Improve drawing accuracy with point cloud data! Method for creating floor plans that faithfully reflect existing conditions

Table of contents

• What is point cloud data?

• Benefits of using point clouds for floor plan creation

• Steps to create high-precision floor plans from point clouds

• The importance of drawings that faithfully reflect existing conditions

• Challenges and countermeasures for using point cloud technology

• Recommendation for simple surveying using LRTK

• FAQ

First, we will outline the creation of floor plans using the increasingly prominent “point cloud data.” Point cloud data are three-dimensional measurement data that represent objects such as buildings and terrain with countless points. Their greatest characteristic is that they can record site shapes and dimensions with high density and high accuracy.

Conventional floor plan creation generally involved manual surveying and hand-drawn drafting. However, by using point cloud surveying, you can measure wide areas in a short time and efficiently create accurate floor plans from the vast information obtained. The use of point cloud data is spreading as a new method to obtain drawings that faithfully reflect existing conditions.

This article explains point cloud data from the basics to the benefits of use, the drawing creation process, and key points for implementation. Please use it as a reference for tips on creating high-precision drawings that faithfully reflect existing conditions.

What is point cloud data?

Point cloud data are collections of countless points acquired by devices such as laser scanners or photogrammetry, with each point containing X, Y, Z coordinate information (and sometimes color information). For example, if you laser-scan the interior of a building, you obtain countless points that make up surfaces in the space, from walls, floors, and ceilings to fixtures. At first glance a point cloud may look like a coarse photograph, but in reality it is digital measurement data in which each point has exact position coordinates.

When displayed on dedicated software, the assembly of points can reproduce the site in three dimensions. In this way, point cloud data are attracting attention as a powerful means to digitally archive physical space as it is.

The first advantages of point cloud data are its speed and large amount of information. Tasks that previously required people to measure point by point with tapes or surveying instruments can be recorded en masse over wide areas in a short time with point cloud surveying. The number of coordinate points that can be obtained can reach millions, so it is unlikely that you will later realize “I forgot to measure that part,” and you can freely measure required dimensions on the software. It is truly like “bringing the whole site back as data,” and a major feature is the ability to record existing conditions in detail.

Point cloud data are also attractive for their high accuracy. Using high-quality laser scanners or photogrammetry, you can capture object shapes with precision to the millimeter (mm level, 0.04 in). Even complex-shaped structures can be comprehensively recorded, including undulations and fine details, making it easy to later produce high-precision drawings or 3D models. For these reasons, point cloud data are being used across a wide range of fields, from civil engineering and construction to architecture and plant management.

Benefits of using point clouds for floor plan creation

Utilizing point cloud data to create floor plans offers many benefits not found in conventional methods. The main advantages are summarized below.

• Fast and efficient measurement: Point cloud surveying using laser scanners or drones can measure wide areas at once, greatly reducing the time spent on-site. On-site measurements that used to take multiple people several days can sometimes be completed in a short time.

• Comprehensive capture of existing conditions: Point clouds record the site down to the smallest corners with a large number of points, preventing the “missed measurements” that often occur with manual surveying. Details and complex areas that are easy to overlook can be checked in the data, reducing the likelihood of later finding “I should have measured that area too.”

• High-precision drawing creation: Using the obtained point cloud data, you can create floor plans more accurately than before. Since point clouds contain precise three-dimensional coordinates of objects, you can import them into CAD software, create arbitrary section views, and accurately draft floor plans and elevations based on those views. Because drawings can be produced to millimeter-level precision (mm level, 0.04 in) that was difficult with manual surveying, dimensional errors and drafting mistakes can be greatly reduced.

• Derivation into multiple drawings and models: From point cloud data you can derive not only floor plans but also elevations, sections, and various drawings or 3D models. With a single point cloud acquisition you obtain multifaceted materials, avoiding redundant work and improving efficiency. For example, if you scan a building, you can later generate elevations or detail drawings from the point cloud without re-surveying.

• Improved quality and reduced rework: Accurate point cloud–based drawings increase precision in the design and construction stages and prevent rework due to mistakes. Having reliable drawings based on actual measurements reduces discrepancies with the site, contributing to construction quality assurance and more efficient schedule management.

As shown above, creating floor plans using point clouds brings many advantages in terms of speed, accuracy, and efficiency. Especially for renovations of existing buildings or creating drawings for old facilities, point cloud–based drafting is a more reliable and reassuring choice than conventional methods.

Steps to create high-precision floor plans from point clouds

Now let’s look at how to create floor plans from point cloud data. Below is a representative example of the general workflow.

• On-site point cloud measurement: First, scan the target site or building with a 3D laser scanner, drone, or LiDAR-equipped smartphone to acquire point cloud data. Choose the optimal measurement method according to the survey range, and if necessary perform scans from multiple locations to collect data.

• Point cloud data processing and integration: Raw point clouds may contain unwanted points or noise. Use dedicated software to remove noise and interpolate missing parts. If multiple point clouds were measured, align and integrate them into a single coordinate system. This process results in clean point cloud data suitable for floor plan creation.

• Extraction of data for floor plans: To create floor plans from point clouds, extract the required sections or projection planes. For an architectural floor plan, for example, slice a horizontal cross-section at a certain height above the floor and extract the outlines of walls and columns that appear in that section. For a top-down plan (plan view) of terrain, generate a projection view from directly above. Using CAD software or point cloud processing tools that support point clouds, you can freely create such cross-sections and projections.

• CAD drafting (tracing): Using the extracted sections or projections as a base, trace the linework in CAD to draw the floor plan. Some software can automatically detect walls and columns and extract lines, but correct manually as needed to produce an accurate drawing. By drafting while referencing the point cloud, you can create drawings that faithfully reflect existing conditions while preventing dimensional mistakes or annotation errors.

• Finishing and verification: Finally, add dimension lines and notes to the completed floor plan to finish it as a drawing. Then perform a final check to ensure the finished drawing is consistent with the point cloud data. Verify by measuring important dimensions on the point cloud and comparing them to the drawing values to check for omissions or errors. Conducting such checks results in a floor plan that is faithful to the point cloud data and highly accurate.

The above is the basic workflow. In actual projects, the detailed steps vary depending on the object and the software used, but the flow is generally “point cloud measurement → data processing → section extraction → tracing → drawing finishing.” Compared to traditional manual measuring and hand drawing, the major difference is that digital measurement and CAD use allow greater efficiency and precision.

The importance of drawings that faithfully reflect existing conditions

Accurate floor plans are essential documents that form the basis for all decisions in design and construction. If drawings diverge from existing conditions, it can lead to construction errors on-site, mistakes in material procurement, and even safety risks. Therefore, preparing drawings that faithfully reflect existing conditions is indispensable for project quality control.

However, with existing buildings and infrastructure, it is not uncommon to have only old drawings and insufficient understanding of the current state. Also, even if design changes were made during construction, they may not be fully reflected in the as-built drawings. In such cases, acquiring point cloud data gives you a digital original that records current conditions as they are. For example, if you scan a structure after completion, you can later produce accurate drawings or 3D models from the point cloud even if paper drawings are lost. As-built drawings created from point cloud data have high consistency in shape and dimensions and serve as reliable base materials.

Point cloud data also have the advantage that they can be updated to the latest state at any time. Once a drawing is created, it must be revised with each change, but if you periodically scan the site with point clouds, you can create drawings or models that reflect the latest conditions each time. Point clouds are effective for progress management and as-built records, allowing all change histories during the project to be kept digitally. In this way, point cloud data are a very reassuring ally for obtaining drawings that faithfully reflect existing conditions.

Challenges and countermeasures for using point cloud technology

Although point cloud technology has many advantages, there are some challenges when introducing and using it. Here we discuss representative challenges and countermeasures, as well as recent trends.

First is the issue of equipment cost. High-precision laser scanners and surveying drones used to be very expensive and a barrier to initial investment. Large-scale point cloud surveys also required specialized surveying teams, increasing labor costs. One solution to this has been the recent emergence of compact and affordable measurement devices. For example, palm-sized laser scanners, LiDARs for consumer drones, and smartphone-mounted surveying devices such as the ones described later have become available, lowering costs and making equipment easier to handle. Using such devices, a single person can more easily perform point cloud surveys.

Next is the challenge of data processing and handling. Point cloud data can become extremely large, which can burden a computer’s performance, and specialized software operation requires a learning curve. This issue has also been mitigated recently by the use of cloud services and faster software, making data processing smoother than before. Additionally, strategies such as generating point clouds only for necessary parts or adjusting point density to lighten the data can significantly reduce file sizes. User-friendly point cloud viewers and automatic analysis tools have also appeared, expanding scenarios where point clouds can be used even by those without specialized knowledge.

Finally, there is the human resources and skills challenge. Because point cloud utilization is a relatively new technology, some sites may lack sufficient know-how or personnel. However, this is being overcome through improved training and technician education as well as the proliferation of easy-to-use tools. Especially with the recent advance of site DX, younger engineers are increasingly comfortable with digital measurement, and there is a trend toward active adoption. Point cloud surveying, which at first was outsourced to specialists, is expected to be brought in-house by more companies over time.

Thus, challenges such as cost, data volume, and skills remain, but solutions are emerging with technological advances. In particular, the arrival of “affordable, easy-to-use” point cloud measurement solutions is a key factor that will accelerate future adoption.

Recommendation for simple surveying using LRTK

Given the challenges described above, smartphone-based easy point cloud surveying tools have attracted attention in recent years. A representative example is “LRTK.” LRTK is a compact positioning device that attaches to a smartphone and transforms the phone into a high-precision surveying instrument.

LRTK incorporates a compact yet high-precision GNSS antenna and an RTK-GNSS receiver, enabling centimeter-level position measurements (cm level accuracy, half-inch accuracy) when combined with a smartphone. At the same time, it works with the smartphone’s built-in LiDAR sensor or camera to allow anyone to easily acquire position-tagged 3D point cloud data. With smartphone-only LiDAR scans, the obtained point clouds lacked georeferencing and had unclear scale or positional relationships. However, by using LRTK, the accurate coordinate reference provided by RTK prevents distortions and positional shifts in the point cloud even when scanning while walking. You can thus easily acquire true-to-scale 3D data of the site on the spot.

Furthermore, LRTK is lightweight and compact at about 165 g, easily attaching to a smartphone with a one-touch dedicated case for portability. Its price is significantly lower than that of large laser scanners, making one-per-person deployment realistic. It also supports cloud integration, making it easy to immediately share measured data for team use. Designed to be user-friendly even for non-specialist surveyors, it is an attractive option for companies wanting to start in-house point cloud measurement or for engineers who need simple surveying on site.

The acquisition of point cloud data, once a high hurdle, has become markedly more accessible thanks to tools like LRTK. If you feel “I want to create drawings that faithfully reflect existing conditions but lack specialized equipment or personnel…,” consider trying such cutting-edge solutions. With easy-to-start point cloud surveying, you can take the first step toward improving drawing accuracy.

FAQ

Q: What equipment and software are required to use point cloud data? A: To acquire point cloud data you can use 3D laser scanners, drones, or increasingly LiDAR-equipped smartphones. For high-precision measurements, laser scanners or RTK-capable drones are effective, but for small sites a smartphone combined with a dedicated device (e.g., LRTK) may suffice. For post-processing, you need point cloud processing software or CAD software; options range from free software to advanced commercial products depending on your needs.

Q: Can inexperienced users create accurate drawings from point cloud data? A: Although you may be bewildered at first, inexperienced users can produce drawings once they grasp the basic steps. Since the point cloud contains all on-site dimensions, you only need to measure and trace required parts in the software. Automatic point cloud processing features and tutorials have become more comprehensive recently, and with training you can learn relatively quickly. A phased approach—outsourcing difficult tasks to specialists while doing finishing work in-house—is also an option.

Q: How much time and cost does point cloud surveying take? A: It depends on the survey range and method, but point cloud surveying generally covers wide areas in a shorter time than traditional surveying and can reduce labor costs. Equipment costs for conventional laser scanners can be high, but renting or adopting small inexpensive devices (e.g., smartphone + LRTK) can greatly reduce expenses. Specific time and cost depend on conditions, but choosing the appropriate method often improves overall efficiency and cost-effectiveness.

Q: Aren’t point cloud data files large and difficult to handle? A: Point cloud data do tend to become large. However, by scanning only necessary ranges or adjusting point density to reduce data volume, you can manage file sizes. Modern computers have improved performance, and with appropriate software you can smoothly display and process point clouds of tens of millions of points. You can also thin out unnecessary parts or use cloud services for processing, so there is no need to worry excessively about data size.

Q: Is point cloud surveying meaningful if I ultimately only need 2D drawings? A: Yes. Even if your goal is only 2D drawings, point cloud surveying has great benefits. Once you acquire point clouds, you can later generate elevations or sections without re-surveying if needed. Creating drawings from point clouds yields high-accuracy, reliable drawings that match existing conditions. It is safer to record the whole site with point clouds than to take partial measurements targeting only the initial 2D plan. If additional drawings are needed later, the point cloud data allow smooth対応, and above all, it is an effective means for improving drawing accuracy.