Products · Intelligent Payloads and Accessories

LiDAR Integration Assessment

LiDAR integration is a mass, power and accuracy question in that order, and each one can independently rule out a platform.

3 of 10Method note

Overview

Mass and mounting rigidity come first: a flexible mount destroys the geometric accuracy the sensor is bought for.
Power budget and duty cycle determine whether the platform can carry it for the length of your mission.
Trajectory quality sets point cloud quality. A better IMU can matter more than a better scanner.
Vibration isolation is a two-sided problem: isolate the sensor, but do not let the mount resonate.

LiDAR answers a different question from photogrammetry

Photogrammetry reconstructs surfaces by matching features across overlapping images, and it works well on surfaces that are visible, textured and static. LiDAR measures range directly, which means it does not need texture, it works in darkness, and — critically — it can return multiple elevations along a single pulse as it passes through gaps in vegetation.

That last property is the reason LiDAR is chosen, and it is the reason it costs more. Under a closed canopy, no amount of camera resolution produces a ground surface, because the camera never sees the ground. LiDAR returns the canopy surface, intermediate structure and the ground as separate returns, which is what produces a terrain model beneath vegetation.

The trade is real and should be stated plainly: LiDAR adds mass, power, cost and a substantially heavier calibration, georeferencing and processing workload. Where the deliverable does not require penetration or does not require an accurate surface under cover, photogrammetry is usually the more economical answer and the better one.

Specifying LiDAR by the output it has to produce

LiDAR is specified by point density, range accuracy and the number of returns it can resolve per pulse. Point density on the ground is the output-side specification and the one that maps to the deliverable: how many points per square metre are needed to represent the surface or the structure at the required fidelity.

Point density is a function of pulse rate, scan pattern, flight speed and altitude — which means it is a flight planning variable as much as a sensor specification. A sensor quoted at an impressive pulse rate delivers that density only if the flight plan is built to realise it. Assess the sensor and the plan together, and check the achievable density at the speed you intend to fly.

Returns matter for vegetated terrain. A single-return system records one elevation per pulse; a multiple-return system can record canopy, understorey and ground separately. For terrain modelling under vegetation, multiple returns are the requirement, not a refinement. Confirm what the system actually records rather than what the brochure headline says.

Georeferencing is where LiDAR projects succeed or fail

A LiDAR point cloud is only useful if it is accurately positioned, and positioning depends on the combination of a GNSS solution, an inertial measurement unit, and the calibration between the sensor and that navigation solution. The airborne part is only half of it; the other half is how the trajectory is processed and how the point cloud is corrected afterwards.

This is the part most often underestimated in planning. It requires survey discipline — a properly executed calibration, control or check points, and processing that is validated rather than assumed. A sophisticated sensor flown without that discipline produces a point cloud that looks correct and is not, and the error is usually discovered when someone tries to measure from it.

Where deliverables must be compared over time — change detection, volume monitoring, growth assessment — consistency of the georeferencing approach matters as much as its absolute accuracy. Two epochs processed differently are not comparable even if each is individually accurate. Our multi-epoch survey data management note covers this discipline.

Integration on the aircraft

The four standard checks apply, and for LiDAR each is more demanding than for a camera. Mass is higher. Power draw is higher and continuous. The data interface has to carry a substantial stream, and onboard storage usually has to hold it because downlinking a full-density stream in real time is frequently impractical.

Mounting is also more sensitive. LiDAR is a measuring instrument, and its accuracy depends on knowing where it is and where it is pointing relative to the navigation solution. A mount that flexes, or that is not rigidly located relative to that solution, introduces error that no amount of processing removes. Treat the mount as part of the instrument.

Confirm that the platform publishes allowable payload mass, centre-of-gravity envelope, mounting provisions and available power. Where it does, integration is scoped work. Where it does not, it is development. Our payload interface and platform compatibility note and payload integration methods describe how we handle both.

Assessing for your programme

The assessment sequence is: confirm the deliverable actually requires penetration or an accurate surface under cover; write the requirement as point density on the ground; confirm return capability against the vegetation; confirm the georeferencing and validation approach; then confirm the four integration properties.

Where the answer to the first question is no, we will say so. LiDAR is frequently specified where photogrammetry would serve, and the difference in cost and workload is substantial. Where the answer is yes, the sensor is only part of the system and the survey discipline around it is what produces the deliverable.

Send your area, vegetation cover, required deliverable and accuracy, platform and flight speed to our engineering team. We will assess whether LiDAR is required, what density the flight plan must deliver, and what the integration and validation work involves.