properly when everyone states which of the three they mean.
Is a drone survey accurate enough to pay earthwork contractors?
With RTK or PPK positioning, surveyed control and independent checkpoints, a drone terrain model is a sound basis for measured quantities, and YelloSKYE's drone volumes come within 1 to 3 per cent of a traditional survey. The bigger risks are usually agreeing the wrong comparison surface, leaving vegetation or machines in the model, and mixing bank and compacted figures.
Should cut and fill be measured against the design or against the last survey?
Both, for different questions. Comparing today's ground with the design shows how much work remains. Comparing it with the previous survey shows what was moved in the period, which is the figure a progress payment claim should be checked against.
Why do drone and ground survey cut and fill figures sometimes disagree?
A ground survey builds the surface from a limited set of picked points and straight lines between them, while a drone surface follows every hump and hollow. The two also often differ in the base surface, the boundary of the area measured and whether topsoil or pavement layers were removed. Agree those inputs first and most of the gap usually disappears.