3D mapping used to mean a tripod, a sequence of fixed stations, and a slow march from one setup to the next. A handheld LiDAR scanner changes the shape of the job: you walk the space and the device builds a registered point-cloud map in real time, stitching your path together as you move. A multi-level building or a long corridor that would take a day of station setups can be covered in a fraction of the time, and the output is the same kind of measurable point cloud downstream software already understands.
How SLAM mapping works
The enabling algorithm is SLAM: simultaneous localisation and mapping. The scanner has to solve two problems at once — where am I, and what does the space look like — and it solves them together, fusing LiDAR returns with inertial measurements to estimate its own trajectory while accumulating points along it. The Lixel K2 does this at 200,000 points per second, colourising the cloud from its cameras as you go, with built-in RTK tying the map to real-world coordinates outdoors.
Because the map is built from a continuously estimated trajectory, small drift errors can accumulate on long walks. Good SLAM technique manages this with loop closures: routes planned so you return to previously scanned areas, letting the processing software recognise the revisit and pull accumulated drift back out of the map. Walk steadily, avoid featureless dead ends, close loops, and the registration largely takes care of itself in LixelStudio afterwards.
The deliverables are standard. LixelStudio exports the registered map to the formats downstream software expects — LAS/LAZ for GIS and survey tools, E57 for exchange, and from there RCP for Revit and AutoCAD — plus mesh and 3DGS outputs from the same capture when the visual layer matters. Nothing about a handheld capture locks you into a proprietary ecosystem; the walk is faster, the data is conventional.
Where handheld LiDAR mapping wins
- Interiors and multi-level buildings. Walk room to room and up stairwells without resetting a tripod each time; the stairwell itself becomes the connection between floors.
- Corridors and infrastructure. Tunnels, walkways, and linear assets map naturally as a continuous path rather than a chain of dozens of overlapping stations.
- Occupied and access-constrained sites. A handheld walk fits the short window a site manager can actually give you, and low-light plant rooms and voids are no obstacle to LiDAR.
- Outdoor and site-scale work. With a longer-range device, whole sites and facades come within reach in a single session, geo-referenced via RTK.
Where handheld does not win: tolerance-critical work specified in millimetres. Tripod terrestrial scanners remain the accuracy benchmark for contained, stable spaces, and the honest comparison is laid out in handheld SLAM vs tripod scanner.
How accurate is handheld LiDAR mapping?
Two numbers to hold onto for the K2: 1 cm relative accuracy within the cloud, which is what measurement and modelling consume, and around 3 cm RMSE absolute with RTK, which is what matters when the map must sit correctly in a coordinate system or merge with other datasets. Real-world results still depend on capture technique, loop closures, and processing settings; the specification is the ceiling, not a guarantee against poor technique. For deliverables with stated tolerances, plan control points and verification measurements before the walk, and agree the requirement with whoever consumes the data. For most as-built documentation, renovation measurement, and indoor mapping in Singapore, centimetre-level is comfortably sufficient.
How long does a LiDAR scan take?
Capture runs at walking pace, so scan time roughly tracks how long the space takes to walk attentively. A showflat or small office is a session of tens of minutes; a full commercial floor might be an hour; a multi-storey building is typically a day or less on site. The K2's hot-swappable batteries run about 1.5 hours each and swap without stopping the capture, and 512 GB of onboard storage means a full rental week of capture without offloading anxiety. In practice the schedule bottleneck is rarely the scanning; it is site access, lift waits, and locked rooms, so walk the route once with the site contact before you start recording. Budget separately for processing: LixelStudio registration and colourisation is workstation time after the walk, usually same-day. The practical upshot for renters is that a single week covers several sites, which is what makes the weekly rate work.
Choosing the scanner: K2, K1, or L2 Pro
For most interiors and mixed sites, the Lixel K2 (S$2,300/week) is the recommended mapper: about 1.2 kg in hand, built-in RTK, and outputs covering point cloud, mesh, and 3DGS from one walk. The Lixel K1 (S$2,200/week) is the value pick when the job is straightforward interior mapping and the K2's extras are not needed — the differences are itemised in K2 vs K1. When the work is large, outdoor, or corridor-heavy, the long-range Lixel L2 Pro (S$3,500/week) reaches up to 300 m and suits infrastructure-scale mapping.
3D mapping in Singapore: what rental costs
All rates are weekly, GST-exclusive, dry rental: you capture, you keep the data. First rentals include a one-time onboarding session, S$250 if you collect the kit yourself or S$400 with delivery to site, so your first walk is not also your first lesson. Against commissioning each survey separately, the arithmetic favours rental as soon as a week covers more than one site — and at walking-pace capture, it almost always does. Availability and terms are on the rental page, or send us your site type, area, and dates and we will recommend the K2 or L2 Pro and confirm a quote.