As-Built Floor Plans from Laser Scans: Complete Workflow

You've just completed a laser scan of an existing building. The point cloud data is perfect—millions of precise measurements showing every wall, door, window, and structural element exactly as it exists. Now you need to turn this raw scan data into an as-built floor plan—a professional CAD drawing that accurately documents the current state of the building.

Whether you're working on a renovation project, creating facility management documentation, or preparing for a design-build project, as-built floor plans are essential. They serve as the foundation for all subsequent design work, ensuring that architects and engineers are working with accurate existing conditions.

In this comprehensive guide, we'll walk through the complete workflow for creating as-built floor plans from laser scans. We'll compare the traditional manual drafting approach with modern AI-assisted methods, show you real cost and time savings, and provide a step-by-step process you can follow for your next project.

What is an As-Built Floor Plan?

An as-built floor plan is a technical drawing that documents the existing conditions of a building exactly as it stands today. Unlike design drawings (which show what will be built) or construction drawings (which show what should be built), as-built drawings show what actually exists.

Why As-Built Floor Plans Matter

As-built documentation is critical for several reasons:

The Challenge: Buildings Rarely Match Original Drawings

Here's the problem: existing buildings rarely match their original construction drawings. Over time, walls get moved, doors get relocated, windows get replaced, and rooms get reconfigured. Original drawings from 20, 30, or 50 years ago are often inaccurate or missing entirely.

This is why laser scanning has become the gold standard for as-built documentation. A laser scan captures the building exactly as it exists today, with millimeter-level accuracy. But the scan itself is just raw data—you still need to convert it into usable CAD drawings.

The Traditional As-Built Workflow: Field → Scan → Manual Drafting

For decades, creating as-built floor plans has followed the same labor-intensive process. Let's break down each step and understand why it's so time-consuming and expensive.

Step 1: Field Scanning (1-2 Days)

The process begins in the field with laser scanning. A surveyor or technician:

1. Sets up the scanner at multiple positions throughout the building
2. Captures scans from each position (typically 10-50 scans per floor, depending on building size)
3. Registers scans using targets or cloud-to-cloud registration
4. Exports point cloud in E57, LAS, or PLY format

For a typical 500m² office space, this might involve:

• 20-30 scan positions
• 4-6 hours of scanning time
• 1-2 hours of registration and quality checking
• Total field time: 1 full day

Step 2: Point Cloud Processing (2-4 Hours)

Back in the office, the raw scan data needs processing:

1. Import scans into registration software (Leica Cyclone, Faro Scene, Trimble RealWorks)
2. Register and align all scan positions
3. Clean point cloud (remove noise, unwanted objects)
4. Export unified point cloud for CAD import

This step typically takes 2-4 hours for a medium-sized project, depending on scan quality and registration complexity.

Step 3: Format Conversion (30-60 Minutes)

Most CAD software (AutoCAD, Revit) cannot directly import point cloud formats like E57 or LAS. You need to convert through Autodesk ReCap Pro:

1. Import E57/LAS/PLY into ReCap Pro
2. Convert to RCP (ReCap Project) format
3. Export to RCS (ReCap Scan) for Revit, or use ReCap directly in AutoCAD

Step 4: Manual Drafting in CAD (1-2 Weeks)

This is where the real time investment happens. A CAD technician manually traces the point cloud:

1. Import point cloud into AutoCAD or Revit
2. Set up drawing layers (walls, doors, windows, dimensions, annotations)
3. Manually trace walls by following the point cloud
4. Identify and draw doors (measuring openings, swing directions)
5. Identify and draw windows (measuring openings, sill heights)
6. Add dimensions to all walls and openings
7. Add room labels and annotations
8. Clean up geometry (fix gaps, align walls, correct angles)
9. Quality check against point cloud
10. Final review and export to DXF/DWG

For a typical 500m² office space, manual drafting takes:

For larger commercial spaces (1,000m²+), this can easily stretch to 1-2 weeks of full-time work.

Step 5: Quality Control and Revisions (4-8 Hours)

After initial drafting, the drawings go through quality control:

• Review against point cloud for accuracy
• Check dimensions and measurements
• Verify door/window locations and sizes
• Client review and revision requests
• Final corrections and updates

This typically adds another 4-8 hours to the project timeline.

Total Time and Cost: Traditional Workflow

For a typical 500m² office space as-built project:

This doesn't include software licensing costs (ReCap Pro at $2,000+/year, AutoCAD/Revit at $2,000+/year). For a firm doing multiple as-built projects per year, software costs add another $4,000-$6,000 annually.

The Modern Workflow: Field → Scan → AI-Assisted Drafting

What if you could skip 95% of the manual drafting time? That's where AI-powered point cloud conversion comes in.

The modern workflow uses the same field scanning process, but replaces manual CAD tracing with automated AI detection. Here's how it works:

Step 1: Field Scanning (Same as Traditional)

The field scanning process remains the same—you still need to capture accurate point cloud data. No changes here.

Step 2: Point Cloud Processing (Same as Traditional)

Registration and cleaning still happen in your scanning software. No changes here either.

Step 3: AI-Powered Conversion (2-4 Minutes)

Instead of manual tracing, you upload your point cloud to an AI-powered conversion tool like ScanToBIM-CAD:

1. Upload point cloud directly (E57, LAS, PLY, or LAZ—no ReCap conversion needed!)
2. AI automatically detects walls, doors, windows, and furniture
3. AI generates clean geometry with proper connections and dimensions
4. Processing completes in 1-3 minutes (depending on file size)

Step 4: Download Your Files

Once AI processing completes, download both file formats:

• DXF file - 2D floor plan ready for AutoCAD or any CAD software
• IFC file - 3D BIM model ready for Revit or ArchiCAD

Step 5: Refine in Your CAD Software (1-3 Hours)

Import the downloaded DXF file into your CAD software (AutoCAD, LibreCAD, etc.) to review and refine the AI-generated geometry:

• Review detected walls, doors, and windows
• Correct any AI detection errors (typically 5-10% of elements need minor adjustments)
• Add missing elements that AI might have missed
• Verify dimensions and measurements against the point cloud
• Add room labels and annotations
• Add title blocks and project information
• Final quality check

This refinement step typically takes 1-3 hours, compared to 20-31 hours of manual drafting. The AI does 90-95% of the work automatically, and you focus on quality control and final touches in your familiar CAD environment.

Total Time and Cost: Modern AI Workflow

For the same 500m² office space project:

*ScanToBIM-CAD pricing: $29/month for 10 conversions. First 2 conversions are free.

**Drafting time only. Field scanning time is the same for both workflows.

Cost Comparison: Traditional vs Modern Workflow

Let's compare the total costs for a typical 500m² as-built project:

*No ReCap needed—ScanToBIM-CAD accepts E57/LAS/PLY directly

**ScanToBIM-CAD subscription: $29-$87/month. Traditional workflow requires ReCap ($2,000+/year) + AutoCAD/Revit ($2,000+/year each)

Step-by-Step: Creating As-Built Floor Plans with ScanToBIM-CAD

Here's exactly how to create as-built floor plans using the modern AI-assisted workflow:

Step 1: Prepare Your Point Cloud

After completing your field scan and registration, export a unified point cloud file:

• Format: E57, LAS, PLY, or LAZ (all supported)
• Size: Up to 2GB per file (larger projects can be split into multiple files)
• Quality: Ensure good point density (typically 1-5mm spacing for interior scans)

Step 2: Upload to ScanToBIM-CAD

Go to ScanToBIM-CAD and create a free account (no credit card required). You get 2 free conversions to start.

1. Click "Upload Point Cloud"
2. Drag and drop your file, or click to browse
3. Wait for upload to complete (typically 1-2 minutes for files under 500MB)

No ReCap conversion needed! ScanToBIM-CAD accepts E57, LAS, and PLY files directly.

Step 3: AI Processing (1-3 Minutes)

Once uploaded, the AI automatically:

• Analyzes the point cloud structure
• Detects walls, doors, and windows
• Generates clean BIM/CAD geometry
• Creates proper wall connections and openings

You'll see real-time status updates. Small files (under 100MB) typically complete in under 90 seconds. Larger files (500MB+) may take 2-3 minutes.

Step 4: Download Your Files

Once processing completes, download both file formats:

• DXF (2D floor plan) - Ready for AutoCAD, LibreCAD, or any CAD software
• IFC (3D BIM model) - Ready for Revit, ArchiCAD, or Blender

Both files are generated automatically. The DXF gives you a clean 2D plan view perfect for as-built documentation, while the IFC provides a full 3D BIM model with proper architectural relationships.

Step 5: Refine in Your CAD Software

Import the downloaded DXF file into your CAD software (AutoCAD, LibreCAD, etc.) to review and refine the AI-generated geometry:

1. Review detected elements: Check that walls, doors, and windows are correctly identified and positioned
2. Correct any errors: Fix any AI detection mistakes (typically 5-10% of elements may need minor adjustments)
3. Add missing elements: Manually add any elements the AI might have missed
4. Verify dimensions: Check measurements against the original point cloud for accuracy
5. Add room labels (Office 101, Conference Room, etc.)
6. Add dimensions (if not already included, or to verify accuracy)
7. Add title block with project information
8. Add door/window schedules if required
9. Final quality check against point cloud

This refinement step typically takes 1-3 hours, compared to 20-31 hours of manual drafting. The AI does 90-95% of the work automatically, and you just need to make final adjustments and add annotations.

Case Study: Office Space As-Built Documentation

To illustrate the real-world impact, let's look at a typical project scenario:

Traditional Workflow Timeline

Using the traditional manual drafting approach:

• Field scanning: 1 day (8 hours)
• Point cloud processing: 4 hours
• ReCap conversion: 30 minutes
• Manual drafting: 8 hours (1 day)
• Quality control: 2 hours
• Total time: 22.5 hours (2.8 days)
• Total cost: $900 (at $40/hour)

Modern AI Workflow Timeline

Using ScanToBIM-CAD AI-assisted workflow:

• Field scanning: 1 day (8 hours) - same as traditional
• Point cloud processing: 4 hours - same as traditional
• AI conversion: 3 minutes
• Download and refinement: 2 hours
• Finalization: 30 minutes
• Total time: 14.55 hours (1.8 days)
• Total cost: $582 (at $40/hour) + $29 (ScanToBIM-CAD) = $611

Results Comparison

The AI-generated floor plan was accurate to within 2cm of manual measurements, meeting all project requirements. The client received the as-built drawings 1 day earlier than with the traditional workflow, allowing the design team to start work sooner.

Best Practices for As-Built Floor Plans

Whether you use traditional or modern workflows, here are best practices for creating accurate as-built documentation:

1. Ensure Good Scan Quality

High-quality scans produce better results:

• Point density: Aim for 1-5mm spacing for interior scans
• Coverage: Ensure all areas are captured (no blind spots)
• Registration: Use targets or cloud-to-cloud registration for accuracy
• Color data: RGB color helps AI detection (if available)

2. Clean Your Point Cloud

Before conversion, remove unwanted elements:

• People and moving objects
• Furniture (if not needed in as-built)
• Noise and outliers
• Exterior elements (if doing interior-only as-built)

3. Verify AI Results

Always review AI-generated geometry:

• Check wall positions against point cloud
• Verify door and window locations
• Confirm dimensions are accurate
• Add any missing elements manually

4. Include Standard As-Built Information

Your final drawings should include:

• Room labels and numbers
• Wall dimensions
• Door and window schedules
• North arrow and scale
• Title block with project information
• Revision history (if applicable)

5. Document Limitations

Be transparent about what the as-built drawings show:

• Date of scan
• Areas not scanned (if any)
• Accuracy tolerance (typically ±2-5cm for laser scans)
• Elements not included (furniture, temporary structures, etc.)

When to Use Traditional vs Modern Workflows

Both workflows have their place. Here's when to use each:

For 90% of as-built projects, the modern AI workflow is the better choice—it's faster, cheaper, and produces results that meet or exceed traditional quality standards.

Conclusion: Cut Your As-Built Drafting Time by 50%

Creating as-built floor plans from laser scans doesn't have to take weeks or cost thousands of dollars. Modern AI-powered tools like ScanToBIM-CAD can reduce your drafting time by 95% and cut project costs by 50-60%.

The traditional manual drafting workflow is still viable for specialized projects, but for most as-built documentation work, the modern AI-assisted approach is faster, cheaper, and more efficient.

Key takeaways:

• ✅ Traditional workflow: 1-2 weeks, $1,960-$4,260 per project
• ✅ Modern AI workflow: 2-4 hours drafting time, $1,069-$2,887 per project
• ✅ Time savings: 95% reduction in drafting time (18-27 hours saved)
• ✅ Cost savings: 32-45% reduction per project ($891-$1,373 saved)
• ✅ Software savings: No ReCap license needed ($2,000+/year saved)

Ready to create as-built floor plans faster? Try ScanToBIM-CAD free—upload your point cloud and get your as-built floor plan in minutes. No credit card required, 2 free conversions to start.

Have questions about as-built workflows or point cloud conversion? Check out our guides on how to convert point clouds to floor plans and point cloud format comparisons for more detailed information.