Improving Efficiency and Accuracy of Earthwork Volume Calculation Using Point Cloud Scanning

In construction sites for land development, excavation, slope maintenance, and roadworks—where survey tasks are handled by civil contractors and municipal staff—the calculation of earthwork volumes (volume calculation) associated with fill and cut is an essential task for progress and quantity management. Traditionally, this volume calculation was commonly done by measuring terrain cross-sections with surveying instruments such as total stations and levels, and manually computing volumes using methods like the average end area method. However, in recent years the spread of point cloud scanning technology has begun to significantly change the earthwork volume calculation process. Beyond drone photogrammetry and terrestrial laser scanners, 3D point cloud surveying using smartphone LiDAR has also emerged, and from the high-density point cloud data acquired by these methods it has become possible to determine fill and cut volumes more quickly and accurately.

This article explains in detail the principles of volume calculation using point cloud scanning and how it differs from conventional methods, as well as the key points of improved work efficiency and accuracy improvement (prevention of human error) resulting from its adoption. We also introduce concrete use cases on actual sites such as land development, excavation, slope management, and road construction, and touch on technical topics such as point cloud data processing, mesh modeling, and cross-section comparison. At the end of the article, we present an easy 3D surveying solution using smartphones with the latest simple surveying tool LRTK and propose methods for introducing it to the field.

Principles of Volume Calculation with Point Cloud Scanning and Differences from Conventional Methods

The basic principle of earthwork volume calculation using point cloud scanning is simple: compute the volume difference between a reference model (ground surface) and a comparison model. For example, in land development, compare the pre-construction original ground model with the completed fill model after construction to calculate the fill volume. By overlaying two terrain datasets in three-dimensional space, the gaps that remain where they do not fully overlap indicate the “missing soil,” while the parts that protrude beyond the overlap indicate “excess soil.” Integrating those volumes numerically yields the cut and fill volumes respectively. Traditionally, surveyors measured the site and drew cross-section diagrams, then calculated volumes using methods like the average end area method; but by using point cloud data, one can compare actual measurements of the entire ground surface, enabling high-precision volume calculations that reflect minute undulations that manual methods might overlook. Also, once a point cloud is acquired, it is easy to perform additional volume calculations for arbitrary areas later, allowing the data to be used for a variety of quantity estimations as needed. Even if you simply want the volume of a single fill or spoil pile, you can outline the fill portion on the point cloud and use the surrounding ground as a reference plane to automatically measure the volume protruding from that plane. In short, by digitizing the entire terrain with point cloud scanning, high-accuracy volume calculation can be achieved for both fills and cuts.

Before the spread of point cloud scanning, the following conventional methods were used. Each method has advantages, but they also had various challenges in terms of manpower and time.

• Staff and level cross-section surveying: The most basic method uses a level and a staff to measure ground elevations at regular intervals, create cross-section drawings, and compute volumes using the average end area method. While simple, on large sites the number of measurement points is limited and local undulations are difficult to capture, and the process from surveying to calculation and drawing creation required substantial effort and time.

• Total station (TS) surveying: TS surveying using an electronic distance meter and prism is a standard method for obtaining high-accuracy three-dimensional coordinates. Many points are observed to create an as-built terrain model, from which volumes are calculated. However, surveying generally requires two or more people, and covering a wide area requires repeatedly relocating the tripod and taking measurements, which is time-consuming. If measurement point intervals are coarse, there is a risk of missing fine irregularities that cause volume errors, and to increase accuracy manually one must increase the number of measurement points.

• UAV (drone) photogrammetry: Drone aerial photography combined with software to generate 3D models or point clouds from multiple photos allows rapid measurement of wide areas from the air, and recent advances in image analysis have improved accuracy. However, drone operation requires expertise and flight permissions, and cannot be used in flight-restricted areas such as urban areas or around airports. High-accuracy surveying also typically requires the placement of ground control points (GCPs) or RTK-GNSS position correction, and the image processing and point cloud generation demand time and advanced software. It is also weather-dependent, so although convenient, it is a method that has site limitations.

• Terrestrial laser scanner (TLS): Using a high-performance laser scanner mounted on a tripod to directly measure the ground surface can acquire high-density point clouds numbering in the millions. It can measure down to the millimeter level and provides extremely high accuracy for earthwork volume calculation. On the other hand, the equipment is large and expensive and requires specialized knowledge to operate. To measure wide areas with obstructed lines of sight, multiple scans from different setup locations are needed, and the acquired data must later be registered (aligned), which is time-consuming. The operational and data processing burdens make frequent surveying difficult.

As described above, conventional earthwork measurement methods each have strengths, but share a common issue: heavy demands on manpower and time. Not only does the surveying itself take time, but considerable manual work is required to organize the acquired data, calculate volumes, and compile reports. Also, because multiple specialized devices and skills are required, sites suffering from chronic personnel shortages may be unable to conduct surveys frequently enough. In addition, even advanced methods such as drones and TLS have constraints due to weather and safety management, so they are not “available anytime, anywhere.” Against this background, there is a growing demand at sites for a more convenient and rapid way to measure earthwork volumes. The new solution that meets this need is the introduction of 3D point cloud scanning technology. In the next chapter, we will look in detail at how using point cloud scanning improves work efficiency and accuracy.

Improved Work Efficiency: Measuring Earthwork Volumes with Fewer People and in Shorter Time

Introducing point cloud scanning technology dramatically improves the efficiency of earthwork measurement. Processes that used to take days to weeks from surveying to calculation have in some cases been significantly shortened. The main efficiency points are summarized below.

• Significant reduction in work time: The time required for field surveying and volume calculation is dramatically reduced. For example, at a large land development site, a measurement and calculation task of finished earthwork that previously required a survey team of four people working one week (a total of 28 person-days) was completed with point cloud scanning by two people in one day (2 person-days). That represents about 1/14 in manpower-days and about 7% of calendar days. This efficiency comes from the ability of point clouds to measure the entire site as a surface at once; the traditional staged process of “survey → drawing → manual calculation” can be largely skipped because measurements immediately yield results on a computer. This directly shortens construction schedules and reduces labor costs, improving overall site productivity.

• Reduction in required personnel: By using point cloud scanning, surveying tasks can be completed with a small number of people. Conventional TS surveying required two persons (an operator and a staff holder, sometimes with an assistant), and drone photogrammetry required at least one to two people for operation and monitoring. In contrast, handheld laser scanners or smartphones used for point cloud acquisition can be operated by one person in many cases. A single field operator carrying the device can obtain data without stopping other workers, significantly reducing the labor cost of assigning personnel for surveying. This impact is especially large in situations with severe labor shortages, as one device per person enables as-built management.

• Immediate processing and rapid decision-making: Data obtained from point cloud measurement can be processed for volume calculation and analysis immediately on dedicated software or cloud services. Previously, data measured on site had to be brought back to the office and entered into specialized software for volume calculation, and creating drawings and reports could take a full day or more. After introducing point cloud scanning, these processes are automated or omitted, allowing results to be known the same day. This enables field personnel to determine on the spot “how many cubic meters of fill are still needed” or “how much has been excavated,” facilitating immediate decisions for subsequent work. Dispatch and scheduling of dump trucks and heavy equipment can be optimized in real time, allowing quick responses and plan adjustments to changing conditions. Eliminating the need to wait for reports speeds up the entire construction management cycle.

• Improved safety and flexible measurement: Site environment and safety also affect surveying efficiency. Traditionally, when measuring steep slopes, workers had to climb dangerous inclines to hold a staff or measure distances. With point cloud scanning, an entire slope can be measured non-contact from a safe location such as the slope foot. Moreover, terrain can be scanned remotely even in areas with poor footing or near operating heavy machinery without interrupting work, achieving both safety and efficiency. Point cloud scanning is also effective where drones cannot fly, such as in urban areas or indoor spaces, because a person can walk to perform measurements. Laser scanners are less affected by light rain, and the flexibility of being able to “measure whenever you want” is a major advantage. Thus, point cloud scanning reduces time and personnel requirements while providing real-time feedback and improved safety, dramatically increasing overall operational efficiency.

Accuracy Improvement and Prevention of Human Error: Reliability of Point Cloud Data

Introducing point cloud scanning also significantly contributes to improved measurement accuracy. Because point cloud data consist of numerous points, the ground surface can be measured almost continuously, capturing minute irregularities and local depressions that manual surveying may miss. As a result, volume calculations reflect details that human surveying might overlook, producing reliable figures that do not rely on estimation. In comparative verification between conventional methods and point cloud-derived volumes, volumes calculated from point cloud surveys have been confirmed to be within about 1–2% error of values obtained by traditional cross-section methods. When point cloud data are acquired following appropriate procedures, volumes can be obtained with accuracy comparable to conventional methods.

Point cloud scanning also helps reduce human error. In conventional workflows, reading measurement points, manual calculations, and spreadsheet processing can introduce reading errors, recording mistakes, or calculation errors. With point cloud surveying, the process from measurement to calculation is digitally continuous, reducing the opportunity for subjective or manual errors. Differences in skill level or judgment among surveyors are minimized, making it easier to ensure consistent quality regardless of who performs the measurement. If scanned volumes that are auto-calculated are shared on-site, transmission errors among multiple personnel are prevented. In this way, digitalization and automation prevent human error and enhance the reproducibility and objectivity of survey results.

The reliability of acquired point cloud data itself is also extremely high and useful as a digital record of the site. A 3D point cloud acquired once can be reused for other analyses later as needed. For example, if you later need a cross-section of a specific area or want to know the settlement amount of a filled portion from that time, analysis can be performed from the originally acquired point cloud without additional measurement. Because point clouds can archive the site situation at a level not possible with paper drawings or photos, they serve as insurance against re-measurement due to omissions or mistakes in records.

Furthermore, the use of point cloud-based quantity calculation methods is gradually being recognized by public institutions. Guidelines such as the Ministry of Land, Infrastructure, Transport and Tourism’s “Guidelines for Quantity Calculation of Civil Engineering Works (draft)” have begun to include methods using point cloud data obtained from drone surveys or 3D laser scans. If conditions are met, such data can be treated as equivalent results to conventional surveying under evolving regulatory frameworks, and the number of cases where point cloud-derived figures are formally used in progress reports and inspections is likely to increase. From a regulatory standpoint, point cloud scanning is becoming a new standard with guaranteed reliability.

Implementation Examples and Specific Use Cases

Point cloud scanning for earthwork volume calculation is being applied across various civil and construction site operations. Below are some typical use cases.

• Progress management of fill and cut: For road construction or land development, accurately grasping terrain changes due to fill and cut is crucial. By acquiring point cloud data of the ground surface before construction and after completion—or for each construction section—and calculating the differential volume between the design surface and the original terrain, objective quantities for progress can be shown. On actual development sites, point cloud data are used to compare planned and actual volumes early on to detect shortages or surpluses of fill material.

• Application to slope management: Point cloud scanning is also powerful for maintaining steep slopes and embankments. For example, at a site where a slope collapsed after heavy rain, quickly calculating the volume of collapsed soil allows immediate determination of how much material must be removed and restored. During slope finishing, cross-section comparisons of point cloud data can check deviations from the design slope and shape. Since dangerous slopes can be measured remotely without personnel entry, slope geometry and volumes can be monitored frequently while ensuring safety.

• Management of spoil and material stockpiles: Point cloud data are useful for managing volumes of spoil piles generated on-site or stockpiles of materials such as crushed stone and gravel. Where pile volumes were previously roughly estimated based on shape, drone scanning from the air or ground-based handheld scanners and smartphones can generate point clouds to calculate accurate volumes. This enables data-driven inventory control and transport planning. In one reported case, a site supervisor quickly scanned a small spoil pile with a smartphone, determined its volume on the spot, and decided the number of dump trucks needed.

• Verification of excavation and backfill volumes: For works involving excavation and backfilling such as sewer installation or foundation embedding, point cloud-based volume verification is effective. By acquiring point clouds before and after excavation and calculating the difference, excavation and backfill volumes can be computed to verify whether the prescribed volume was properly backfilled or whether over-excavation occurred. The volume of rock extracted during excavation, which is hard to predict on-site, can be immediately calculated with point cloud measurement and used for transport planning (note: for complex-shaped boulders the point cloud may not capture cavities on the backside, which can be a source of error, so scanning from multiple angles is necessary).

• Progress management during construction (time-series recording of as-built): In large earthworks, it is becoming common to regularly acquire as-built point clouds weekly or monthly and record construction progress in 3D. Comparing time-series point clouds allows visualization of “how much volume changed at what stage.” This enables early correction of deviations from plans and serves as evidence data for progress reports. Point cloud data can be shared via the cloud among stakeholders, so during site inspections and meetings with clients and supervisors the same 3D model can be viewed for efficient explanation and confirmation.

Technical Explanation of Point Cloud Data Processing, Mesh Modeling, and Cross-Section Comparison

To accurately calculate volumes from point cloud data acquired by scanning, several technical points should be kept in mind. The main processing steps and tips are summarized below.

• Unified surveying reference (alignment): When comparing multiple point cloud datasets, it is essential that they are measured in the same coordinate system and reference. For example, if taking differences between pre- and post-construction point clouds, both must be based on a common survey reference. In drone photogrammetry, set control points (targets) in advance; in laser scanning, perform initial instrument alignment and GNSS-based positioning correction. Proper position calibration at the acquisition stage is necessary. If the reference is misaligned, post-processing software can perform coordinate transformation and registration (alignment of point clouds) to ensure the point clouds overlap correctly.

• Filtering out unwanted points: Point cloud data can include unwanted points not part of the ground surface (moving vehicles, workers, and other noise), and points outside the target area for volume calculation can remain and cause errors. Therefore, it is important to remove unnecessary data before analysis. Use a dedicated viewer to delete noise points or clip the data outside the analysis range to prepare as clean a point cloud as possible. Filter out non-ground points such as vegetation and structures as needed so that only the terrain surface remains, enabling accurate calculations unaffected by extraneous noise.

• Mesh modeling (surface generation): Point cloud data are a collection of discrete points, but to perform volume calculations they must be treated as a surface. Traditionally, point clouds are converted into a TIN (triangulated irregular network) or grid mesh and modeled before computing volumes. Recently, some point cloud viewers and analysis software can directly compare point clouds and compute volume differences, but behind the scenes they automatically convert the data to meshes or voxels for shape representation. In any case, it is important that the conversion from point cloud to surface is performed accurately. During mesh generation, appropriate interpolation should be applied according to point density to create a smooth ground surface model. Extremely sparse point clouds can produce inaccurate mesh shapes, so securing sufficient point density during measurement is also important.

• Volume calculation and difference visualization: Once the reference and comparison surfaces (point clouds or generated surface models) are prepared, analyze the height differences between the two to compute differential volumes. By specifying the comparison targets in analysis software, results such as “X cubic meters of fill and Y cubic meters of cut within the specified area” are automatically displayed. Some software also shows a color-coded heat map of differences. For example, when overlaying a design model and an as-built point cloud, surfaces meeting the design height can be shown in blue, while areas below the design indicating fill shortages can be shown in red, providing color visualization of the terrain. This makes it immediately apparent where soil is sufficient and where it is over-excavated, allowing both numerical and visual confirmation. Using such visualization together with numeric volume results facilitates smooth information sharing among site personnel and as-built inspections.

• Creating and comparing cross-sections: It is also easy to extract longitudinal and transverse cross-sections from point cloud data at arbitrary locations. Traditionally, shape could only be understood along predefined section lines set during surveying, but with point clouds you can generate as many cross-sections as needed afterward. For example, in roadworks you can overlay the design cross-section and the measured as-built section, or evaluate deviations between the planned slope line and the actual slope cross-section—all with one click. Cross-sections are useful for confirming subtle shape differences that may not be clear from volume numbers alone and serve as material for judging as-built quality. Data obtained from point cloud scanning thus enable flexible multi-faceted analysis after measurement, which is a major strength.

Easy Field Survey with a Smartphone: Promoting On-site DX Using LRTK

Some may worry, “I understand that point cloud scanning is useful, but isn’t it necessary to prepare an expensive laser scanner or drone to introduce it?” To address that concern, we introduce here a convenient 3D surveying tool that uses a smartphone. A representative example is the smartphone surveying system called LRTK.

LRTK is a solution consisting of an ultra-compact RTK-GNSS receiver (weighing only about 125 g) that attaches to a smartphone and a dedicated app. RTK (real-time kinematic) is a technology that corrects satellite positioning (GPS/GNSS) errors in real time to improve positional accuracy to within a few centimeters (a few in). When an LRTK device is mounted on a smartphone, the phone’s usual position measurement error of several meters (several ft) is immediately elevated to surveying-equipment-class high accuracy. By combining this high-accuracy position information with the smartphone’s built-in LiDAR sensor and camera, anyone can easily acquire high-precision 3D point clouds. Because all you need is a smartphone and a pocket-sized antenna, the system can be carried on-site at all times and measurements can be taken whenever needed.

When measurements are performed with the dedicated LRTK app, the acquired point cloud data and surveyed point coordinates are synchronized in real time to the cloud service “LRTK Cloud.” On LRTK Cloud you can view point clouds in 3D via a browser and intuitively measure distances, areas, and volumes. For example, if you scan a fill area with your smartphone and upload it to the cloud, the volume is automatically computed and displayed on the spot, eliminating the need to bring data back to a PC for analysis. Cloud-based 3D data can be shared immediately with stakeholders, enabling supervisors and clients in the office to view the same point cloud model in real time during site checks.

The advantages of smartphone surveying with LRTK are that it is easy for anyone to operate and is overwhelmingly low-cost compared to conventional equipment. The app is designed to be intuitive for field staff without specialized training, creating an environment where users can “measure themselves whenever they want.” Tasks that previously required expensive laser scanners or outsourcing to surveying companies can, in many cases, be handled with a single smartphone and LRTK. It aligns with the Ministry of Land, Infrastructure, Transport and Tourism’s i-Construction initiative and can be considered a state-of-the-art on-site DX tool.

Thus, point cloud surveying technology is evolving toward forms that are easy to use on-site. If you feel “I want to introduce this to my site” or “I want to try it first,” starting with simple 3D surveying using a smartphone and LRTK is one option. For more detailed product information and consultation on implementation, please check the [LRTK official site](https://www.lrtk.lefixea.com/). Consider adopting advanced technologies appropriately to improve site productivity and strengthen safety management.