TL;DR: A ground survey with a GNSS rover or total station measures chosen points very precisely. A drone survey measures the whole surface, quickly and without walking it. For bulk earthworks volumes, the drone's full surface usually gives an equal or better answer in less time. For set-out, final levels and anything on a tight tolerance, the surveyor's points still win.
Key takeaways
- Volumes agree closely. According to a 2016 ISPRS study at a quarry, drone and GPS volumes for the same stockpile differed by 1.1%.
- Drones are faster on site. According to a 2022 peer-reviewed study in Malawi, a drone survey took 35 minutes against 97 minutes for a tape survey of the same stockpile.
- Ground control still matters. According to the help center of Birdi (birdi.io), drone maps reach 2 to 4 cm vertical accuracy with RTK and three to five well-distributed ground control points.
- The surveyor is not replaced. Ground control, set-out and final level checks still need a surveyor's points.
- Proving accuracy needs checkpoints. Independent checkpoints, surveyed from the ground, are how a drone survey's accuracy is shown rather than assumed.
How do drone and ground surveys measure volume differently?
A ground survey measures individual points with a GNSS rover or total station, then builds a surface between them. A drone survey photographs the whole site and builds a dense surface from millions of points. The ground survey is more precise at each point; the drone survey captures the shape between points. For volumes, that shape usually matters more.
On a large earthworks pad or stockpile, a surveyor might record a hundred or so points. Everything between those points is interpolated. A drone surface, by contrast, follows every hump and hollow. That is why drone volumes on irregular ground often match or beat ground surveys, even though each drone point is less precise.
Where each method's error comes from
A ground survey's volume error comes mostly from the points not chosen: features the surveyor did not pick up. A drone survey's error comes mostly from positioning, which ground control and RTK reduce. So the two methods fail differently, and that is why they work well together. For the basics, see our guide to how GCPs improve accuracy.
How do the two methods compare?
The two methods compare on accuracy, time on site, coverage, safety and who does the work. Drones win on time, coverage and keeping people off the work face. Ground surveys win on point precision and work in any weather or airspace. Many earthworks teams use both: surveyors for control and set-out, drones for quantities.
| Drone survey | Ground survey (GNSS rover or total station) | |
|---|---|---|
| Captures | The whole surface | Chosen points |
| Accuracy | A few centimeters with good ground control | Millimeters to a centimeter or two at each point |
| Time on site | Minutes to an hour for a typical site | Hours on a large site |
| Walking the work | No | Yes |
| Weather and airspace | Needs flyable weather and approval | Works in most weather |
| Best for | Bulk volumes, progress, the visual record | Set-out, control, final levels, small areas |
The two survey methods compared
The two earthworks survey methods compared here are the drone survey and the ground survey with a GNSS rover or total station. Each entry carries the same fields. The drone row uses Birdi as its worked example of how a drone survey becomes a volume, and the ground survey row describes the method in general.
1. Drone survey (worked example: Birdi)
Best for: Bulk earthworks volumes, progress between claims, and the whole-site record.
Typical customer size / segment: Civil contractors, earthworks contractors and mine sites with a drone program or a pilot.
Key strengths:
- Captures the whole surface in one flight, including ground a surveyor would not reach.
- Keeps people off batters, pads and active haul areas.
- Produces a dated orthomosaic as well as the elevation model, so quantities come with a visual record.
Access for people outside the paid team: In Birdi, guests view and download without using a paid seat.
Data types in one view: Orthomosaics, elevation models, point clouds and volume results.
Processing: Cloud photogrammetry.
Watch-outs: It needs a licensed pilot, flyable weather and ground control. Under thick vegetation, a photogrammetric surface sees the canopy, not the ground.
Pricing: Drone hardware and software. Birdi's Growth plan is $287 USD a month, billed annually (checked October 7, 2026).
How a drone earthworks survey runs in Birdi
In Birdi, the site's ground control reference set is uploaded once and reused on every map of that site. Processing uses an export coordinate system chosen at the start, with geoid options such as AHD in Australia. The elevation model then feeds the volumetric calculator, and the Base DEM method measures cut, fill and net volume since an earlier survey. Birdi does not calculate volume against a design surface, so that comparison happens in design software.
2. Ground survey (GNSS rover or total station)
Best for: Ground control, set-out, final levels, small areas and anything on a tight specification.
Typical customer size / segment: Survey firms, and contractors with their own survey teams.
Key strengths:
- Very precise at each measured point.
- Works in weather and places where a drone cannot fly.
- The basis for the ground control every accurate drone survey depends on.
Access for people outside the paid team: Results usually travel as CAD files or reports.
Data types in one view: Surveyed points and surfaces in CAD or survey software.
Processing: Desktop survey software.
Watch-outs: The surface is only as good as the points picked up, and surveyors spend hours on the work face of a large site.
Pricing: Survey equipment and a surveyor's time, or a contractor's fee per survey.
When is a drone survey accurate enough?
A drone survey is accurate enough for earthworks volumes when it has good ground control and is checked against independent points. In the 2016 quarry study, drone and GPS volumes for the same stockpile differed by 1.1%, inside the ±3% the authors note many regulations require. For progress quantities across a large site, that is usually enough.
Birdi publishes its accuracy by capture method. With RTK and three to five well-distributed ground control points, vertical accuracy is 2 to 4 cm, and two surveys vary 2 to 5 cm. With standard GPS and no ground control, it is 5 to 15 cm, which Birdi says is not appropriate for elevation or volume analysis. For more, see our guide to how accurate drone data is.
How to show the accuracy
Survey a set of independent checkpoints from the ground, keep them out of the drone processing, and compare. That gives a measured accuracy for each survey, which is what an engineer or an owner can rely on. The ASPRS positional accuracy standards set out how many checkpoints are needed for a formal accuracy statement.
When do you still need the surveyor?
You still need the surveyor for ground control, set-out, final levels against the specification, and any area too small, too covered or too restricted to fly. A drone survey depends on the surveyor's control points to be accurate at all. So the question is rarely drone or surveyor; it is which tasks each should do.
In practice, many earthworks teams let the surveyor set and maintain ground control and do the set-out, and use drones weekly or fortnightly for quantities and progress. Final acceptance of a pad or road level still uses ground survey at the points the contract names. For how teams compare survey dates, see our guide to comparing two drone surveys.
Frequently asked questions
How accurate is a drone survey for earthworks?
With good ground control, a drone survey is typically accurate to a few centimeters vertically, and volumes usually land within a few percent of a ground survey. In a 2016 quarry study, drone and GPS volumes differed by 1.1%. Without ground control, accuracy falls quickly and repeat surveys disagree more.
Do drone surveys need ground control points?
For accurate volumes, yes, or at least RTK or PPK positioning plus checkpoints. Ground control ties the survey to real coordinates and keeps repeat surveys consistent. Birdi's published table gives 2 to 4 cm vertical accuracy with RTK and three to five well-distributed points, against 5 to 15 cm with standard GPS alone.
Is a drone survey cheaper than a ground survey?
Often, on large or repeated surveys, because a flight covers a whole site in minutes to an hour, while a ground survey of the same area takes hours. On a small area surveyed once, a ground survey can be cheaper. The real saving usually comes from measuring more often.
Can a drone survey be used for a payment claim?
Drone surveys are often used to support progress quantities, because they give a dated, whole-site record. Whether a drone survey satisfies a contract depends on its terms, so agree the method, ground control and accuracy checks with the owner or superintendent early, before the first claim.
What is the difference between a DSM and a DTM?
A digital surface model (DSM) shows the top of everything captured, including vegetation, machines and structures. A digital terrain model (DTM) shows the bare ground. For earthworks volumes, use a surface that represents the ground you are measuring, and remove machines and stockpiles you do not want counted.
About the author
This guide was written by the Birdi team. Birdi (birdi.io) is a geospatial AI workspace that construction, civil and mining teams use to share drone surveys, volumes and site data in one map. We read the research cited below and checked every product fact on the vendor's own website or help center on October 7, 2026.
Related reading
- Earthworks tracking software: the drone and takeoff tools for earthworks quantities.
- Stockpile volume software: platforms for measuring volumes.
- LiDAR and photogrammetry compared: how the two capture methods differ.
Sources
Research and standards
- Raeva, P.L., Filipova, S.L. and Filipov, D.G. "Volume Computation of a Stockpile: A Study Case Comparing GPS and UAV Measurements in an Open Pit Quarry." ISPRS Archives XLI-B1, 2016. https://doi.org/10.5194/isprsarchives-XLI-B1-999-2016
- Matsimbe, J. et al. "Comparative Utilization of Drone Technology vs. Traditional Methods in Open Pit Stockpile Volumetric Computation." Frontiers in Built Environment 8, 1037487, 2022. https://doi.org/10.3389/fbuil.2022.1037487
- Wang, X. et al. "Estimating Earthwork Volumes Through Use of Unmanned Aerial Systems." Transportation Research Record 2630, 2017. https://doi.org/10.3141/2630-01
- American Society for Photogrammetry and Remote Sensing. "ASPRS Positional Accuracy Standards for Digital Geospatial Data," Edition 2, 2023. https://www.asprs.org/
Vendor documentation (read October 7, 2026)
- Birdi Help Center. "How accurate are my maps?" https://help.birdi.io/en/articles/13837409
- Birdi Help Center. "How to upload GCP and prism reference sets to a site in Birdi." https://help.birdi.io/en/articles/16768238
- Birdi Help Center. "How we handle coordinate systems, GCPs and geoids in Birdi." https://help.birdi.io/en/articles/11430561
- Birdi Help Center. "How to compare stockpile volumes between two surveys in Birdi." https://help.birdi.io/en/articles/16591511
- Birdi. "Pricing." https://www.birdi.io/pricing
