From Point Cloud Data to CAD Floor Plans! A New Workflow Connecting Surveying and Design

Table of Contents

• Challenges of Traditional Surveying and Floor Plan Creation

• What Is Point Cloud Data?

• The Process of Converting Point Cloud Data into CAD Floor Plans

• A New Workflow Enabled by Point Cloud Utilization

• Recommendation: Simple Surveying with LRTK

• FAQ

Challenges of Traditional Surveying and Floor Plan Creation

In architectural and civil engineering design, accurately measuring the dimensions of existing buildings and sites and creating floor plans is an indispensable process. In recent years, the construction industry has loudly called for digital transformation (DX) of on-site operations, yet measurements taken in the field and subsequent drawing creation still rely heavily on manual work, making efficiency a major challenge. Until now, designers and survey technicians needed to measure dimensions on site and produce drawings from those measurements to create floor plans of existing buildings. A typical procedure is as follows.

• On-site survey (measurement): Visit the building or site and measure necessary dimensions one by one—such as wall-to-wall distances, room sizes, heights, and the positions of openings—using tapes or laser measuring devices, and record them. It is not uncommon for this work to take two people several hours.

• Sketches and notes: Write measurement results on paper sketches at the site and leave handwritten notes. When there are many measurement points, omissions are likely, and partial photos are often taken to prevent oversights.

• Drafting (clean copy): Return to the office and, based on the on-site measurement notes and sketches, produce a clean drawing in CAD software. Draw wall lines, place openings such as doors and windows, and add dimension lines to complete an accurate floor plan.

• Checking for missing information: Often, once you start drafting in CAD you realize you forgot to measure some necessary dimensions. If critical dimensions are missing, you must revisit the site to take additional measurements.

With these traditional methods, there is inefficiency caused by going back and forth between fieldwork and office work. Repeated site visits are a significant burden for designers, especially when buildings are complex or located far away. Manual surveying inevitably introduces errors of several centimeters (a few inches) and recording mistakes, and stories like “I took the data back to the office only to find I forgot to measure a crucial dimension” are not uncommon. The time and human-error risks involved in creating floor plans that accurately reflect current conditions have been long-standing problems.

What Is Point Cloud Data?

A new technology attracting attention that can help solve these issues is the use of "point cloud data." Point cloud data digitally records a large number of points in space as a collection of XYZ coordinates. In other words, it is 3D data that contains the positional information of countless points that make up an object or space (and in some cases color information as well). It is acquired using technologies such as laser scanners and photogrammetry, and can precisely represent the shape of buildings and terrain.

Traditionally, acquiring point clouds required expensive specialized 3D laser scanners, but recently some smartphones have begun to include LiDAR (infrared laser distance sensors). This makes it possible to easily capture point cloud data using a smartphone. For example, scanning an interior with a modern smartphone can convert the entire indoor space—including walls, floors, and ceilings—into point cloud data in a short time.

The characteristic advantage of point cloud data is that it can comprehensively capture spatial geometry. The number of points a person can measure with a tape is limited, but with point clouds you can record virtually every location in the space as a dense collection of points. Because even slight wall tilts or protruding beams can be captured, you don’t have to worry later about “having missed something” or “forgetting to measure.” Once scanned, any required dimension can be measured later within the data. Moreover, if photogrammetry is combined to add color information to the point cloud, you can recreate the current condition as a photorealistic 3D model.

Thus, compared to traditional manual surveying, point cloud data provides far more detailed and accurate spatial information. In architecture and civil engineering, its use is expanding widely—from understanding existing conditions for renovations and seismic retrofits to as-built management and cultural heritage recording. Another important benefit is the ability to safely measure areas that were previously difficult for people to access. High places, busy roads, and dangerous slopes can be measured from a distance using laser scanning or drone photography, allowing data acquisition without physically entering hazardous areas.

The Process of Converting Point Cloud Data into CAD Floor Plans

So how do you create 2D floor plans from acquired point cloud data? Here is a general workflow.

• Acquiring point cloud data: For building interiors, walk around holding a smartphone or a handheld laser scanner to capture the point cloud of the entire room. For outdoor terrain surveys, point clouds may be acquired using drone photogrammetry or ground-based laser scanners. The acquired point cloud data consists of millions of points and is imported into dedicated software or cloud services for use.

• Adjusting the point cloud data: Perform processing as needed, such as registering multiple point clouds, removing noise, or performing coordinate transformations. Modern software has become highly capable, and even large point clouds can now be smoothly displayed and edited on a PC. In some cases, unnecessary points are thinned out, or filtering is applied to extract only main structures like walls and floors. However, the latest point-cloud-compatible CAD software increasingly allows handling data without extensive pre-processing.

• Tracing in plan view: Draw the floor plan based on the point cloud. A typical method is to display a horizontal slice of the point cloud and trace wall lines over it. For example, if you display a thin slice of the point cloud at a height of 1 m (3.3 ft), the outlines of walls and columns project onto a 2D plane, and you can use these as guides to draw CAD geometry. Door frames and window positions can also be placed while checking the point cloud. If the point cloud is densely captured, you can produce plans with high dimensional accuracy.

• Finishing and drafting: Clean up the traced line work, organize drawing symbols and dimensions, and complete the CAD drawing. The resulting floor plan data can be saved in DXF or DWG format for editing and sharing in general CAD software. Also, if you preserve the point cloud itself as a deliverable, you can later reuse it to create elevations or sections, or easily share 3D information with designers and contractors.

This point-cloud-to-CAD “digitization” process dramatically improves the accuracy of drawings that were previously created from on-site notes and manual measurements. Because the positions of all components are recorded by scanning, there are no omissions in the drawings. Using point clouds also reduces variability due to the drafter’s skill, allowing consistently high-quality drawings to be produced.

In the past, point cloud processing and drawing creation were sometimes done separately in non-integrated software, but now there are also software solutions that integrate point cloud display and CAD drawing. Because you can display point clouds and create/edit line drawings in a single application, work proceeds efficiently without data conversion overhead. For creating terrain maps with elevation differences, technologies that automatically extract the ground surface from point clouds, fill gaps between points to make edges clearer, and automatically track alignments have been commercialized, making accurate drafting possible with less effort.

A New Workflow Enabled by Point Cloud Utilization

By utilizing point cloud data, the processes of surveying (field measurement) and design (drawing creation) can be connected via a single digital dataset, creating a workflow that is markedly different from the traditional one. The biggest change is that you can cover almost all the information needed for design with a single data acquisition, without having to visit the site multiple times.

For example, for a building with a total floor area of several hundred square meters, a survey that used to require two people and half a day can now be completed by one person in just a few minutes to tens of minutes by scanning with a smartphone. The acquired point cloud data can be taken back to the office for detailed analysis and drafting, enabling a clear division of roles: “quick data acquisition in the field, detailed drafting in a calm office environment.” This dramatically shortens the time lag between fieldwork and office work, reducing unnecessary travel and waiting time. In practice, there have been cases where on-site survey time was reduced to less than one-tenth of the conventional time, and this effect is more pronounced on large projects.

Also, by using point cloud data as a common format, surveyors and designers can share the same source of information. Traditionally, when the person who measured and the person who drew the plans were different, miscommunications or differences in interpretation could occur. With point clouds, designers themselves can collect on-site data, and even if someone else performed the measurements, viewing the point cloud allows accurate 3D understanding of site conditions. With a point cloud that serves as a “digital copy of the site,” team members can collaborate as if they “brought the site back with them” to share.

This new workflow accelerates DX (digital transformation) in architectural design and civil engineering. As seen in initiatives like i-Construction promoted by the Ministry of Land, Infrastructure, Transport and Tourism, digitalization and labor-saving on sites are industry-wide trends. Integrating point cloud measurement and drafting is a concrete measure toward that goal. In practice, point cloud measurement plus CAD workflows are powerful in situations such as:

• Renovation and remodeling: Even old buildings without existing drawings can have detailed dimensions acquired quickly with a point cloud scan. Having accurate as-built drawings improves the precision of design plans and reduces rework during construction.

• Additions and extensions: When adding new structures to existing buildings, point cloud data allows accurate measurement of existing dimensions and tilts. Verifying the interface between old and new components in advance prevents on-site issues such as parts not fitting.

• Equipment upgrades and interior design: Pipes and ducts routed in ceilings and under floors can be confirmed in 3D if point clouds are captured. Considering hidden systems in layouts reduces oversights during office moves or store renovations.

• Current-condition presentations: Placing models of new walls or furniture on the acquired 3D point cloud helps clients intuitively understand changes to the space. Overlaying the existing condition and proposed plan makes presentations more persuasive and facilitates consensus building.

Given the deepening labor shortage, workflows that achieve both efficiency and improved accuracy will become increasingly important.

Recommendation: Simple Surveying with LRTK

Finally, as a modern surveying and drafting solution using smartphones, we introduce LRTK. LRTK is an innovative system that turns a smartphone into a high-precision surveying instrument. By attaching a dedicated ultra-compact RTK receiver to a smartphone, you can obtain position coordinates with an error range of a few centimeters (a few inches). Combining this with smartphone LiDAR scanning and photogrammetry enables anyone to easily acquire high-precision point cloud data.

Survey data acquired with LRTK is automatically saved and synchronized to the cloud from the smartphone app on site. Access the cloud from an office PC and you can immediately view and use the uploaded point cloud data. In the cloud viewer you can measure distances and areas on the captured point cloud, and if needed, trace with a mouse to create floor plans. Completed drawing data can be exported in DXF format for direct editing and use in commercial CAD software. The point cloud itself can also be downloaded in XYZ or LAS formats, making integration with existing BIM/CAD software easy.

There is no complicated operation required—with just a smartphone and LRTK, anyone can complete surveying through drafting—and that ease of use is a major appeal. Users who have adopted LRTK report feedback such as “It’s overwhelmingly easier to use than traditional surveying equipment yet offers high accuracy” and “Carrying one device per person makes on-site responses much faster.” For architects and civil engineers, LRTK can be a reliable partner that seamlessly connects field surveys to drawing creation.

The LRTK series combines high-precision GNSS positioning technology with cloud services to dramatically improve on-site surveying accuracy and operational efficiency. By utilizing such modern tools, implementing a new point-cloud-based workflow should become even easier. For details, please also visit the [LRTK official website](https://lrtk.lefixea.com/). Please consider using LRTK to evolve your sites to the next stage.

FAQ

Q: Do I need special equipment to acquire point cloud data? A: Not necessarily. While high-precision 3D laser scanners used to be mainstream, recently smartphones and drones have made it easy to acquire point clouds. For small indoor measurements, LiDAR-equipped smartphones are often sufficient, and for wide-area terrain surveys, drone photogrammetry is an option. By choosing the appropriate method for your use case, you can start point cloud measurement without special equipment. For large projects, you can also rent high-performance 3D laser scanners temporarily or outsource to specialized point cloud measurement services.

Q: Is the accuracy of point clouds acquired by smartphones sufficient? A: In many cases, the accuracy of smartphone-acquired point clouds is adequate for creating floor plans. Point clouds obtained by smartphone LiDAR typically fall within errors of several centimeters (a few inches) for indoor measurements, depending on the environment. This is usually sufficient for architectural planning and renovation drawings. However, for geodetic control surveys requiring millimeter-level accuracy, a smartphone alone may be insufficient. In such cases, combining RTK-GNSS like LRTK to improve positioning accuracy is effective.

Q: Can floor plans be generated automatically from point cloud data? A: At present, fully automatic floor plan generation remains difficult. Research and development in automating point cloud processing are progressing, and attempts exist to detect planar surfaces like walls and floors from point clouds and convert them into line drawings. However, because building shapes and conditions vary, completely eliminating human judgment is challenging. In practice, operators typically trace point clouds to create drawings efficiently. Point clouds significantly reduce manual labor compared to hand measurements, so the realistic view is “humans supplement the parts that are difficult to automate,” rather than “it’s manual because it’s not automated.”

Q: Do I need a high-performance PC to handle point cloud data? A: Point cloud datasets can be large, containing millions of points, but improvements in PC performance and software optimization mean that a typical business PC can handle a reasonable amount of data. Of course, better CPU, memory, and graphics performance improve usability, but strategies such as splitting or compressing unnecessary parts or using cloud services can mitigate local hardware limits. For example, LRTK manages and displays point clouds in the cloud, allowing smooth viewing regardless of local PC specs. Except for extremely large point clouds, standard PC environments are generally sufficient.

Q: Is the cost of introducing a new point-cloud workflow high? A: While traditional 3D laser scanners could cost millions of yen and posed high barriers to adoption, lower-cost options have increased recently. Using smartphones is a prime example—you can operate with only dedicated apps and auxiliary devices at relatively low cost. Moreover, introducing point cloud workflows can shorten on-site survey time and improve work efficiency, so overall benefits can outweigh costs. For example, reducing multiple site visits to a single visit cuts labor and travel expenses. The ability to improve work quality and productivity with minimal investment is a major advantage.