For design firms and renovation teams adopting digital measurement, a common inflection point arrives after the first successful scan. The point cloud is on screen — dense, detailed, complete. The immediate question is practical: how does this data become a usable CAD drawing?

This guide answers that question in full. It covers the complete SHARE3DCAM Scan to CAD workflow from on-site capture through to the final DWG export, with step-by-step instructions, software screenshots, and honest guidance on what each stage delivers — and what it does not.

One clarification upfront: point cloud to CAD is not a one-click conversion. No software — including SHARE PointClouds Studio — produces finished, code-ready construction drawings at the push of a button. What it does provide is a measurable, always-accessible digital record of the actual site, paired with a set of tools that help teams produce CAD drawings faster and with fewer return visits than traditional measurement methods allow.

Below is the complete workflow, stage by stage.

The Four Stages of Point Cloud to CAD

Every Scan to CAD project — whether it is a 60 m² apartment or a 2,000 m² commercial floor — follows the same four-stage structure:

1. Capture the space with a handheld SLAM LiDAR scanner
2. Process and prepare the point cloud in SHARE PointClouds Studio
3. Choose the CAD path — auto-generation or RCS-based manual drafting
4. Refine in CAD software (AutoCAD, DraftSight, BricsCAD, or equivalent)

Stage 1: Capture the Space

Everything downstream depends on what you capture on site. Walk the space with a SHARE SLAM S20 or SHARE SLAM S100 handheld scanner — a single walkthrough, no tripods or targets required for most indoor jobs.

The result is a dense 3D record of the site: walls, door openings, windows, beams, irregular corners, alcoves, ceiling variations, and exposed service runs. Elements that would require individual measurement in a conventional workflow are captured as part of the whole.

Why this matters downstream: In a traditional workflow, a missed measurement — or a client question three weeks later about a dimension not recorded on site — means scheduling a return visit. With a point cloud, the data is available from any workstation. The scan does not eliminate site visits entirely, but it dramatically reduces the situations where missing information forces a return trip.

Key practices at this stage:

• Maintain a steady walking pace — 0.5 to 1.0 m/s. Avoid running, frequent stopping, or abrupt speed changes.
• Create loop closures wherever possible. Returning to the starting point or crossing a previously scanned area gives the SLAM algorithm additional constraints to correct accumulated drift.
• At doorways and narrow corridors, slow down and keep both spaces in view during the transition to maintain feature continuity.
• Glossy floors, large mirrors, and floor-to-ceiling windows present challenges for LiDAR. Scan from multiple angles or introduce temporary visual markers where feasible.

• Review the scan preview before leaving site. If a wall appears distorted or a room is clearly incomplete, rescan while still on location. These issues cannot be corrected in post-processing.

Stage 2: Process and Prepare the Point Cloud

Back at the workstation, launch SHARE PointClouds Studio. Import the scan data folder, select the processing parameters — the defaults work for most projects — and begin processing. Duration depends on scan length and hardware, but typical jobs complete in a few minutes.

Once processing completes, the full point cloud appears. The step most operators overlook comes next:

Clean the point cloud before you begin drafting.

A few minutes of cleanup here will save considerably more time during the drawing stage:

• Rotate and level. Align walls with the viewport axes for easier navigation. Use the rotation controls, the leveling tool, or reset the view as needed.
• Crop out unwanted data. Every scan captures noise — reflections from windows, passersby in corridors, equipment temporarily placed in rooms. Use the polygon, rectangle, or 3D crop tools to isolate the area you intend to draw.
• Save a clean working copy. Keep the raw scan data archived, but work from a cropped version containing only the relevant rooms.

What cleaning cannot address: If a wall was occluded by immovable furniture, or a corner fell within a scan shadow, those gaps will persist. Point clouds represent captured data — they do not interpolate missing geometry. Plan to fill these gaps during the drafting stage, or better, identify and rescan them while still on site.

Stage 3: Choose the CAD Path

From here the workflow splits into two routes, both supported by SHARE PointClouds Studio:

Many teams use both: Path A for a rapid floor plan draft, Path B for detailed sections or complex areas where precision is critical.

Path A: Auto-Generated Plans and Sections

This route uses wall detection within SHARE PointClouds Studio to create a DWG/DXF line drawing that can be exported and refined in external CAD software.

Floor Plan Generation

When to use it: Small to medium residential or commercial spaces (approximately up to 150 m²), rectangular layouts, clearly defined wall boundaries, and minimal occlusion.

1. With the cleaned point cloud selected, open CAD Drawing → Floor Plan Generation.
2. Set the section height. The default is 1500 mm — the horizontal plane at which the software detects walls. This works for most interior walls. Adjust for unusual configurations.
3. Enter the wall thickness if known (measure a few reference points on site or from the point cloud). Accurate input improves wall offset precision.
4. Click confirm. The software runs wall detection; progress is shown in the bottom-right panel.
5. Once complete, click Edit Drawing. The generated line work appears overlaid on the point cloud.

Correction is the critical step.

Auto-detected wall lines will be close, but rarely exact. Use the built-in drawing tools — line, rectangle, trim, extend — to correct wall positions, close gaps, and adjust door and window openings. Verify every adjustment against the point cloud underlay. A wall offset of 15 cm is immediately visible in overlay comparison.

When the drawing is ready, click Export Drawing. Configure the paper size, drawing information, file name, and export location.

Uploading pasted image...

What the export contains: A ZIP archive. After extraction, the folder includes the DWG and DXF files with the generated line work, plus the point cloud underlay. Open the DWG in AutoCAD to view the drawing frame, the draft line work, and the point cloud background.

Section Generation

When to use it: Large or multi-zone spaces, projects requiring cross-sections at specific positions or heights, and situations where structural relationships need to be assessed in elevation.

Unlike floor plan generation — which slices horizontally at a fixed height — section generation places a vertical or angled cut plane at any user-defined position. This is useful for checking wall profiles, ceiling heights, door openings in elevation, and structural relationships not visible from a plan view.

1. Open CAD Drafting → Section Generation.
2. Click Create under "Draw Section."
3. Click twice in the viewport to define the section line — once for the start point, once for the end point.
4. Adjust section thickness. Wider sections capture more context but introduce additional noise. Narrower sections yield cleaner profiles. Iterate as needed.
5. Click confirm to run the extraction.
6. Export using the same procedure as floor plans — ZIP archive with DWG/DXF and point cloud underlay.

Important limitation: When a section export is opened in AutoCAD, the point cloud underlay displays as a 2D image — not a navigable 3D RCS point cloud. The wall profile is visible and can be traced, but the view cannot be rotated or zoomed through in three dimensions. For full 3D navigation within AutoCAD, use Path B.

Summary of Path A: SHARE PointClouds Studio provides a draft faster than drawing from a blank file. However, the output is a starting point — it requires review, correction, and completion. The point cloud underlay is the reference against which accuracy is verified.

Path B: RCS Point Cloud Import for Manual Tracing

When a project involves angled walls, curved elements, irregular openings, or geometries that do not fit automated wall detection well, skip the auto-generation and proceed directly to RCS-based manual drafting.

Path B is also the appropriate choice when point cloud data must be viewed from multiple angles in 3D while drafting, or when project accuracy requirements justify tracing each element individually rather than correcting auto-generated approximations.

Procedure:

1. In SHARE PointClouds Studio, right-click the point cloud and select Export.
2. Choose RCS File (.rcs) as the output format. Set the file name and save location, then confirm. Export time scales with scan size.
3. Open AutoCAD. Create a new drawing or open the existing project file.
4. Navigate to Insert → Attach. Confirm the file type filter reads "Autodesk Point Cloud" and select the RCS file.
5. Click in the drawing area to set the insertion point. The point cloud loads as a 3D reference.

Working with RCS data in AutoCAD:

• Right-click the point cloud and use Point Cloud → Crop → Rectangular to isolate specific rooms or areas.
• Switch between top, front, and isometric views to assess wall profiles and structural relationships.
• Use standard AutoCAD drawing tools — line, polyline, rectangle, arc — to trace directly over the point cloud.
• For dimensions requiring confirmation, return to SHARE PointClouds Studio to measure within the point cloud, then adjust the CAD drawing accordingly.

The value of this approach: Every line drawn is placed against the actual site geometry. Whether a wall runs straight or a corner is square is visible in the data — not a matter of interpretation from field notes. For renovation projects where existing conditions are irregular, this alone changes the quality and reliability of the final drawing.

Stage 4: Refine in Your CAD Software

Whether the starting point was Path A (auto-generated draft) or Path B (RCS manual tracing), the final stage is the same: open the drawing in the team's preferred CAD environment — AutoCAD, DraftSight, BricsCAD, or another compatible platform — and complete the drafting work.

This stage is not a formality. The point cloud provides a reference, not a finished deliverable. The following steps remain essential:

• Verify all critical dimensions against the point cloud
• Add annotations, dimensions, and hatching
• Layer and organize the drawing according to project standards
• Cross-check key measurements — particularly structural elements, door and window openings, and any features with defined construction tolerances
• Add details the scanner could not capture clearly: cabinet interiors, areas behind fixed furniture, and extremely narrow gaps

If a dimension appears inconsistent — or if the point cloud is noisy in a particular area — verify with a laser distance meter or physical measurement. No digital dataset replaces judgment at tolerance-critical positions.

Which Path Should You Choose?

There is no single correct answer. The table below provides a decision framework based on project characteristics:

The guiding principle: match the method to the complexity, not the reverse. If achieving acceptable accuracy through auto-detection correction takes longer than tracing manually, the efficiency argument for Path A disappears. Conversely, manually tracing a simple, rectangular room is unnecessary effort when floor plan generation produces an adequate draft in far less time.

FAQ

Does SHARE3DCAM's point cloud to CAD generate complete construction drawings automatically?

No. The software produces a CAD draft — floor plans or section profiles with a point cloud underlay — exported as DWG/DXF for continued work. The drawing still requires human review, correction, annotation, and finalization to meet project standards. The tool addresses two common bottlenecks: starting from a blank file, and returning to site for missed dimensions.

What are the main methods for bringing point cloud data into CAD?

Three methods: (1) Export the point cloud as an RCS file, attach it in AutoCAD, and trace manually. (2) Use floor plan generation in SHARE PointClouds Studio to create a draft DWG, then refine externally. (3) Use section generation to extract cross-section profiles and trace those. Most teams adopt a combination depending on project complexity.

Why is the point cloud underlay important?

It represents the actual site conditions. Drawn lines can be compared against real wall positions, door and window openings verified against the scan, and structural elements cross-referenced with captured data. Without it, drafting relies on memory and field notes — precisely the limitations digital measurement is designed to address.

After exporting to CAD, is manual work still required?

Yes. The exported DWG is a draft. It must be checked against the point cloud, corrected for auto-detection inaccuracies, annotated, layered, and elevated to the project's drafting standard. The scan provides a faster and more reliable starting point — not an instant finished drawing.

Where Point Cloud to CAD Delivers the Greatest Value

This workflow is not the right choice for every project. It provides the strongest return when:

• Pre-renovation measurement requires documenting existing conditions before design begins.
• The space contains irregular or complex geometry that is tedious or error-prone to measure manually.
• Commercial fit-outs demand accurate as-built records for multi-trade coordination.
• Cross-section references are needed for structural review or MEP coordination.
• Construction progress documentation at milestones requires an archive more useful than photographs alone.
• Existing-conditions capture precedes a renovation where original drawings are missing or unreliable.

For projects involving a single rectangular room with no unusual features, a laser distance meter and a short manual sketch may be the more practical choice. Point cloud to CAD workflows become most valuable when the alternative involves multiple return visits, disputed dimensions, or drawings that do not reflect site reality.