How to Create Floor Plans from Handheld LiDAR Scans

Handheld LiDAR scanners have revolutionized interior scanning workflows. Unlike traditional tripod-mounted scanners that require stopping at each position, handheld devices let you walk through a building while continuously capturing millions of 3D points. The result? Faster scanning times, easier data collection, and the same high-quality point clouds needed for accurate floor plan creation.

But once you've captured that handheld scan, how do you turn it into a usable floor plan? In this comprehensive guide, we'll walk through everything you need to know about creating floor plans from handheld LiDAR scans—from choosing the right device to exporting your final CAD files.

What is Handheld LiDAR Scanning?

Handheld LiDAR scanning (also called mobile or portable LiDAR) uses a compact, battery-powered scanner that you carry while walking through a space. The scanner continuously captures 3D point cloud data as you move, using simultaneous localization and mapping (SLAM) technology to track its position and orientation in real-time.

Unlike tripod scanners that capture from fixed positions, handheld scanners are designed for continuous motion. You simply walk through the space at a normal pace (typically 1-2 meters per second), and the scanner does the rest.

Key Advantages of Handheld LiDAR

Limitations of Handheld LiDAR

Handheld scanners excel at interior spaces, but have some limitations:

• Interior-focused: Best for indoor spaces; outdoor scanning requires good lighting and stable conditions
• Accuracy: Typically ±2-5cm accuracy (tripod scanners achieve ±1-2cm)
• Range: Effective range of 20-100m (vs 100-300m for tripod scanners)
• Complex geometries: May struggle with very large open spaces or highly reflective surfaces

Handheld vs Tripod Laser Scanners: When to Use Each

Both handheld and tripod scanners have their place. Here's when to choose each:

Decision Guide: Handheld vs Tripod

Popular Handheld LiDAR Devices Overview

Several manufacturers offer handheld LiDAR scanners, each with different strengths. Here's an overview of the most popular options:

1. SatLab SLAM

SatLab's SLAM scanner is a popular choice for interior scanning workflows. It offers good accuracy for handheld scanning and exports in standard formats (E57, LAS, PLY) that work directly with ScanToBIM-CAD and other conversion tools.

2. GeoSLAM ZEB Series

GeoSLAM's ZEB series is known for its lightweight design and fast scanning capabilities. The ZEB Horizon is particularly popular for multi-story buildings and large interior spaces. GeoSLAM's processing software (GeoSLAM Hub) provides good registration and export options.

3. Leica BLK2GO

The BLK2GO is Leica's entry into handheld scanning. It's one of the lightest options available and integrates well with Leica's existing software ecosystem (Cyclone, CloudWorx). For floor plan export, you'll typically export from Cyclone to E57 format, which can then be converted to DXF using ScanToBIM-CAD.

4. NavVis VLX

NavVis VLX is a more premium option with higher accuracy and longer range. It's designed for large commercial and industrial spaces where precision matters. The VLX also captures panoramic images alongside point cloud data, which can be useful for documentation and visualization.

5. Other Handheld Options

• Faro Freestyle: Faro's handheld solution, integrates with Faro software ecosystem
• Trimble X7 (with handheld mode): Can be used in handheld mode for faster scanning
• iPhone/iPad LiDAR: Consumer-grade option for basic scanning (lower accuracy, but very accessible)

Typical Workflow: From Scan to CAD

Creating floor plans from handheld LiDAR scans follows a consistent workflow. Here's the step-by-step process:

Step 1: Field Scanning (15-30 Minutes)

Begin by scanning the space with your handheld device:

1. Prepare the scanner: Charge battery, check storage space, calibrate if needed
2. Plan your route: Walk through the space systematically (e.g., start at entrance, follow perimeter, then interior rooms)
3. Start scanning: Begin recording and walk at normal pace (1-2 m/s)
4. Cover all areas: Ensure you scan all rooms, corridors, and important spaces
5. Close loops: Return to starting point or overlap with previous path to help registration
6. Stop recording: End scan when complete

Step 2: Data Transfer and Initial Processing (10-20 Minutes)

After scanning, transfer data to your computer and run initial processing:

1. Transfer scan data: Connect scanner to computer or use manufacturer's software to download
2. Open in processing software: Use manufacturer's software (GeoSLAM Hub, Leica Cyclone, etc.)
3. Run SLAM processing: Software processes IMU and LiDAR data to create registered point cloud
4. Review quality: Check for gaps, registration errors, or missing areas
5. Clean if needed: Remove noise, outliers, or unwanted objects

Processing time varies by device and software, but typically takes 10-30 minutes for a 500m² scan.

Step 3: Export Point Cloud (5-10 Minutes)

Export your processed point cloud in a format compatible with CAD conversion:

• E57 format: Industry standard, preserves metadata (recommended)
• LAS format: LiDAR standard, good compression with LAZ
• PLY format: Simple format, widely supported

Export settings to consider:

• Point density: For floor plans, 1-5mm spacing is typically sufficient
• Color data: Include RGB if available (helps with AI detection)
• File size: For large buildings, consider splitting by floor

Step 4: Convert to Floor Plan (2-4 Minutes)

Now convert your point cloud to a floor plan using AI-powered tools:

1. Upload to ScanToBIM-CAD: Go to ScanToBIM-CAD and create a free account
2. Upload your file: Drag and drop your E57, LAS, or PLY file (no format conversion needed!)
3. AI processing: Wait 1-3 minutes while AI detects walls, doors, and windows
4. Review results: Check the 3D viewer to verify detection quality
5. Download files: Get both DXF (2D floor plan) and IFC (3D BIM model)

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

Import the downloaded DXF/IFC file into your BIM/CAD software (Revit, Archicad, AutoCAD, LibreCAD, etc.) for final refinement:

• Review AI detection: Check that walls, doors, and windows are correctly identified
• Fix any errors: Correct any AI mistakes (typically 5-10% of elements need minor adjustments)
• Add missing elements: Manually add anything the AI missed
• Add dimensions: Verify and add wall dimensions
• Add room labels: Label rooms and spaces
• Add annotations: Title block, north arrow, scale, etc.
• Final quality check: Compare against original point cloud

This refinement step typically takes 1-3 hours, compared to 8-20 hours of initial manual tracing. The AI does 90-95% of the work automatically.

Total Workflow Time

For a typical 500m² interior scan:

Compare this to traditional manual tracing, which would take 8-20 hours for the same project!

Data Formats and Converter Tools

Handheld scanners export in various formats. Here's what you need to know:

Common Export Formats

Format Conversion Tools

If you need to convert between formats, here are common tools:

• CloudCompare: Free, open-source tool for format conversion (E57 ↔ LAS ↔ PLY)
• Manufacturer Software: GeoSLAM Hub, Leica Cyclone, etc. can export to multiple formats
• PDAL: Command-line tool for advanced conversions
• ScanToBIM-CAD: Accepts all formats directly—no conversion needed!

Accuracy and Best Practices

Getting accurate floor plans from handheld scans requires following best practices:

1. Scanning Best Practices

• Walk at consistent speed: 1-2 meters per second is optimal
• Keep scanner level: Avoid tilting or rotating rapidly
• Close loops: Return to starting point or overlap paths to improve registration
• Scan important features: Walk past doors and windows from multiple angles
• Avoid rapid movements: Sudden stops or direction changes can cause registration errors
• Ensure good coverage: Don't skip rooms or areas

2. Processing Best Practices

• Use manufacturer's software: Each scanner has optimized processing software
• Check registration quality: Look for gaps, misalignments, or duplicate geometry
• Clean point cloud: Remove noise, outliers, and unwanted objects before export
• Verify accuracy: Compare against known measurements if possible
• Export appropriate density: 1-5mm spacing is typically sufficient for floor plans

3. Conversion Best Practices

• Use AI-powered tools: Automated conversion is faster and more consistent than manual tracing
• Review AI results: Always check the generated floor plan against the point cloud
• Refine in CAD: Make final adjustments in your familiar CAD environment
• Verify dimensions: Check key measurements against the point cloud
• Add annotations: Room labels, dimensions, title blocks, etc.

Expected Accuracy

For handheld LiDAR scans converted to floor plans:

This accuracy is sufficient for most as-built documentation, facility management, and renovation design work. For survey-grade precision (±1cm), tripod scanners may be required.

Common Pitfalls and Troubleshooting

Here are common issues you might encounter and how to solve them:

Problem 1: Poor Registration (Gaps or Misalignments)

Symptoms: Point cloud has gaps, duplicate walls, or misaligned sections

Solutions:

• Re-process scan with different SLAM parameters
• Ensure you closed loops during scanning (returned to starting point)
• Check for rapid movements or stops that might confuse SLAM algorithm
• Use manufacturer's software optimization tools
• For severe issues, re-scan the problematic area

Problem 2: Missing Walls or Rooms

Symptoms: AI-generated floor plan is missing some walls or entire rooms

Solutions:

• Check point cloud density—ensure walls have adequate point coverage
• Verify scan coverage—make sure you scanned all areas
• Manually add missing elements in CAD software (still faster than full manual tracing)
• Re-process with different AI parameters if available

Problem 3: Incorrect Door/Window Detection

Symptoms: Doors or windows are missing, incorrectly positioned, or wrong size

Solutions:

• Check point cloud quality around openings—ensure good coverage
• Manually adjust door/window positions in CAD (quick fix)
• Verify scan captured openings from multiple angles
• Add missing doors/windows manually in CAD

Problem 4: Low Point Cloud Density

Symptoms: Sparse point cloud, difficult to identify features

Solutions:

• Walk slower during scanning (allows more points per area)
• Scan important areas multiple times
• Check scanner settings for point density/scan rate
• Use higher-quality scanning mode if available

Problem 5: Large File Sizes

Symptoms: Point cloud files are very large (several GB), slow to process

Solutions:

• Export with lower point density (1-5mm spacing is sufficient for floor plans)
• Split large buildings by floor into separate files
• Use compressed formats (LAZ instead of LAS, binary PLY instead of ASCII)
• Remove unnecessary data (exterior points if doing interior-only floor plan)

Problem 6: Accuracy Issues

Symptoms: Floor plan dimensions don't match known measurements

Solutions:

• Verify scanner calibration
• Check for registration errors in point cloud
• Compare against known measurements (door widths, room dimensions)
• Re-scan if accuracy is critical and errors are significant
• Consider using tripod scanner for higher-precision requirements

Real-World Example: Office Space Floor Plan

Let's walk through a real-world example:

Workflow Timeline

• Field scanning: 20 minutes (walked through entire space systematically)
• Data processing: 15 minutes (GeoSLAM Hub SLAM processing)
• Export to E57: 5 minutes
• AI conversion (ScanToBIM-CAD): 3 minutes
• CAD refinement: 2 hours (reviewed, added room labels, dimensions, title block)
• Total time: 2 hours 43 minutes

Results

The AI-generated floor plan was 95% complete—only minor adjustments were needed:

• ✅ All walls correctly detected and positioned
• ✅ 11 of 12 doors correctly identified (1 door needed manual adjustment)
• ✅ 7 of 8 windows correctly identified (1 window needed manual addition)
• ✅ Room boundaries accurate to within ±3cm
• ✅ Final floor plan delivered same day (vs 1-2 weeks with manual tracing)

Conclusion: Convert Your Handheld LiDAR Scan to Floor Plans in Minutes

Handheld LiDAR scanners have made interior scanning faster and more accessible than ever. Combined with AI-powered conversion tools like ScanToBIM-CAD, you can now create accurate floor plans in hours instead of weeks.

Key takeaways:

• ✅ Handheld scanners: Fast, portable, ideal for interior spaces
• ✅ Popular devices: SatLab, GeoSLAM, BLK2GO, NavVis all export standard formats
• ✅ Complete workflow: Scan → Process → Export → Convert → Refine (2-4 hours total)
• ✅ Format support: E57, LAS, PLY all work with ScanToBIM-CAD (no ReCap needed!)
• ✅ Time savings: 95% faster than manual tracing (2-4 hours vs 8-20 hours)
• ✅ Cost savings: 70-90% reduction compared to manual workflows

Ready to convert your handheld LiDAR scan to a floor plan? Try ScanToBIM-CAD free—upload your E57, LAS, or PLY file and get your floor plan in minutes. No credit card required, 2 free conversions to start.

Have questions about handheld scanning workflows? Check out our other guides on point cloud to floor plan conversion, point cloud formats, and as-built documentation workflows.