3.6 Scan Plan Tool

Step 1: Scope and Constraints

Answer these before site arrival. They are not optional, and several of them determine choices in Step 2 that cannot be undone once scanning starts.

• Total scan area in square feet or square meters, by floor or zone.
• Deliverable: point cloud, 3DGS, BIM model, ATIS.cloud project, measurement-grade survey, or a combination. This decides which software you fuse in, and the two have different limits.
• Required absolute accuracy: 3DGS visualization with no absolute reference, 3 cm RMSE georeferenced with RTK or control points, or survey-grade with extensive control.
• Baseline or rescan. Rescans must match prior session structure: same targets, route, boundaries, and capture settings.
• Access window: hours per day, total days, escort requirements, after-hours availability.
• Access restrictions: secure zones, sensitive equipment, photography limits, adhesive limits. Adhesive is the documented target mounting method, so an adhesive prohibition changes the plan rather than inconveniencing it.
• Devices: L2 Pro for large facades and long range, K2 for indoor and close-quarters work, PortalCam where 3DGS is the only deliverable. The PortalCam produces no georeferenced output, so it cannot serve a job with an absolute-accuracy requirement.
• Site contact and the escalation path if something goes wrong during the scan.

Step 2: Decisions That Lock Before You Press Record

Some capture choices cannot be changed mid-scan and cannot be applied afterward in processing. Getting one wrong means a recapture, not a reprocess. Settle each of these in the plan rather than on site.

Step 3: Route and Session Planning

Turn the scope into an executable route: session boundaries, startup points, and the order of areas.

Fusion Limits Are Not the Same in Both Applications

A session is one continuous scan from startup to completion. Multiple sessions join afterward with Map Fusion, and the two applications impose different limits. Plan segments against the one you will actually process in. If the project needs both point cloud and 3DGS from the same captures, plan to the stricter figure in every row.

Alpine plans at 12 to 15 minutes per capture. That sits under both limits and leaves room for the coverage you did not anticipate.

Session Boundary Rules

• Each session starts and ends at the same physical location with at least 50 ft (15 m) of overlap into the neighbouring session. Less than that aligns inconsistently.
• Place boundaries at structural features that are easy to return to and rich in texture: cross-aisles, distinctive doorways, intersections. Avoid boundaries in low light, on stairs, in narrow corridors, and in front of reflective surfaces.
• No segment may sit entirely inside another. Every segment needs unique coverage beyond the shared zone or the fusion cannot resolve the spatial relationship.
• Multi-floor: one session per floor, joined at stairwells rather than elevators.
• Multi-day: each day is at least one session. Plan the day-to-day overlap explicitly, and schedule it at a similar time of day so lighting matches.
• Fusion and enhancement jobs finish within about 2 hours. Segments captured across a major lighting change stitch with visible seams that no processing setting removes.

Startup Point

• Initialize in a feature-rich area with the device on flat, stable ground and nothing obstructing the LiDAR. Avoid open plains, large glass surfaces, and anywhere with many moving objects.
• Choose a point you can return to cleanly. The final loop closure needs the same position and a return line of sight within about 40 degrees of the original, followed by a stationary hold before you stop.
• Without RTK, the initialization position becomes the opening viewpoint of the delivered 3DGS model. Pick the space you want a client to see first.
• For RTK, reach Fixed outdoors under clear sky, walk at least 33 ft (10 m) while Fixed, then carry the device to the indoor start.

Step 4: Georeferencing and Control Point Placement

If the deliverable needs absolute coordinates, plan control placement before the scan. Number and spacing depend on device and area.

Control Point Spacing

Space control so that no continuous run without a fix exceeds the device's tolerance. The maximum figures are published; the preferred figures are Alpine practice and buy margin.

Surveying the Control

• Surveyed coordinates are what make a control point georeference anything. Without them it is an anchor target with a known location: useful to SLAM, useless for absolute coordinates.
• Use the same coordinate system as the RTK base where RTK is in play. Mixing systems forces a transformation and loses precision.
• Document the coordinate system explicitly: datum, projection, zone. Future rescans must reuse it.
• Match the survey method to the accuracy need: total station for survey-grade, RTK rover for general georeferencing, building geometry for relative-only deliverables.
• Use the standard control point base when marking, and set the device down rather than marking handheld.

Coverage Pattern

• Distribute control as a polygon, never a line. Collinear control does not constrain rotation about that line, so the model can twist around it while every residual still looks acceptable.
• The project needs at least 3 absolute control points across all sessions, not collinear. This is a whole-project requirement for coordinate transformation, and it is easy to satisfy inside each session and still fail across the set.
• Place one near the startup point and one near the final loop closure point.
• Add 20 percent redundancy over the minimum, so a marking or survey failure does not break the dataset. Alpine practice.

Step 5: Anchor Target Placement

Anchor targets and control points do different jobs. Control provides georeferencing; anchor targets give SLAM high-contrast features to track where natural geometry is uniform. Some projects need both, some only one, and one physical marker can serve both roles if you survey it.

When to Place Anchor Targets

• Long featureless corridors over 200 ft (60 m).
• Dark environments where visual tracking is unreliable.
• Repetitive geometry: data centers, uniform-racked warehouses, tunnels.
• Multi-session projects where fusion depends on shared anchors.
• Rescan engagements where target continuity drives change detection.

Target Spacing

Plan placement every 100 to 150 ft (30 to 45 m) along the path. Closer in higher-challenge zones such as very dark or very repetitive areas, wider where some natural features exist. Alpine practice, not a manufacturer figure. In repetitive environments, place them at irregular intervals rather than regular ones, so the spacing itself does not repeat.

Placement Checklist

• Scanner can approach within 3 to 6 ft (1 to 2 m) and circle each target during the scan.
• Each target is on a stable surface, not a vibrating panel, near vents, or on a flexible substrate.
• Each target's position is documented with a photo and a room or rack reference before the scan.
• For rescans, target positions match the baseline documentation.
• Adhesive policy verified with the site operator before placement.
• Removal plan in place: targets and tape collected before site exit, and the removed state photographed.

Step 6: Time and Resource Estimation

Estimates that miss reality by more than 25 percent cause access overruns, fatigue, or incomplete coverage. Build from per-zone baselines rather than a single area ratio.

Pace by Device

Coverage rate follows walking speed, and the published limits differ by device. The area figures below assume the slow end of the L2 Pro and K2 band, which is what indoor work requires. A PortalCam job runs at roughly half that pace throughout and should be budgeted accordingly.

Baseline Scanning Time

These assume the indoor pace above, single-pass coverage, and no significant transitions or environmental challenges. Use them as starting points, then apply the adjustment factors.

Adjustment Factors

• Multiply by 2.5 to 3.0 for three-pass coverage. The published three-pass method walks the same routes three times, so the cost is close to arithmetic. Apply it wherever the deliverable needs ceiling and floor detail or high 3DGS quality. Alpine estimate of a published method.
• Multiply by 1.5 to 2.0 where there are many transitions: multi-floor, many small rooms, stairwell-heavy layouts.
• Multiply by 1.5 to 2.0 for any site you cannot clear before scanning. Occupied offices and active construction require repeated passes and longer dwell at transitions.
• Add 30 to 60 minutes per session for setup, startup-point selection, RTK Fix, and final checks.
• Add target time: 3 to 5 minutes per anchor target, 10 to 15 minutes per surveyed control point.
• Add escort coordination at secure sites: 15 to 30 minutes per access window.
• Plan a 5-minute rest every 45 to 60 minutes of active scanning. Fatigue degrades posture before you notice it, and hour-three posture produces less recoverable data than a planned break.

Step 7: Risk Review

Review the plan against known failure modes before arrival. Catching them in planning is far cheaper than after the fact.

• Locked decisions settled: capture mode, layered optimization state, and which application will fuse, all recorded in the plan rather than left to the operator on site.
• Segment plan fits the fusion limits for the application you will actually process in, including the 20-minute per-capture limit if point cloud output is in scope.
• Session boundaries verified: each with at least 50 ft (15 m) of overlap, placed on texture-rich features, with no segment sitting inside another.
• Control placement verified: spacing within device tolerance, polygon distribution, at least 3 non-collinear absolute points across the project, redundancy added, coordinate system documented.
• Anchor target inventory confirmed: enough targets, adhesive permitted, removal planned.
• Compound-challenge zones identified such as dark and featureless, or reflective and outdoor, with extra time and specific technique assigned to each.
• Site preparation agreed with the client: all doors and passages open in advance, all fixed lights on, curtains and doors fixed for the duration, movable objects removed or secured, lightweight items near air vents secured, mirrors and glass covered or noted.
• Site contacts current: primary, escalation, and after-hours numbers confirmed in the last 30 days.
• Access window fits the estimate: scan time plus setup, breaks, and contingency, with a 25 percent buffer.
• Deliverable expectations documented in writing: format, accuracy, publishing, BIM handoff, timeline.
• Equipment ready: devices charged, spare batteries charged, lenses clean and fog-free, mobile devices charged with the field app installed and tested, processing machine has free disk space.