How Drone LiDAR Surveys A Brazilian Mine Without Stopping Production
- by Stefan Gandhi
At a copper and gold surface mine in Brazil, a topographic survey used to mean standing down heavy machinery for a few minutes at every face, over and over, day after day. Lundin Mining now captures the same survey grade topography from the air using a DJI Matrice 400 carrying a DJI Zenmuse L3, without pausing the operation underneath it. This article looks at what that change involves, why LiDAR suits an active mine, and what the results mean for anyone running survey work on a live industrial site.
The Problem With Surveying A Working Mine
Mine surveying is not a once a year exercise. Excavation volumes have to be tracked continuously, because they drive production reporting, reconciliation against the block model and the planning of the next cut. On a large surface operation that means recurring survey work across a site measured in thousands of hectares.
Ground based methods force a choice nobody wants to make. A surveyor working at a face needs the machinery around them to stop. Multiply a few minutes of downtime by every face, every survey cycle, across a full year, and the cumulative cost of lost production becomes significant on its own, before you count the safety exposure of putting people next to active equipment and unstable ground.
Lundin Mining's Brazilian operation sits at Alto Horizonte in Goiás and spans more than 10,500 hectares, with excavation volumes monitored on a weekly basis. That is exactly the profile where ground survey time compounds fastest.
Why LiDAR Rather Than Photogrammetry
Photogrammetry builds a model from overlapping images, which works well on clean, well lit, texture rich ground. Mine environments frequently offer none of those things. Dust, harsh shadow across benches and haul roads, uniform rock surfaces with little visual texture, and vegetation creeping over old faces all degrade a photogrammetric result.
LiDAR measures range directly with a laser rather than inferring geometry from pixels, so it holds up where photogrammetry struggles. It penetrates light vegetation to return ground points beneath, it does not depend on surface texture, and it is far less troubled by low or uneven light. For volumetric calculations across benches and stockpiles, that reliability is the whole point.
The Zenmuse L3 is DJI's long range LiDAR payload for the Matrice 400, designed for survey grade point cloud capture at altitude. The combination is what lets a single aircraft cover a large pit from a safe height rather than working close to the ground.
The Efficiency Gain That Actually Landed
The headline result from the Brazilian operation is a straightforward one. A survey that previously consumed three to four batteries now takes one.
That is not simply a battery saving. Fewer batteries means fewer landings, fewer swap cycles and fewer interruptions to the capture, which in turn means a shorter total window during which the survey is running at all. On a site where the survey window used to be the thing forcing machinery to stand down, compressing it to a single flight changes the operational maths entirely.
The team's earlier work with DJI aircraft had already established the pattern. Using DJI Matrice 300 RTK aircraft with Zenmuse L1 LiDAR and Zenmuse P1 photogrammetric payloads, supported by a D-RTK 2 base station, the operation reported around 3cm accuracy, full site coverage with a detailed 3D mesh and roughly a 50 percent reduction in execution time compared with ground based GNSS RTK and terrestrial laser scanning.
The Outputs A Mine Actually Uses
A point cloud on its own is not a deliverable. The value comes from what gets derived from it, and on this site that means orthomosaics, digital terrain and surface models, contour lines and textured 3D meshes, processed through DJI Terra alongside third party photogrammetry and mesh tools.
Those outputs feed several departments at once. Surveying uses the terrain models for volumes. Geology works from the same capture. Engineering plans from the contours. Environmental and security teams get current site imagery without a separate mobilisation. Archaeology has used it too, which is a reminder that a single flight over a large site tends to answer questions nobody commissioned it for.
Lundin Mining built an internal drone centre around this, with standardised flight protocols, managed equipment and around 30 trained personnel across those departments. That structure matters more than the hardware. A drone programme that lives with one enthusiast produces occasional data. A programme with protocols and trained staff across departments produces a dependable weekly dataset the business can plan against.
What Transfers To Other Sites
Very little of this is mining specific. Any operation that has to survey the same large area repeatedly while it continues working faces the same trade off, including quarries, landfill and waste transfer sites, large construction projects and stockpile yards.
The pattern is consistent. Identify the recurring survey that currently forces work to stop, check if the accuracy requirement genuinely needs LiDAR or if photogrammetry would do, then build the flight into a routine rather than treating each capture as a project. The saving comes from repetition, not from any single flight.
FAQs
How accurate is drone LiDAR surveying?
Survey grade drone LiDAR systems typically deliver centimetre level accuracy when flown with RTK or PPK positioning and validated against ground control. Lundin Mining reported around 3cm accuracy on its Brazilian operation. Actual results depend on flight altitude, ground control quality and the terrain being surveyed.
What is the difference between LiDAR and photogrammetry for drone survey?
LiDAR measures distance directly with a laser, so it works in poor light, ignores surface texture and penetrates light vegetation to reach the ground. Photogrammetry reconstructs geometry from overlapping photographs, which produces excellent colour detail but struggles with dust, shadow and vegetated or featureless surfaces.
Can drones survey an active mine site?
Yes, and that is one of the strongest arguments for using them. Because the aircraft works from the air rather than on the ground, machinery below can usually keep running, removing the production stoppages that ground survey methods require.
How long does a drone LiDAR survey take?
It varies with site size and required point density, but the Brazilian operation cut a survey that previously needed three to four battery sets down to a single battery. For a large surface mine, that means a capture window measured in a single flight rather than a working session.
Final Thoughts
The interesting result here is not the accuracy figure. It is that the survey stopped being an event the mine had to make room for. Once a capture fits into one flight and the machinery below never stands down, weekly survey data becomes something the operation simply has, rather than something it schedules around. That is the shift worth copying, and it applies to any site large enough that surveying it the old way costs production time.
Specifying LiDAR survey capability for a live industrial site? See the DJI Zenmuse L3 Worry-Free Plus Combo at the Coptrz official online store.




