In construction management, post-completion as-built verification is essential for quality assurance, but conventionally it has required substantial effort. Recently, however, a new method that visualizes the differences between 3D design data and point cloud data has emerged, drawing attention for making as-built verification markedly easier. This article explains in detail what construction management issues can be solved by this “3D design data × point cloud difference visualization,” and outlines its concrete benefits and ways to use it. With the power of digital technology, construction management sites are evolving to become smarter and more reliable.

Table of contents

• Challenges in construction management: Why is as-built verification difficult?

• Utilizing 3D design data (CIM models, BIM, and DWG drawings)

• Digitally recording as-built conditions with point cloud data

• What is 3D model × point cloud difference visualization?

• Construction management issues solved by difference visualization

• Closing: Start construction management DX with simple surveying using LRTK

• FAQ (Frequently Asked Questions)

Challenges in construction management: Why is as-built verification difficult?

On construction sites, verifying and recording whether completed structures or developed land were built “according to the design”—known as as-built verification (as-built management)—is indispensable. In public works, it is necessary to measure actual as-built conditions and show differences from the design values in accordance with client-specified as-built management standards. However, traditional methods of as-built verification have many challenges.

Traditionally, measurements were taken manually using tape measures, staffs, and levels, and dimensions and elevations at various locations were compared against the design on drawings. For example, in road construction, the width, thickness, and elevation of the subgrade and pavement are measured at numerous locations after completion, and records and drawing checks are done manually. Because this method measures only certain points, it is difficult to comprehensively grasp the as-built condition of an entire site.

Manual measurement requires time and manpower, and the number of measurement points is limited. As a result, oversights like “passing at measured points but failing across the surface” tend to occur. Subtle differences from the design in unmeasured areas can be missed, leading to the risk of being surprised by “this differs from the drawings” during later inspections. Also, for buried elements that become invisible after backfilling, adequate records may not be left before covering, making it difficult to locate them later. The larger the structure, the more limited manual measurement becomes; it cannot capture the whole quality, and the process tends to be person-dependent and inefficient.

As described above, traditional as-built verification suffers from lack of coverage due to the limited points measured and from manual-work burdens and human errors. To solve these issues, a shift to more efficient and precise digital measurement was needed.

Utilizing 3D design data (CIM models, BIM, and DWG drawings)

Today, the construction industry is moving toward using three-dimensional digital models from the design stage. In civil engineering, CIM models (Construction Information Modeling) and in building construction BIM (Building Information Modeling) are becoming widespread as 3D design data, and the delivery of these as design documents is increasing. Unlike conventional 2D drawings (for example CAD DWG drawings), 3D design data contain the shape information of the finished form in three dimensions, allowing them to be used directly as digital reference surfaces during construction and as-built inspections.

With a 3D design model, construction managers can perform digital verification. For example, what used to be checked with a red pen on paper drawings can now be reproduced and compared on a digital model. Because the design surface itself is reproduced in the computer, it is also easy to automatically detect differences by comparing on-site measurement data against the model.

The Ministry of Land, Infrastructure, Transport and Tourism’s push for i-Construction advocates the active introduction of construction planning and management using 3D models, and the use of CIM/BIM is expected to become increasingly standardized. 3D design data are not merely substitutes for drawings but are key information assets for digitizing construction management. It would be wasteful not to leverage them for as-built verification.

Digitally recording as-built conditions with point cloud data

On the other hand, point cloud data are attracting attention as a means of recording on-site as-built conditions in detail. Point cloud data are collections of countless points obtained by laser measurement or photogrammetry, where each point has XYZ coordinates and thus digitally captures the site’s shape. In other words, it is a 3D scan of the entire site and a digital record that includes minute bumps and hollows that drawings or photos cannot fully capture.

There are various methods for point cloud measurement. For high accuracy, terrestrial laser scanners (TLS) are representative and can measure structural details with millimeter-level precision (mm-level accuracy (0.04 in)). For efficiently recording wide areas, drone photogrammetry is effective; aerial photos can be used to create a point cloud model of the entire terrain. Recently, mobile LiDAR built into smartphones and tablets makes it possible to easily acquire close-range point clouds. Depending on the site scale and purpose, the appropriate measurement method can be chosen.

Using the acquired point cloud data greatly improves as-built management efficiency. First, accuracy and coverage are greatly increased. Point cloud measurement can capture numerous points throughout the site, so tiny surface irregularities or dimensional differences that manual methods would miss are not overlooked. Where conventional methods could only check key locations as “points,” point clouds capture the shape of structures and ground as “surfaces,” enabling comprehensive, three-dimensional understanding of the as-built condition. Elements hidden after construction, such as embedded items or the interior of concrete, can be scanned immediately after construction and preserved as digital data.

There are also major benefits in reduced work time and increased efficiency. A single scan can acquire a large amount of measurement data, so tasks that used to take half a day can sometimes be completed in tens of minutes. For example, to know the volume of an embankment, instead of measuring cross-sections and calculating manually, simply scanning around the embankment allows for immediate volume calculation. Because measurements are non-contact, waiting time for heavy equipment is reduced and rework due to missed measurements is eliminated. Point cloud measurement can be operated by a small team, making it easy to introduce even at sites with labor shortages, and reducing surveying in hazardous high places or slopes contributes to improved safety.

Furthermore, the acquired point cloud data are valuable as digital records. Because dimensions for needed locations can be extracted later, worries like “I forgot to take a photo” or “I forgot to measure” disappear. If the point cloud is preserved, cross-sections can be created and dimensional checks performed on that data even after time has passed. If shared in the cloud, stakeholders in remote offices can review the 3D data and proceed with inspections. Unlike paper records, digital data do not degrade and are easy to store long-term, making them useful as future evidence or for planning maintenance.

What is 3D model × point cloud difference visualization?

So what exactly does it mean to combine 3D design models (design data) and point cloud data to visualize differences? This involves overlaying the measured point cloud on the 3D design model and color-coding each point to show its deviation from the design. Areas with little difference from the design are shown in green, while areas that are higher or lower than the allowable range are shown in red tones, displayed as a heat map. This makes it immediately clear which areas are overfilled and which are undercut.

This 3D difference check reveals overall finishing trends that were previously overlooked. For example, in pavement work, only a few thickness measurements could be taken conventionally, but with difference visualization the entire finished surface’s irregularities can be evaluated, allowing quantitative assessment of construction precision variability. The Ministry of Land, Infrastructure, Transport and Tourism has recently introduced a new method called “surface management” that evaluates pavement and embankment as-built conditions using areal measurement data such as point clouds, enabling detection of subtle quality differences that single-point inspections could not reveal.

Difference visualization also speeds up and solidifies as-built inspections. Once point cloud data are obtained, there is no need to hurry around the site measuring dimensions manually. Because discrepancies from the design can be automatically extracted in post-scan data processing, nonconforming areas can be identified on the spot and corrective actions taken immediately. Defects that used to be corrected only after being pointed out in final inspections can be discovered and addressed immediately after construction with difference visualization, minimizing rework.

Moreover, the results of such 3D comparisons are useful as explanatory materials for clients and owners. Overlaying as-built point clouds and the design model with images or colored distribution maps intuitively communicates “areas constructed according to the design” and “areas where deviations occurred.” Even those who find technical drawings difficult to understand can grasp the situation at a glance with a color-coded heat map. This reduces misunderstandings and facilitates smoother communication with stakeholders.

Construction management issues solved by difference visualization

As discussed, visualizing differences between 3D design data and point cloud data greatly improves the issues inherent in traditional construction management. The greatest effect is the prevention of oversights and improvement in quality accuracy. By capturing the entire site as data, construction errors and out-of-spec areas can be detected comprehensively, preventing human oversight. As a result, the pass rate for as-built inspections increases and the risk of later being told “this differs from the drawings” is reduced.

Reducing rework is another important effect. When defects are discovered and corrected early through difference visualization, large-scale remedial work or schedule delays can be avoided. By eliminating quality issues early, additional costs can be controlled. By digitally checking as-built conditions continuously, site supervisors can conduct inspections with confidence.

Also, if comparison results between point clouds and design models are stored, they can be used as objective evidence for a long time. Information that cannot be fully conveyed by paper documents or photo albums is detailed in 3D data, helping grasp current conditions for future maintenance or renovation planning. If data are shared in the cloud, supervisors or inspectors in remote locations can share as-built information in real time, enabling smooth remote inspections and meetings. This is a major advantage from the perspectives of work-style reform and promoting DX.

In this way, as-built management that incorporates difference visualization has the potential to improve speed, accuracy, safety, and trust. Moving away from management that relies on craftsmanship and experience, everyone can evaluate quality based on objective data, improving organizational productivity and reliability. Data-driven management also enables consistent quality checks without relying on experience or intuition, which is beneficial for human resource development and skill transfer. The Ministry of Land, Infrastructure, Transport and Tourism is promoting advanced construction management through use of 3D data, and such methods are expected to become the standard going forward.

Closing: Start construction management DX with simple surveying using LRTK

The use of 3D design data × point cloud difference visualization is exactly the kind of initiative that symbolizes DX (digital transformation) on construction sites. Surveying and measurement technologies that support this are also rapidly evolving. One notable development is LRTK, which enables anyone to perform high-accuracy surveying easily using a smartphone. LRTK consists of a small device that attaches to a smartphone and a dedicated app, providing a solution that allows acquisition of high-precision 3D data through simple surveying even without specialized surveying skills.

For example, with LRTK, a construction manager can walk around the site with a smartphone and acquire point cloud data of the site and structures, and measure heights and distances on the spot. As-built measurement that used to require a surveying company can be completed by a single site staff member in a few minutes with LRTK. Based on the 3D data obtained this way, immediate difference checks against the design model and AR-based as-built visualization can be performed—this is a major feature of LRTK. It also supports cloud integration, enabling one-stop sharing of field-acquired data within the company and presentation of 3D views or difference heat maps as reporting materials to owners.

Why not try simple surveying with LRTK on part of your site first? Once you experience its ease and usefulness, it can become an excellent trigger to drive DX across your construction management. Embrace the latest technology and take your company’s construction management to the next stage.

FAQ (Frequently Asked Questions)

Q: Can as-built verification be performed with point cloud data even if there is no 3D design model (CIM/BIM)? A: Even without a 3D model, it is possible to grasp as-built conditions using point cloud data. For example, cross-sections can be generated from the as-built terrain point cloud and compared with design elevations and dimensions. However, it is ideal to have a 3D model prepared from the design stage. With a 3D model, automatic comparison with point clouds becomes easier, enabling more precise and efficient as-built verification. Since CIM/BIM adoption is progressing, it is advisable to utilize design 3D data if possible.

Q: Do you need expensive equipment to acquire point cloud data? A: Not necessarily. While high-precision laser scanners are costly, point clouds can be obtained by relatively accessible means depending on the use case—for example, drones or smartphone LiDAR. Recently, solutions like LRTK have appeared that achieve centimeter-class accuracy (cm-level accuracy (about 0.4 in)) using a combination of a consumer smartphone and a small device. By choosing the appropriate tool according to the measurement target and required accuracy, you can start utilizing point cloud data while keeping costs down.

Q: Can difference visualization be used without specialized knowledge? A: Yes. Modern software and services are designed to be user-friendly, and you can use difference visualization results without advanced CAD skills. Alignment (registration) of point clouds and design data has been automated, and some tools can display color-coded heat maps with one click. The results are intuitive—by simply looking at color differences you can tell where deviations exist. With some operation training if needed, site personnel can sufficiently handle these tools.

Q: How reliable is the measurement accuracy of point cloud data? A: It depends on the measurement method and equipment, but current point cloud measurement technology offers practical accuracy. Terrestrial laser scanners can achieve millimeter-level accuracy, and drone photogrammetry can be within a few centimeters of error if control points are properly placed. Smartphone LiDAR can also achieve a few cm level of accuracy with appropriate techniques. By combining RTK-GNSS as in LRTK, horizontal positioning of about ±1–2 cm (±0.4–0.8 in) and vertical accuracy of about ±3 cm (±1.2 in) can be achieved. With proper calibration and verification, point cloud data have accuracy you can trust for as-built management.

Q: Is it acceptable to use point cloud data for client inspections and approvals? A: Recently, the use of point cloud data in as-built management has been increasingly recognized and recommended by clients. The Ministry of Land, Infrastructure, Transport and Tourism promotes 3D data usage as part of i-Construction, and as-built management guidelines now include areal data evaluation methods such as “surface management” using point clouds. In other words, in addition to traditional point measurements, as-built proof using point clouds is being officially accepted. However, specific procedures may vary by contracting agency or inspection officer, so it is advisable to consult in advance and confirm acceptance of submitted point cloud analysis results. In many cases, color maps showing the entire site created from point clouds are easy for inspectors to understand and are often received favorably.

Q: What is as-built verification using AR like? A: AR (augmented reality) based as-built verification overlays the 3D design model onto the real object via smart glasses or a smartphone, allowing intuitive detection of deviations. For example, with an MR device (commercial smart glasses), inspectors can view a virtual design model superimposed on the actual structure through the goggles and visually check on-site whether construction conforms to the design, enabling inspection without climbing scaffolding and from safer positions. AR is expected as a leading-edge technology for site DX. LRTK can also display acquired point cloud data and design models in AR, helping confirm the locations of buried items and share completion images.