Table of Contents

• Introduction: The Importance of Differential Earthwork Volume in As-Built Management

• Issues with Conventional As-Built Management and Earthwork Volume Calculation Methods

• Innovation in As-Built Management Using Point Cloud Measurement Technology

• Benefits of Smartphone RTK Point Cloud Surveying

• Workflow for Differential Analysis and Automated Pass/Fail Judgment Using Point Cloud Data

• Efficiency Gains in Quality Checks Realized by Automated Pass/Fail Judgment

• Summary: Start Site DX with Simple Surveying Using LRTK

• FAQ: Required Point Cloud Density and Accuracy, Measures for Sites with Poor GNSS Reception, Measurement Points for Flat Areas and Slopes, Methods for Difference Verification and Whether Reports Can Be Output, etc.

Introduction: The Importance of Differential Earthwork Volume in As-Built Management

On civil engineering sites, “as-built management,” which verifies that post-construction shapes and dimensions match the design, and “earthwork volume management,” which accurately quantifies the soil involved in excavation and embankment, are indispensable processes. In particular, correctly calculating and managing the differential earthwork volume—the change in soil volume before and after construction—is critical as it directly affects construction progress (work quantity) management, pass/fail judgments in as-built inspections, and settlement of construction costs. To confirm whether embankments have been placed to the designed volume or whether the required excavation volume has been achieved, it is necessary to survey the terrain before and after construction, calculate the difference in earthwork volume, and check both quality and quantities.

However, conventional surveying methods for grasping differential earthwork volumes and as-built conditions have required a great deal of time and labor, and have faced challenges in terms of accuracy and safety. Recently, point cloud surveying has attracted attention as a technology that solves these issues and dramatically improves on-site productivity and quality control. This article explains in detail the problems of conventional methods and the new as-built management approach using point cloud technology that overcomes them. We hope this provides useful hints for construction managers and survey technicians who want to achieve efficient and high-accuracy site management amid labor shortages and DX promotion.

Issues with Conventional As-Built Management and Earthwork Volume Calculation Methods

Traditionally, post-construction as-built verification and earthwork volume calculations have mainly relied on specialized surveying staff using total stations (TS) or levels to measure key points and then creating drawings and tables to make judgments. For example, in road construction or land development, it has been common to set several baseline cross-section lines, perform longitudinal and cross-sectional surveys at the same positions before and after construction, create cross-sectional diagrams, and calculate volumes from the differences before and after embankment/excavation. On small sites, simplified management that estimates earthwork from the number of dump truck trips and load per truck is sometimes used.

However, these conventional methods have clear limits in measurement coverage and accuracy. Cross-section surveys require interpolation of terrain changes between survey lines, risking oversight of subtle undulations or excesses/deficits in unmeasured areas. Point measurements at several-meter intervals often fail to capture variations between points, and some as-built defects (differences from the design) can be missed. Moreover, it is practically difficult in large-scale earthworks to survey the entire area in detail due to manpower and time constraints, so only key sections were typically measured; this often led to problems where localized over-excavation or insufficient embankment were identified later.

Furthermore, conventional as-built management work requires substantial labor and personnel. Experienced technicians walking the site to take measurements, recording results manually, and reorganizing them into drawings or spreadsheets is highly inefficient. Photographing the as-built and preparing reports also takes time, imposing a heavy burden on site supervisors and technicians. In addition, safety concerns cannot be overlooked. Surveying in hard-to-access areas—high slopes, the underside of bridges, narrow excavation trenches—can be dangerous, and forcing measurements can lead to falls or collapses with serious consequences. As a result, “dangerous-to-measure areas” tend to be insufficiently inspected.

Thus, conventional labor-intensive as-built management faces a dilemma: wanting to measure everywhere but being unable to, and seeking efficiency gains without sacrificing accuracy. The emerging solution is a new method that scans the entire site using three-dimensional point cloud data to understand as-built conditions and earthwork volumes across surfaces.

Innovation in As-Built Management Using Point Cloud Measurement Technology

Point cloud data is a digital record of a surface composed of countless points (each having X/Y/Z coordinates). It can be acquired by laser scanners or photogrammetry, and its characteristic is that it yields a highly detailed 3D model as if the entire site were copied wholesale. By utilizing point clouds, it becomes possible to perform high-density, wide-area as-built measurement at once—something that was difficult with conventional methods—and to detect even slight elevation differences or shape deviations between measured points without omission.

The main reasons point cloud-based as-built management is gaining attention are its efficiency and comprehensiveness. Where conventional site measurement required many personnel and days to measure partially, point clouds can cover the entire site in a short time. Because nothing is left unmeasured and the entire site is digitized, the risk of overlooking quality issues is greatly reduced. Also, since measurement data remains as a digital 3D model, it has high reusability: you can later analyze arbitrary cross-sections or recalculate volumes, reducing the need for additional rework surveying.

From a safety perspective, point cloud measurement is advantageous because it can be done remotely and non-contact. Using drones, long poles-mounted laser scanners, or handheld smartphone LiDAR, dangerous or inaccessible areas can be scanned from a distance. This reduces the need to work on unstable slopes or at heights, enabling both safety and surveying to be ensured. Due to these benefits, the Ministry of Land, Infrastructure, Transport and Tourism is promoting the use of 3D measurement technology by preparing draft guidelines for as-built management using 3D measurement, and point cloud utilization is being advanced across the industry.

Benefits of Smartphone RTK Point Cloud Surveying

There are various methods to obtain point cloud data—terrestrial laser scanners, UAV (drone)-mounted LiDAR, etc.—but smartphone-based point cloud surveying has been attracting particular attention recently. Modern smartphones are equipped with LiDAR sensors and high-performance cameras, enabling easy 3D scanning comparable to dedicated surveying equipment. Combining this with a high-precision GNSS receiver using RTK (real-time kinematic) positioning makes it possible for anyone to perform point cloud surveying with a small, lightweight tool that nonetheless achieves positioning accuracy on the order of a few centimeters.

The benefits of smartphone RTK point cloud surveying can be summarized in three main points. First, ease and labor savings. There is no need to carry heavy tripods or special equipment; you can walk the site with only a smartphone and a small GNSS unit to measure. Work can be completed by a single person, reducing personnel coordination and making it easy to introduce on sites with staff shortages. Second, real-time capability. You can visualize the scanned point cloud on the smartphone screen in the field and proceed while checking for omissions. Because RTK provides continuous position correction, high-precision coordinates are always obtained, allowing instant dimensional checks and volume calculations on site. Tasks that used to require returning to the office for data processing and drawing generation can now be completed in real time on site. Third, low cost and low barrier to entry. Dedicated laser scanners and surveying instruments were expensive and required specialized knowledge, but smartphone RTK uses relatively inexpensive hardware and familiar smartphone apps, so training costs are small. It is easy to roll out to in-house technicians and is a practical first step for DX adoption.

Workflow for Differential Analysis and Automated Pass/Fail Judgment Using Point Cloud Data

Now, let’s look at the basic workflow of how differential earthwork volumes are actually calculated and as-built pass/fail judgments are made using point cloud data obtained by smartphones and GNSS.

1\. Obtaining the as-built point cloud: First, scan the post-construction site using point cloud surveying to acquire the as-built 3D point cloud data. If the point cloud is positioned in a reference coordinate system (public coordinates or a local coordinate system), subsequent comparison work will proceed smoothly. With smartphone RTK, this point cloud acquisition can be completed quickly on site.

2\. Comparison with design data and reference point clouds: Next, overlay the acquired as-built point cloud data with the design data (3D model of the design surface) or the pre-construction ground point cloud. The differences between them are automatically calculated in analysis software, yielding the elevation differences (positive and negative) at each location. Because point clouds obtained with RTK are already in correct coordinate positions, they can generally be compared with high accuracy without additional alignment.

3\. Automatic calculation of differential earthwork volume: From the differences between point clouds, total embankment and excavation volumes are automatically calculated. For example, the volume difference between the as-built point cloud and the design surface immediately reveals the amounts of “excess embankment” or “remaining excavation,” allowing numeric confirmation of soil surpluses or shortages. While volume aggregation was formerly done manually per section, point cloud analysis completes it at the push of a button.

4\. Visualization by heat map: Difference results can be intuitively displayed as a heat map. Concretely, areas higher than the design surface (excess embankment or uncut areas) are shown in red tones, and lower areas (insufficient embankment or over-excavated areas) in blue tones, colored on the point cloud. This makes it immediately apparent where and by how much the site deviates. Subtle trends that would be easy to miss in numeric-only reports can be intuitively grasped in 3D visuals, enabling quick identification of areas requiring correction.

5\. Automated pass/fail judgment: Alongside heat map display, the software can perform automated pass/fail judgments based on as-built management criteria. If you preset allowable tolerances for each work type, the system automatically judges each measurement grid as “pass” or “fail” by comparing the point cloud analysis results. For example, you can display only regions that meet the criterion “within ±5 cm (±2.0 in) of the design surface” in green, and highlight areas exceeding the tolerance—excess embankment or over-excavation—in red. This allows inspection personnel to focus only on the areas the software deems “fail,” dramatically improving the efficiency and reliability of inspections.

6\. Report generation and sharing: Finally, generate reporting materials automatically based on the analysis results. You can output an as-built inspection report summarizing differential earthwork volumes and include heat map images and cross-section diagrams as a PDF, or export the surface model created from the point cloud as CAD data (LandXML or DXF). These electronic outputs can be produced in formats aligned with the Ministry of Land, Infrastructure, Transport and Tourism’s electronic submission guidelines (draft), allowing direct use as digital deliverables. Using a cloud system, point cloud data and reports can be shared internally and externally instantly, facilitating smooth explanations to clients and sharing of correction locations.

Efficiency Gains in Quality Checks Realized by Automated Pass/Fail Judgment

As described above, by performing analysis and automated judgments using point clouds, site quality checks can be made far more efficient while establishing a robust inspection system. Compared with relying on human visual checks and spot measurements, data-driven objective evaluation becomes possible, reducing subjective variability and oversights. Especially when pass/fail judgments are automated, inspectors need only concentrate on areas flagged as “fail” by the software, significantly reducing the burden of checking an entire large site thoroughly.

Records for quality control also improve in accuracy. Previously, records were kept based on paper drawings and photos, but point cloud data and heat maps can store as-built conformity and error distributions at high resolution. In case of disputes, you can later verify precisely when, where, and by how much the site deviated from the design, providing assurance for quality assurance. For communications with clients, color maps of as-built data are persuasive and serve as objective evidence.

Moreover, introducing point cloud-based as-built management contributes to on-site DX (digital transformation). When surveying and inspection processes are digitized, it becomes easier to link and centrally manage data from other processes, and it can form the foundation for advanced applications such as construction automation or AI-based quality prediction. By starting with digital technologies for automating differential volume calculation and pass/fail judgment, you can elevate overall site management to the next stage while achieving both precision and speed in quality checks.

Summary: Start Site DX with Simple Surveying Using LRTK

As discussed in this article, calculating differential earthwork volumes and automating as-built management using point clouds has the potential to dramatically improve on-site productivity and quality. You don’t need to switch everything all at once; even introducing digital measurement in part of your workflow can have significant effects on labor shortages and operational efficiency. Tasks that once required multiple personnel can now be completed with a single smartphone, producing real-time results on site. Experience this ease and high accuracy at your own sites.

The first step we recommend is adopting simple surveying with LRTK. The LRTK series combines smartphones with small GNSS receivers to support high-precision RTK positioning, offering an innovative surveying solution. It delivers ease of use comparable to conventional specialized equipment while achieving competitive positioning accuracy, and supports the workflow from point cloud acquisition and differential analysis to cloud sharing. LRTK also aligns with the Ministry of Land, Infrastructure, Transport and Tourism’s i-Construction initiatives, making it a flagship tool for site DX. For more details, please refer to the official LRTK website. Consider using LRTK to evolve your site management to the next stage.

FAQ: Required Point Cloud Density and Accuracy, Measures for Sites with Poor GNSS Reception, Measurement Points for Flat Areas and Slopes, Methods for Difference Verification and Whether Reports Can Be Output, etc.

Q: What point cloud density and accuracy are required for calculating differential earthwork volumes? Is point cloud obtained by a smartphone acceptable? A: For public works as-built management, a point cloud density of several dozen points per 1 square meter is sometimes recommended. Smartphone LiDAR is not as high-density as a dedicated laser scanner, but if scanned slowly and carefully, it can acquire practically sufficient point clouds. Combining with RTK high-precision positioning can keep positional errors of each point to around a few centimeters, which is sufficient for typical embankment and excavation volume calculations. Except in special cases that require extremely strict millimeter-level accuracy, smartphone point clouds are generally adequate for differential earthwork calculations. If you remain concerned, you can supplement and cross-check key locations with conventional surveying—this hybrid approach is an option.

Q: Can smartphone point cloud surveying be used in locations with poor GNSS satellite reception (mountainous areas, under viaducts, out of communication range, etc.)? A: RTK positioning can become unstable in environments with severely obstructed satellite visibility, but there are countermeasures. First, it is important to obtain reference coordinates at a location with good visibility before starting the scan. In areas where satellites are difficult to capture, using CLAS (the wide-area augmentation service) provided by Japan’s Quasi-Zenith Satellite System “Michibiki” enables real-time corrections reception even in some communication-limited areas, which helps improve accuracy. In places such as tunnels or underground spaces where satellite reception is impossible, real-time position correction is difficult, but you can handle this by aligning point clouds later using reference points obtained near the entrances. In short, even in sites where GNSS is challenging, point cloud surveying is feasible with planning, use of augmentation signals, and post-processing adjustments.

Q: Are there differences in point cloud surveying methods for flat sites and sloped surfaces (cut slopes)? A: The basic scanning procedure is the same, but on slopes you need to pay attention to scanning positions and angles. Smartphone LiDAR is effective within a range of a few meters, so it is difficult to capture a tall slope entirely from the bottom in one pass; you may need to split the scan into upper and lower segments and scan up close in stages. Obtain point clouds from both the top and bottom sides of the slope as needed, then merge the data afterward to ensure full coverage. If you want to efficiently survey wide slopes, combining photogrammetry (smartphone camera) to capture wide-angle records and converting them to point clouds later is also effective. On flat sites, line-of-sight is usually good, making it easier to scan large areas at once; however, if the survey area is extremely large, set several known points (targets) to maintain accuracy through periodic alignment. In any case, planning scan paths with overlapping coverage is key to ensuring accuracy and preventing omissions.

Q: How can differences between point clouds be verified? I’m worried about whether comparisons are being performed correctly. A: Dedicated point cloud processing software and cloud services provide functions to overlay multiple point clouds and automatically calculate differences. As mentioned above, difference results can be displayed on a colored 3D model (heat map), with areas higher than the design in red and lower areas in blue, making deviations immediately apparent. You can also extract longitudinal and cross-sections at arbitrary locations and overlay the cross-section lines from two point clouds to compare them. These features enable visual and quantitative verification of “where” and “how much” the errors are. Point clouds positioned with RTK generally share the same coordinate system, so differential analysis can be done with high accuracy. If you are concerned, check invariant on-site points (edges of immovable structures, etc.) in both datasets to confirm that reference elevations match.

Q: Can drawings and forms for reports be output from data obtained by point cloud surveying? Can it support electronic submission? A: Yes. Point cloud analysis systems can output various deliverables based on acquired point cloud data and calculation results. For example, you can output a PDF report summarizing differential earthwork calculations with one click, or save automatically generated longitudinal/cross-sections from the point cloud in DXF format. Exporting the surface model of the as-built terrain in LandXML format allows direct use in other civil CAD or machine guidance systems. These electronic outputs conform to the Ministry of Land, Infrastructure, Transport and Tourism’s electronic submission guidelines (draft) and fit smoothly into conventional inspection and delivery workflows. In other words, data from smartphone point cloud surveying is designed so that it can be submitted as official deliverables, making post-introduction documentation reliable.