Terrestrial LiDAR vs Mobile LiDAR

TL;DR

Terrestrial LiDAR sits a scanner on a tripod and captures the world one fixed station at a time, trading speed for the tightest accuracy a 3D scan can give — roughly 2–6 mm on structure. Mobile LiDAR straps the same class of sensor to a vehicle, backpack or trolley and captures kilometres while you walk or drive, at the cost of accuracy that drifts to 10–50 mm depending on the positioning solution behind it. For a mill trunnion or a flange face you want the tripod; for a 4 km haul road or an entire processing plant on a shutdown clock, the mobile system earns its keep.

Key takeaways

• Terrestrial laser scanning (TLS) holds the accuracy edge — a Leica RTC360 or FARO Focus delivers 2–6 mm range noise per station, against 10–50 mm for most mobile systems whose error is governed by their GNSS/IMU trajectory rather than the laser itself.
• Mobile LiDAR's advantage is throughput: a backpack or trolley unit such as a Leica Pegasus or NavVis VLX captures 1–2 km of corridor or a full plant level in the time a tripod scanner manages a handful of static setups.
• Indoors and in covered plant where GNSS is unavailable, mobile systems rely on SLAM (simultaneous localisation and mapping), which is fast but accumulates drift over long loops; closing the loop or dropping survey-grade control targets is what keeps a mobile job honest.
• Both register into GDA2020 / MGA2020 with AHD heights once tied to control, and neither replaces a total station for ±2 mm dimensional control on a baseplate or rail — that stays a contact-measurement task.
• On cost, the field day rates are comparable; the real difference is data volume and processing time, where a sprawling mobile capture can generate 100 GB+ and several days of registration and classification work.

What terrestrial LiDAR actually is

Terrestrial laser scanning is static, tripod-mounted capture. The instrument occupies a fixed station, sweeps a near-spherical field of view, and records millions of range-and-angle measurements before you move it to the next setup. Because the scanner is stationary and levelled, its internal geometry is the dominant accuracy term — there is no moving platform introducing trajectory error.

A modern phase-based scanner such as a Leica RTC360 or a FARO Focus Premium captures around 2 million points per second with single-point range noise in the 1–4 mm band and registers adjacent stations either on overlapping geometry (cloud-to-cloud) or on surveyed targets. Stations are tied to a control network so the whole job sits in MGA2020 with AHD levels. The deliverable is a dense, low-noise point cloud good enough to model pipework, extract steelwork, run clash detection against a retrofit design, or measure a tank shell for ovality.

What mobile LiDAR actually is

Mobile LiDAR (also called mobile mapping or kinematic scanning) puts the laser on something that moves — a survey vehicle, a backpack, a push trolley, or a handheld unit. Instead of occupying fixed stations, the sensor scans continuously while the operator covers ground, and the system reconstructs the cloud by knowing precisely where the scanner was at every instant.

That position comes from one of two engines. Outdoors, a tightly coupled GNSS receiver and inertial measurement unit (IMU) compute a trajectory in GDA2020; the cloud is only as accurate as that trajectory, so accuracy degrades under canopy, beside high walls, or in any GNSS-shadowed cut. Indoors and in covered plant, the system switches to SLAM, matching successive scans to build its own internal map without GNSS. SLAM is remarkably quick but drifts over distance — walk a long open corridor and the start and end can disagree by centimetres unless you close a loop or anchor to control targets.

SLAM (simultaneous localisation and mapping) lets a mobile scanner work where there is no GNSS by continuously matching what it sees to what it has already mapped. Its weakness is cumulative drift, so disciplined mobile surveying drops surveyed control targets at intervals and closes capture loops to constrain that drift.

Accuracy and capability: side by side

Key point: the headline difference is not the laser — both classes use comparable sensors. It is what carries the laser. A stationary tripod removes the largest error source; a moving platform reintroduces it as trajectory uncertainty. That single fact decides most jobs.

Speed and the shutdown clock

The reason mobile LiDAR exists is time. On a planned outage at a Pilbara processing plant or a Gladstone alumina refinery, the survey window is measured in hours and the plant is losing production every one of them. A surveyor walking a NavVis VLX or a Leica BLK2GO through a plant level can capture in a single pass what would take a tripod scanner a day of leapfrogging static setups around the same structure.

That speed is the whole argument for as-built capture of large brownfield assets, conveyor galleries, multi-level steelwork and access-restricted areas where standing still is itself a hazard. The trade is accuracy: if the deliverable feeds a retrofit clash-detection model where pipe spools must drop in to a few millimetres, the mobile cloud's 10–50 mm may not be defensible, and a hybrid capture — mobile for bulk context, terrestrial for the accuracy-critical tie-in zones — is the honest answer.

Cost comparison: 2026 Australian rates

Field day rates sit in the same ballpark, so cost is rarely decided by the rate card. It is decided downstream. A large mobile capture generates far more data — 50–150 GB is routine on a plant or corridor job — and that data needs trajectory processing, deskewing, control adjustment and classification before it is usable. On a tight, high-accuracy terrestrial job the processing is lighter, but you have paid for the extra field days the static method demands. Match the method to the deliverable and the total cost usually falls where it should.

Use-case scenarios

Mill, kiln and rotating-equipment alignment

Recommendation: neither — use a total station, supported by terrestrial LiDAR for context. At the ±2 mm a SAG mill trunnion or kiln tyre demands, no scan-based method is the primary tool. Terrestrial LiDAR can document the surrounding structure and clearances; the critical alignment stays a contact-measurement task.

Processing-plant as-built for retrofit design, WA or QLD

Recommendation: terrestrial LiDAR, or a hybrid. Pipe routing, nozzle orientation and spool generation need 2–6 mm geometry for clash detection. Use mobile LiDAR to capture bulk context quickly, then terrestrial scans in the tie-in zones where the new steel must actually fit.

Conveyor corridor and haul-road as-built

Recommendation: mobile LiDAR. Kilometres of linear asset are exactly where a vehicle or backpack system pays for itself. A few centimetres of absolute accuracy is ample for corridor design, drainage and clearance checks, and the single-pass capture is a fraction of the static effort.

Underground and GNSS-denied capture

Recommendation: mobile LiDAR with SLAM. Decline drives, underground workings and covered plant have no satellite signal. A SLAM backpack or a drone-mounted Emesent Hovermap maps them rapidly — just constrain the drift with control targets and loop closures on longer runs.

Heritage façade, tank shell ovality or structural deformation

Recommendation: terrestrial LiDAR. Low noise and repeatable station geometry make TLS the right tool for precise surface measurement and epoch-to-epoch comparison, where a moving platform's trajectory noise would swamp the signal you are trying to detect.

The decision table

Combining both on one job

The two methods are complements far more often than rivals. A common WA and QLD pattern on a large brownfield as-built is to walk a mobile system through the whole asset first — galleries, levels, access ways — to build a complete spatial context fast, then bring a terrestrial scanner back to the handful of zones where new equipment, pipe spools or structural ties must fit to a few millimetres. The mobile cloud carries the coverage; the terrestrial scans carry the accuracy; both register onto the same control network so the model is seamless. It costs more than either alone, but it removes the compromise — and on a single mobilisation that serves engineering, maintenance and operations at once, it is usually the cheapest path to a model everyone can trust.

Honest limitations

Terrestrial LiDAR: it is slow. Each setup takes minutes, occlusions force extra stations to see behind structure, and a large plant can run to many days of fieldwork. It also has no native colour beyond the scanner's own imagery, and standing in active areas to capture them is its own safety problem.

Mobile LiDAR: absolute accuracy is the constant caveat. Outdoors it is hostage to GNSS quality — multipath beside steel, signal loss in cuts and under canopy all degrade the trajectory. Indoors, SLAM drifts over distance, so undisciplined capture without control or loop closure produces a cloud that looks complete but does not hold dimensionally. Data volumes are large and processing is heavier than most clients expect.

Both share the same registration reality: a scan is only as good as the control it is tied to. Whichever method captures the points, the survey discipline behind the control network is what makes the deliverable defensible.

Frequently asked questions

Is terrestrial LiDAR more accurate than mobile LiDAR?

Yes, as a rule. Terrestrial scanning from a fixed tripod typically holds 2–6 mm on registered structure, while mobile systems sit in the 10–50 mm band because their accuracy is governed by the GNSS/IMU trajectory or SLAM solution carrying the moving sensor, not by the laser itself.

When is mobile LiDAR the better choice?

Whenever coverage and speed dominate over millimetre accuracy — long corridors, haul roads, conveyor galleries, and full plant as-builts on a tight shutdown window. It is also the practical choice underground and in any GNSS-denied space, where a SLAM backpack or drone-mounted scanner maps areas a tripod could never cover in the time available.

Can either method replace a total station for alignment?

No. At the ±2 mm a mill trunnion, kiln tyre or crane rail demands, scan-based methods are documentation tools, not the primary measurement. Precision alignment stays a contact-measurement task with a total station; LiDAR captures the surrounding structure and clearances around it.

Does mobile LiDAR work indoors without GNSS?

Yes — that is one of its strengths. Indoors it switches to SLAM, building its own map by matching successive scans. The catch is drift over long runs, so a careful operator drops surveyed control targets and closes capture loops to keep the cloud dimensionally honest.

Which produces more data to process?

Mobile LiDAR, usually by a wide margin. A continuous plant or corridor capture commonly produces 50–150 GB and needs trajectory processing, control adjustment and classification before it is usable. Terrestrial jobs are lighter to process but demand more field time to acquire.

Ready to choose the right method?

Choosing the wrong platform wastes a mobilisation and produces a cloud you cannot rely on — a fast mobile capture that drifts past the tolerance your retrofit needs, or a painstaking terrestrial job on a corridor where a single-pass walk would have done in an hour. Industrial Spatial Solutions runs both terrestrial and mobile LiDAR across mining, plant and civil sites Australia-wide, and we recommend the method — or the hybrid — your site, accuracy specification and shutdown window actually require. For work demanding ±2 mm dimensional control, we pair scanning with total-station measurement so nothing slips through the tolerance gap. Call us on 0407 057 015 to talk through your asset, your accuracy requirement and your timeline, and we will quote the right approach.