Dramatic Efficiency with Point Cloud Data! Approaching the Front Lines of Automated Civil Engineering As-Built Management

As-built management (dekigata kanri) is indispensable in civil engineering work, but traditional methods are plagued by time-consuming tasks and the risk of errors. “Forgot to take the crucial photos,” “The as-built records don’t match the drawings”—have you ever experienced this? As-built management is the process of verifying and recording that the shape and dimensions of completed structures conform to the design, but manual measurements and photo management often leave gaps, and omissions or inconsistencies may be discovered later, causing trouble. Point cloud data–based automation is attracting attention as a trump card that dramatically solves these issues and dramatically improves the accuracy and efficiency of as-built management.

Point cloud data are three-dimensional datasets consisting of countless points obtained by laser measurement or photogrammetric analysis, where each point has XYZ coordinates (position) and color/intensity information. They can record detailed shapes as if the entire site were scanned, essentially serving as a full-scale 3D copy of the site. In recent years, supported by the Ministry of Land, Infrastructure, Transport and Tourism’s *i-Construction* initiative, point cloud technologies such as drone surveying and 3D laser scanners have become more widespread. In the as-built management field, the era in which “measuring and verifying with point clouds is the norm” is approaching, and the introduction of high-precision point cloud measurement is expected to enhance both quality control and efficiency. This article provides a comprehensive explanation of the front lines of point cloud utilization in civil engineering as-built management, from methods of acquiring point cloud data to utilization techniques and the mechanisms of automatic judgment. At the end, we also introduce the latest easy point cloud surveying method using smartphones combined with compact GNSS, LRTK, offering practical hints for field implementation.

Means of Acquiring Point Cloud Data (Drone, Laser Scanner, Smartphone)

First, let’s look at representative ways to acquire point cloud data. Depending on the purpose and scale, the following methods are available.

• Drone (UAV) Photogrammetry / LiDAR: Small unmanned aerial vehicles equipped with cameras or lasers measure the site from above. There is a photogrammetry technique that reconstructs 3D shapes from many aerial photos via software, and a method that directly acquires point clouds by mounting a laser scanner. The advantage is being able to survey wide areas in a short time, making it suitable for grasping as-built conditions of large sites such as reclaimed land or dam construction. However, special measures are required to measure shaded areas that are not visible from above.

• Terrestrial Laser Scanner (TLS): Tripod-mounted or mobile laser scanners are used on the ground to irradiate laser light horizontally and vertically to acquire surrounding point clouds. They can obtain millions of high-density points at once and can record fine surface irregularities of structures with high accuracy. They are powerful for as-built measurement of complex shapes such as bridges, tunnel interiors, and retaining walls. Equipment is expensive, but accuracy is extremely high—on the order of a few millimeters to a few centimeters—and suitable for detailed as-built inspection.

• Smartphone-mounted LiDAR / Photo Scanning: Some high-end modern smartphones have small LiDAR sensors built in, enabling easy acquisition of surrounding point clouds. There are also apps that create 3D reconstructions by taking multiple photos with a smartphone or tablet camera. Smartphone 3D scanning is convenient for measuring as-built conditions in confined areas or for quick supplementary measurements, and its ease of use by field staff is attractive. Although a standalone smartphone has limitations compared to laser scanners in accuracy and range, practical accuracy can be achieved through measures such as combining with GNSS, discussed later.

By combining these means, point cloud data that cover the site from air to ground can be obtained. For example, on a large earthwork site, a drone can scan the overall terrain, terrestrial laser scanners can supplement the details, and a smartphone can be used for narrow spaces or fine checks. While manual surveying used to be mainstream, these methods that can now measure 3D in high density over wide areas in a short time are becoming the new standard.

Measurement Workflow and Points for Improving Work Efficiency

Let’s outline a typical workflow for as-built management using point clouds and organize the key points for improving efficiency. The rough steps are as follows.

• Planning and Establishing Reference Points – First, plan according to the measurement range and required accuracy. If necessary, install known points (reference points or control points) on site to ensure the accuracy of subsequent point clouds. Even in traditional as-built management guidelines, surveying reference points is indispensable, and similarly with point cloud measurement a reference coordinate network is important.

• Point Cloud Data Acquisition – Acquire the site’s point cloud data using the methods described in the previous section (drone, TLS, smartphone, etc.). A laser scanner can sweep 360 degrees of the surroundings at the push of a button, and a drone can perform aerial photography and laser measurement with automated flight. Compared to conventional methods that measured each location manually, one measurement can collect millions of points in a short time, dramatically improving surveying and recording efficiency. There are reports where surveying that used to take several days on large reclaimed land could be completed in about half a day with a drone.

• Data Processing and Analysis – Process the acquired point cloud data with dedicated software or cloud services. If there are multiple scan datasets, perform alignment (composition/integration) and remove unnecessary noise points. In photogrammetry, point cloud generation from images is performed at this stage. Modern point cloud processing software is fast and increasingly automated, allowing analysis on a field laptop or tablet in a short time. Using cloud services, heavy processing can be executed on servers, leaving the field to simply wait for results. The ability to perform near-real-time as-built verification is a key point for improving efficiency.

• Extraction and Judgment of As-Built Conditions – Once point cloud processing is complete, extract the information necessary for as-built management from the point cloud. Specifically, this includes creating cross-sections as described later, measuring deviations by comparing with design data, confirming dimensions at specified locations, and calculating volumes. Many of these can be automatically calculated and judged in software, significantly reducing the need for humans to manually calculate or annotate drawings. As a result, the inspection process speeds up, reducing the burden on staff and helping to prevent human errors through double-checking.

• Deliverables Preparation and Sharing – Finally, organize the inspection results and point cloud data and compile deliverables such as reports and drawings. Combining cross-sections generated from point clouds, deviation heat maps, and quantity calculation sheets makes it possible to create as-built documentation with higher objectivity than before. The point cloud itself can also be saved and submitted as electronic delivery data to share information with the client. The ease of data sharing and utilization is another efficiency advantage of point cloud adoption.

By following this workflow, as-built management can be digitized and automated from field measurement through inspection and reporting. In particular, if acquisition and analysis/verification proceed in parallel, “once measurement is complete, immediate pass/fail judgment on site” becomes realistic, enabling rapid construction management. In practice, ICT construction sites have reported completing surveying and as-built inspection within 1–2 days, greatly improving productivity.

Methods of As-Built Evaluation Using Point Cloud Processing (Cross-Section Comparison, Surface Deviation, Volume Calculation, etc.)

By utilizing point cloud data, advanced as-built evaluations that were previously difficult can be performed efficiently. Here are the main evaluation and analysis methods.

• Comparison of Arbitrary Cross-Sections: You can extract cross-section drawings from point clouds at any desired locations and overlay them with the design standard cross-section. For example, you can continuously create cross-sections of roads or embankments and check whether the design alignment and the actual as-built are misaligned. With point cloud measurement, all locations—not just scheduled measurement points per control criteria—can be cross-sectioned, enabling detection of local inconsistencies that were previously overlooked.

• Surface Deviation Heat Map: This analysis overlays the acquired as-built point cloud with the designed 3D data (CAD model or design surface) and color-codes the elevation differences of each point. Known as a deviation map, higher-than-design areas appear red and lower areas blue in a heat map, allowing immediate visualization of the distribution of errors across the site. This is also used for automatic judgment in as-built inspection; because it visually indicates the areas outside the standard, subsequent rework planning becomes easier.

• Volume and Quantity Calculation: Volumes of embankment or excavation can be accurately calculated from point cloud data. By comparing pre- and post-construction ground point clouds and calculating the differential volume, progress quantity calculations and the creation of as-built quantity statements can be automated and accelerated. Areas and thickness measurements are also freely available; for example, you can batch-check whether specified structure dimensions (widths or thicknesses) remain within tolerance throughout the entire section. Once point clouds are acquired, various quantity calculations and dimensional verifications can be performed without additional field surveying, which is a major advantage.

Beyond these, point cloud data can be used to create as-built drawings (as-built plans) or integrate measured data into design BIM/CIM models to generate as-built models, expanding the scope of applications. All these methods rely on objective data obtained from a large number of points, enabling evaluations more reliable than traditional methods. The groundbreaking point of point cloud utilization is that construction quality can be judged digitally without relying on human subjectivity.

Accuracy Management and Verification Methods for Point Clouds

Accuracy management and verification of data are indispensable when using point cloud measurement for as-built management. No matter how convenient it is, if accuracy is lacking it cannot be used for official inspections. The following measures help ensure reliability.

First, alignment using reference points and control points is fundamental. Known markers (targets) installed before measurement are measured at the time of point cloud acquisition, and by assigning their coordinate values, the point cloud is corrected to the correct position and elevation. For example, even with laser scanner measurement, placing multiple control points around the site and reading them allows absolute coordinates of the point cloud to be aligned to millimeter-to-centimeter accuracy. The Ministry of Land, Infrastructure, Transport and Tourism’s as-built management guidelines also stipulate performing accuracy checks using reference points, and the introduction of control points has become an official requirement. The same applies to drone photogrammetry: appropriately placing and surveying ground control points (GCPs) can greatly improve the accuracy of photogrammetry-derived point clouds.

Next, it is important to understand the accuracy characteristics of equipment and methods. With laser scanners, measurement distance accuracy and scan range affect results; with photogrammetry, pixel resolution and the number of photos influence point cloud density and accuracy. Conducting trial measurements in advance and checking errors over known distances and elevation standards provides assurance. For example, measure the distance between two widely separated points and verify that the difference from the known value falls within required accuracy. It is desirable to perform practical accuracy verification under the conditions of each site and adopt point cloud data as as-built deliverables based on those results.

Additionally, perform post-acquisition point cloud data inspection. You can check registration errors between point clouds in software or compare the point cloud to coordinates of check points measured separately on site. For example, for important structural points, compare the point cloud’s elevation values with heights measured by leveling to record errors. If such verification supports a statement like “point cloud measurement accuracy is within ±○ cm,” you can confidently submit the data to the client.

*Note*: The Ministry of Land, Infrastructure, Transport and Tourism has recently developed a draft “As-Built Management Guidelines Using 3D Measurement Technologies,” indicating procedures for using point cloud data for as-built management in earthworks and slope works. Point cloud measurement is increasingly recognized as a formal as-built management method conforming to these official guidelines, and with appropriate accuracy management it can be used for inspections equivalent to traditional methods.

Mechanisms and Examples of Automatic Judgment Functions

One of the most notable uses of point cloud data is the automatic judgment function for as-built conditions. This system digitally compares the acquired point cloud with design data on a computer and makes pass/fail judgments for construction. Tasks that used to require inspectors to compare drawings and the site or to manually check errors at measurement points are now executed en masse by software.

The mechanism of automatic judgment is based on results from the aforementioned deviation heat maps and cross-section comparisons, detecting points that fall outside pre-set tolerance ranges. For example, if the allowable deviation for a road is set to ±5 cm (±2.0 in), the software will color-code or output to a warning list any points on the point cloud that are 5 cm or more above or below the design model. There are also functions that automatically judge “pass/needs correction” by measuring concrete thickness or width for each control cross-section. In short, because the machine determines “within standard or out of standard,” humans only need to review the results, shortening inspection time and preventing human error.

Example: In one infrastructure project, point clouds were analyzed on a smart construction system and as-built inspection was automated by calculating differences from the design 3D data. Immediately after construction, the site was scanned and the data imported into a tablet app, where a deviation map was generated within minutes. The inspector could instantly grasp construction accuracy via the tablet’s color-coded display. If any areas fell outside standards, the system immediately showed the extent and degree, enabling prompt rework. Where inspections had previously taken more than half a day, data processing and judgment were completed on site, greatly improving as-built inspection efficiency. Results were also shared in real time via the cloud with office superiors and the client, allowing on-the-spot approval. The use of such automatic judgment functions not only speeds up inspections but also helps prevent rework and facilitate smooth consensus building among stakeholders.

Centralizing Information through Data Organization and Cloud Sharing

A key challenge in handling point cloud data is how to organize and share large volumes of information. Traditionally, records were managed with paper drawings and photo books, but in the point cloud era centralized digital data management is critical. This is where cloud services demonstrate power.

Storing large-capacity data such as point clouds in the cloud allows stakeholders to access it anytime via the Internet. Using a dedicated viewer, you can view and measure 3D point clouds from an office PC without being on site, enabling location-independent information sharing. For example, construction managers, designers, and client engineers can review the same point cloud data in an online meeting to confirm as-built conditions. Uploading to the cloud also automates version control so the latest information is always consolidated in one place.

Cloud services that allow point cloud data processing and analysis online have also emerged. Without installing software on individual PCs, users can create cross-sections, measure distances, and calculate volumes using only a web browser, enabling clients without specialized software to easily utilize the data. There is no need to send massive point cloud files by email; issuing a viewing link lets recipients display 3D data with a single click. Access control lets you finely manage who can view which information, ensuring confidentiality. In short, cloud-based data sharing creates an environment where everyone views the single latest dataset, bridging information gaps between site and office and between contractor and client.

Organizing data by linking point clouds with related materials also improves efficiency. For example, adding metadata to point clouds (acquisition date/time, survey equipment, weather conditions, etc.) or leaving annotations for inspection points on the point cloud makes later review easier. If drawings, photos, and reports are consolidated in a cloud project folder, all information related to as-built management is centralized, saving time searching. When you need to review the past, the digital records stored in the cloud will be a powerful ally.

Benefits of Adoption from the Client and Inspector Perspective

What benefits does adopting point cloud as-built management bring to clients and inspectors? Here are the main points.

• Objective and Detailed Quality Verification: Point cloud data allows verification of construction results with objective 3D information. Compared to traditional methods that judge pass/fail based on limited measurement points, point clouds capture the entire structure’s as-built condition, reducing the chance of oversight and subjective judgment. This increases confidence in quality inspections for clients and facilitates acceptance.

• Efficiency and Labor Saving in Inspections: Automated analysis of digital data speeds up inspections and reduces the burden on inspectors. Tasks like standing on site for long periods measuring dimensions or checking photos decline, enabling inspections with fewer people and shorter time. Because as-built conditions can be confirmed via the cloud even at remote sites, travel time is reduced and remote supervision becomes feasible.

• Long-term Preservation and Traceability of Records: As-built records obtained by point clouds can be stored digitally long-term, with lower risk of degradation or loss than paper drawings or photo books. If problems arise in the future, the original construction details can be accurately reconstructed for verification. For clients, retaining point cloud data as post-completion quality assurance material is highly valuable.

• Prevention of Disputes and Claims: Objective point cloud records can provide clear evidence in disputes with clients over questions like “Was the work really carried out according to design?” or in later claims regarding as-built defects. Data-based, highly transparent explanations strengthen trust between contractor and client.

• Compliance with National Policy and Improved Evaluation: Adopting ICT-enabled construction methods and applying CIM principles promoted by the Ministry of Land, Infrastructure, Transport and Tourism is another advantage. Clients who adopt advanced construction management methods may receive favorable evaluations from higher authorities and enhance internal DX achievements. Point cloud utilization is not only labor-saving but can be presented as smart construction in line with modern trends.

Point cloud as-built management thus offers many benefits not only to contractors but also to clients. It achieves both improved inspection accuracy and efficiency, creating a win-win technology. Field inspectors may feel uneasy about new digital methods, but with proper training and operation they can realize “reliable and easy inspections,” so adoption should be actively considered.

Potential for Integration with Future Automated Construction and Remote Inspection

Point cloud utilization has strong potential to integrate with next-generation construction management such as automated construction (robotic construction) and remote inspection. The construction industry is advancing efforts to reduce manpower and improve efficiency through ICT and mechanization, and point clouds provide key information for these initiatives.

For example, on automated construction sites heavy machinery operates under automatic control based on 3D design data, and point cloud measurement can be used for feedback in as-built verification. A cycle of construction → point cloud inspection → immediate feedback of as-built deviations → additional machine correction as needed would allow machinery to complete precision work without human intervention. Trials are already underway where daily embankment quantities are calculated using drone point clouds and reflected in machine work plans. If point cloud data are linked in real time with construction processes, more advanced autonomous construction will become possible.

Point clouds also have high affinity with remote inspection and remote monitoring. If point clouds are shared via the cloud, inspections can be performed from the office without visiting the site. In fact, during the COVID-19 pandemic when reducing travel became necessary, attempts were made to replace on-site inspections with online review of on-site-scanned point cloud models. In the future, combining high-resolution point clouds with VR/AR technology may enable inspection experiences equivalent to being on site while located remotely. This would dramatically improve client inspections and enable location-independent construction management.

In this way, point cloud utilization becomes more than a mere recording method; it will be a foundational technology supporting construction site DX (digital transformation). It will create synergy with automated construction systems, AI-based quality judgment, and remote site management, contributing to safer and more productive construction sites.

Practical Tips for Field Introduction: Easy Point Cloud Surveying Starting with Smartphone × Compact GNSS

Although point cloud as-built management is attractive, some may worry that specialized equipment and advanced skills are required and the barrier to entry is high. Until recently, 3D laser scanners and high-performance drone surveying systems required investments of several million yen and operators needed extensive training. However, today’s era makes it possible for anyone to easily measure high-accuracy point clouds by simply combining a smartphone with a compact GNSS receiver.

For example, solutions like LRTK (L-R-T-K) use a palm-sized RTK-GNSS device that attaches to the back of a smartphone and a dedicated app, allowing a single field staff member to walk around and quickly scan surrounding point clouds. By combining point clouds obtained with a smartphone’s built-in LiDAR sensor or camera with centimeter-level positioning information provided by RTK, survey accuracy comparable to stationary equipment can be achieved. You can also compare on-site data with design data immediately and check as-built conditions—displaying colorized deviation maps on the screen right after acquisition. Acquired data can be uploaded to the cloud and shared instantly, enabling real-time collaboration with supervisors or clients in the office. It’s truly a revolutionary system that makes 3D as-built measurement and verification available anytime, anywhere, by anyone.

The emergence of smartphone surveying tools like LRTK has dramatically lowered the barrier to point cloud adoption. There is no need to carry specialized equipment; simply keep a smartphone and a small device in your pocket and you can take measurements whenever needed. Apps are designed so that starting/stopping positioning and saving data are one-touch operations, with usability designed for non-specialists. The compact design weighing only a few hundred grams makes it easy to carry daily and incorporate into routine tasks. In addition to point cloud measurement, the same system can be used for navigation of pile-driving positions and AR displays for guiding heavy equipment, contributing comprehensively to on-site productivity improvement.

To fully realize the benefits of point cloud utilization discussed above, active adoption of the latest technologies is essential. Fortunately, the smartphone × compact GNSS approach offers an accessible 3D as-built management solution you can start using as soon as tomorrow. It’s an excellent opportunity to promote site DX while avoiding large initial investments. To evolve civil as-built management into smarter and more reliable practices, consider introducing point cloud data utilization into your sites. Recording the entire site with point clouds and achieving reliable as-built management and trouble-free construction can begin with the trustworthy ally of simple smartphone × GNSS point cloud surveying.