"Do we need LiDAR?" is the most common payload question in mapping — usually asked with a price list open. The honest answer depends on surfaces, vegetation and deliverables, not on which technology is newer. Here's the decision as surveyors and integrators actually make it.
Two ways to make a point cloud
Photogrammetry computes 3D from overlapping photos: cameras are cheap, and accuracy rides on image quality, geometry and georeferencing (see camera payloads and RTK/PPK). It measures what it can see and match — it needs texture, light and a clear view.
LiDAR measures ranges directly with laser pulses, hundreds of thousands per second, each tagged with the sensor's position and attitude from a GNSS-INS. It doesn't care about texture or ambient light, and multiple returns let some energy reach the ground through vegetation gaps — the killer feature.
The decision table
| Condition | Winner | Why |
|---|---|---|
| Open terrain, hard textured surfaces | Photogrammetry | Comparable accuracy, fraction of the cost, orthophoto included |
| Vegetated terrain / forestry ground model | LiDAR | Multiple returns reach ground; photogrammetry maps the canopy top |
| Powerlines, masts, thin structures | LiDAR | Wires are unmatchable in images, trivially resolved by pulses |
| Uniform surfaces (fresh asphalt, sand, snow) | LiDAR | No texture to match — photogrammetry produces mush |
| Visual deliverables (orthos, textured meshes, inspection imagery) | Photogrammetry | LiDAR has no colour of its own |
| Night / poor light | LiDAR | Active sensor |
| Volumes/stockpiles (bare) | Photogrammetry | Industry-standard, cheap, plenty accurate |
| Corridors mixing all of the above | Hybrid pod | Geometry from pulses, interpretation from pixels |
Accuracy: read the whole error budget
Vendors quote sensor precision; deliverables carry system error. For LiDAR: laser ranging (±1–3 cm) ⊕ IMU attitude error × range (a 0.005° attitude error is 1.7 cm at 200 m) ⊕ GNSS trajectory ⊕ boresight calibration. That's why the IMU grade dominates price, why flying lower improves LiDAR accuracy, and why a bargain scanner with a hobby IMU produces expensive noise. For photogrammetry: GSD sets the floor (plan 1–2× GSD horizontal, 2–3× vertical), then image sharpness, overlap geometry (75/75% frontal/side is the modern default), georeferencing quality, and camera calibration stability — the reasons mechanical-shutter cameras and PPS time-stamping are mapping-payload table stakes.
Cost and workflow honesty
- Capex: mapping camera + PPK ≈ low-to-mid four figures. Survey-grade UAV LiDAR ≈ mid five to six figures. That gap buys a lot of extra photogrammetry flights.
- Field time: photogrammetry needs more passes for overlap; LiDAR needs calibration figures and steadier flying. Both live and die on correction workflows done right.
- Processing: photogrammetry is compute-hungry but automated (hours of GPU); LiDAR processing is faster but demands trajectory QC and classification skill. Deliverable QC — control-point residual reports — is what separates professionals in both.
- Data volumes: plan storage and transfer for tens of GB per site either way; corridor LiDAR grows to TB-class programs.
Buy the job's requirement, not the technology. If your work is stockpiles and construction progress: photogrammetry, done excellently, wins on economics. If your contracts say "ground model under canopy" or "wire clearance": LiDAR isn't optional. If you're a service startup, rent LiDAR for the first contracts that need it — utilisation, not capability, justifies the purchase.
Integration notes for either payload
Mapping payloads are the most integration-sensitive cargo a UAV carries: rigid, boresight-stable mounting (no soft isolation between LiDAR and its IMU — they must move together), clean power, PPS time sync throughout, and vibration management at the source per the prop guide. Flight planning matters as much as hardware: consistent height above terrain (not take-off), speed matched to scan/trigger rates, and honest endurance margins so the last line of the survey isn't flown on hope.
Frequently asked questions
Is drone LiDAR more accurate than photogrammetry?
Not inherently. Good photogrammetry with RTK/PPK and ground control achieves 1–3 cm on hard, textured surfaces — comparable to survey-grade UAV LiDAR. LiDAR wins decisively where photogrammetry structurally fails: vegetation, uniform surfaces, thin structures like wires, and low light.
Why is drone LiDAR so much more expensive?
A survey-grade system bundles a laser scanner with a tactical-grade IMU and GNSS in a calibrated assembly — the inertial unit is often half the system cost. Every point's accuracy depends directly on knowing the sensor's position and attitude at the moment of each pulse.
Can one drone carry both LiDAR and a mapping camera?
Yes, and hybrid payloads (LiDAR + RGB) are increasingly standard: LiDAR provides geometry through vegetation, imagery provides texture and colour for interpretation and deliverables. Budget for the combined mass on your payload fraction and for a workflow that fuses both datasets.