How to Combine Laser Scanning and Drones

TL;DR: Combining laser scanning and drones means using terrestrial laser scanners for high-detail, millimetre-grade capture at ground level and inside structures, while a drone captures the roofs, elevated steelwork and broad terrain that scanners cannot reach safely. The two datasets are tied together through shared ground control and registered into a single point cloud on a common datum (GDA2020/MGA2020 and AHD), giving you one complete, dimensionally accurate model of the whole asset.

Key takeaways

• Use terrestrial laser scanning (Leica RTC360, FARO Focus, Trimble X9) for sub-6 mm detail at ground level, and a drone (DJI Matrice 350 RTK or Mavic 3 Enterprise) for elevated and large-area capture you would otherwise need scaffolding or a man-lift to reach.
• The datasets only fuse cleanly if both are referenced to the same control: place and survey 6-10 ground control points (GCPs) with RTK GNSS on GDA2020/MGA2020, then register both clouds to that frame.
• Expect roughly 2-6 mm accuracy from terrestrial scans and 20-50 mm from drone photogrammetry; combine them so each method does the work it is best at rather than forcing one to do everything.
• Drone work in Australia is regulated under CASA Part 101 — a Remote Pilot Licence (RePL) and an operator's certificate (ReOC) are required for commercial flights, plus site-specific approvals near aerodromes or controlled airspace.
• A combined capture of a typical processing plant or mine pit-and-infrastructure site runs around AUD 4,000-15,000 depending on scale, and routinely removes one to three days of working-at-heights access from the program.

Why combine the two methods at all

No single sensor captures an industrial site well. A terrestrial laser scanner standing on the ground produces a dense, colourised point cloud accurate to a few millimetres — ideal for conveyor structures, mill foundations, pipe racks and tank farms. But it sees the world from where the tripod sits. Roofs, the tops of silos, elevated walkways, stockpiles and the wider pit or laydown area are either occluded or simply out of range.

A drone solves the opposite problem. Flown at 40-80 m above ground level it maps hundreds of hectares in a single sortie, captures roof condition and high steelwork without anyone leaving the ground, and produces a bare-earth surface for volumes. What it does not give you is the tight tolerance you need for an alignment, a tie-in or a clash check — drone photogrammetry typically lands at 20-50 mm, and considerably worse on featureless or reflective surfaces.

Combining laser scanning and drones lets each instrument do what it is good at. The result is a single dataset that is millimetre-accurate where it matters (the equipment) and comprehensive where it needs to be (the whole footprint). For a brownfield revamp, a shutdown scope, or a mine site digital twin, that completeness is the difference between a model engineers can design from and one with holes in it.

What each method captures best

The practical division is straightforward: if a person would normally climb to reach it, fly it; if it needs to be within a few millimetres, scan it. Anything in between — say, a 12 m tall crusher — is often captured by both and the two clouds cross-check each other.

How to combine laser scanning and drones: the workflow

This is the workflow Industrial Spatial Solutions follows to merge terrestrial scan data and drone data into one deliverable. It assumes a typical industrial or mine site of a few hectares with plant infrastructure.

Step 1: Plan the capture and define the datum (before mobilisation)

Decide up front what tolerance each part of the site demands, and which method serves it. Agree the coordinate system — for almost all Australian work this is GDA2020 with the relevant MGA2020 zone for horizontal position and AHD for height. Getting the datum agreed in writing before anyone mobilises prevents the single most common integration failure: two beautiful datasets that will not line up because they were referenced to different frames.

Tip: Confirm the MGA2020 zone for the site (for example, Zone 50 for the Pilbara, Zone 56 for the Hunter Valley). A cloud built on the wrong zone is metres out, not millimetres.

Step 2: Establish ground control (day 1)

Place and survey 6-10 ground control points distributed across the site and visible from the air — painted targets or chequerboard mats for the drone, and survey spheres or black-and-white targets for the scanner. Coordinate each with RTK or network GNSS to a few centimetres on MGA2020/AHD. This control is the common backbone both datasets register to, so it is worth doing carefully. Place GCPs at varying elevations where possible, not just on flat ground, to constrain the vertical solution.

Tip: Position several control marks so they are usable by both the drone and a nearby scan setup. Shared targets give you a direct check between the two clouds later.

Step 3: Terrestrial laser scanning (days 1-2)

Scan the plant and structures with the terrestrial scanner, capturing enough setups that adjacent scans overlap by 30-40% for reliable registration. A Leica RTC360 captures around two million points per second and completes a setup in roughly two minutes; a FARO Focus Premium or Trimble X9 gives comparable results. Capture HDR imagery for colourisation. Keep targets or shared geometry in view between setups, and tie selected setups to the surveyed control marks.

Tip: Scan reflective or wet surfaces (stainless tanks, polished steel, puddles) from multiple angles — single-angle returns on these surfaces are noisy and drop out of the registration.

Step 4: Drone capture (day 2)

Fly the drone to cover roofs, elevated structures, stockpiles and the surrounding terrain. For photogrammetry, plan 75-80% front and side image overlap at an altitude that delivers the ground sample distance (GSD) you need — typically 1.5-3 cm/pixel. Use an RTK-equipped aircraft (Matrice 350 RTK) so each image is geotagged to within a few centimetres, which reduces reliance on GCPs and tightens the eventual fit to the scan data. Where vegetation hides the ground, switch to a UAV LiDAR payload to capture bare earth.

Tip: Fly early morning when wind is lowest and shadows are long but soft. Harsh midday shadows degrade photogrammetric matching on structures.

Step 5: Process and register to a common datum (office)

Register the terrestrial scans together first, constraining the registration to the surveyed control. Process the drone imagery into its own point cloud and orthomosaic, also tied to the same GCPs. Then bring both into a single project on the shared datum. Because both were referenced to the same control, they drop into alignment; check the residuals at the shared targets and at overlapping geometry such as the crusher captured by both methods.

Tip: Inspect the seam where scan and drone data meet — usually the lower roofline or top of a structure. A clean transition there tells you the integration is sound; a step indicates a control or scale problem to resolve before delivery.

Step 6: Quality-check and deliver

Verify the merged cloud against independent check points that were not used in registration. Confirm the high-detail zones still hold their few-millimetre accuracy and that the drone-derived areas meet their stated tolerance. Deliver the unified point cloud (typically .e57, .las or .rcp), an orthomosaic, a digital terrain model, and any modelled deliverables — a Revit or AutoCAD model for scan-to-BIM, or volumes for stockpiles.

Equipment and resources checklist

ISS to provide:

• Terrestrial laser scanner (Leica RTC360, FARO Focus Premium or Trimble X9) with current calibration certificate
• CASA-compliant drone (DJI Matrice 350 RTK or Mavic 3 Enterprise) with RTK module, and a UAV LiDAR payload if vegetation requires it
• RTK GNSS rover and base for ground control on MGA2020/AHD
• Survey targets, spheres and aerial GCP markers
• Licensed surveyor plus a RePL-holding remote pilot operating under the ISS ReOC
• Registration and processing software, and the merged deliverables on the agreed datum

Site to provide:

• Site induction and access for survey vehicle and crew
• Confirmation of the coordinate system and any existing site control marks
• Airspace information — proximity to aerodromes, helipads or controlled airspace, and any site flight restrictions
• A site contact with authority to grant access to elevated or restricted areas
• Notice of reflective, hot or live equipment that affects scanning or flight

Note: The site does not need to provide scaffolding or man-lifts for the elevated capture — that is precisely the access the drone removes. Flag any no-fly constraints early, because they change the method mix.

Cost considerations

A combined laser scanning and drone capture is priced on site size, the number of scan setups, flight area and the deliverables. As a guide, a single-asset scan with supporting drone coverage starts around AUD 4,000, while a full processing plant or a mine pit-with-infrastructure capture runs AUD 8,000-15,000 or more.

The return is rarely the survey fee alone. Removing one to three days of working at heights, avoiding a return mobilisation because the roof was never captured, and giving engineers a complete model to design from usually outweighs the cost of the capture many times over.

Common mistakes to avoid

Mistake 1: Capturing the two datasets on different control

The fastest way to waste a combined survey is to scan one week and fly another with no shared, surveyed control. The clouds then have to be aligned by eye or by best-fit, which introduces error and undermines the accuracy the scanner worked so hard to achieve. Establish one set of GCPs and reference both methods to it.

Mistake 2: Expecting drone data to do the scanner's job

Drone photogrammetry at 20-50 mm cannot drive a flange alignment or a tight tie-in. If a deliverable needs millimetres, it must be scanned. Use the drone for coverage, not for the precision work.

Mistake 3: Ignoring CASA before mobilising

Discovering on site that the asset sits under controlled airspace or beside an aerodrome can ground the drone half of the job. Resolve CASA Part 101 requirements and any site flight restrictions during planning, not on the day.

⚠️ Watch out: Reflective and wet surfaces defeat both methods at once — stainless tanks blind the scanner and confuse photogrammetric matching. Plan extra setups and overlapping flight passes for these areas, or accept that they will need traditional measurement to fill gaps.

Frequently asked questions

How accurate is a combined laser scanning and drone survey?

Accuracy is not uniform across the dataset, and that is the point. The terrestrial-scanned zones hold around 2-6 mm, suitable for alignment, clash detection and dimensional control. The drone-captured zones — roofs, terrain, stockpiles — sit at roughly 20-50 mm for photogrammetry. Combining laser scanning and drones gives you each accuracy where it is appropriate, all on one GDA2020/MGA2020 and AHD frame.

Why not just use a drone for everything?

A drone cannot deliver the few-millimetre accuracy that plant, structures and tie-ins require, and it struggles inside buildings and under congested pipe racks. For coverage and elevated capture it is excellent; for precision it is not a substitute for a terrestrial scanner. The two are complementary, not interchangeable.

Do you need ground control points if the drone has RTK?

RTK on the aircraft greatly improves accuracy and reduces how many GCPs you need, but a handful of surveyed control points remain best practice. They tie the drone cloud to the same frame as the scan data, give an independent accuracy check, and constrain the vertical solution that RTK alone can leave slightly weak.

What CASA requirements apply to the drone component in Australia?

Commercial drone surveying falls under CASA Part 101. The remote pilot must hold a Remote Pilot Licence (RePL) and operate under a Remotely Piloted Aircraft Operator's Certificate (ReOC), within standard conditions such as 120 m AGL and visual line of sight. Work near controlled aerodromes or in restricted airspace needs additional approvals. ISS holds the required certifications.

How long does a combined capture take on site?

For a typical few-hectare industrial site, field capture is usually one to two days — control and scanning, then the drone flights. Processing and registration into a single deliverable adds several days in the office depending on the outputs ordered. Larger mine sites or full plants extend the field time accordingly.

Industrial Spatial Solutions captures, registers and delivers combined laser scanning and drone datasets across mining, processing and industrial sites Australia-wide, on the correct GDA2020/MGA2020 and AHD datum and under full CASA compliance. To scope your site and receive a fixed quote, call us on 0407 057 015 or request a quote and we will recommend the right mix of scanning and drone capture for your asset.