Point Cloud Data Transforms the Field! Double the Efficiency in Civil Engineering with High-Precision 3D

The construction industry faces mounting challenges such as labor shortages due to an aging and shrinking population and harsh working conditions often described as the “three Ks” (kitanai—dirty, kiken—dangerous, kitsui—hard). Improving productivity is an urgent priority. In civil engineering worksites, with efficiency increasingly demanded, the Ministry of Land, Infrastructure, Transport and Tourism has set a goal to improve on-site productivity by 20% by fiscal 2025 through the use of ICT (information and communication technology) and 3D data. A key technology in this context is point cloud data, which digitizes the entire site through 3D scanning. By leveraging high-precision 3D point clouds, there is potential to double productivity across all phases of civil engineering—surveying, design, construction, and maintenance. In practice, adoption of point clouds is progressing across a wide range of sites—from major general contractors to small and medium-sized contractors and local governments—with reported improvements in efficiency and quality. This article explains in detail how using point cloud data enhances efficiency at each process, covering specific technologies, tools, workflows, and case examples.

What is point cloud data? Recording the entire site in three dimensions

Point cloud data are three-dimensional data that represent the shape of objects or terrain by a large number of points in space. Each point includes X, Y, and Z coordinates (and sometimes color information), and the collection of points can reproduce terrain and structures with high accuracy. In other words, it is a “digital, full-scale copy of the site” captured as countless points. Point clouds record complex undulations and structural details in three dimensions exactly as they are—details that conventional planar drawings or surveys with only a few points cannot fully capture.

Types of point cloud data vary significantly depending on acquisition methods. Representative types include point clouds obtained by laser measurement and those generated by photogrammetry, each with distinct characteristics. For example, point clouds acquired by laser scanners directly measure distance and therefore offer high accuracy and allow material inference from return intensity. On the other hand, point clouds generated by photogrammetry are based on color images, so the point cloud is textured and visually intuitive. Both approaches share the ability to digitize the site in fine detail, and they are used according to the objective.

Methods for acquiring point cloud data (drones, laser scanners, smartphone LiDAR)

Point cloud data are mainly acquired with 3D laser scanners (LiDAR). These dedicated devices emit laser light and capture the reflected light with sensors to rapidly measure coordinates of numerous surrounding points. This enables collection of a huge number of points in a short time—quantities that would be difficult to measure manually—producing high-precision 3D data. Laser scanners come in various types, such as drone-mounted units for aerial surveys (UAV LiDAR), tripod-mounted terrestrial scanners (fixed type), and vehicle-mounted systems used while driving (mobile mapping systems, MMS). Wide reclamation sites often use drones, while roads and urban areas use terrestrial fixed units or MMS—choosing the method according to site scale and target. Recently, smartphone LiDAR built into devices like iPads and Android devices has appeared and is being used for surveying narrow indoor spaces and small structures. New methods—such as handheld 3D scanners and backpack-type LiDAR units that can be carried while walking—are also being developed one after another.

Separately, photogrammetry—converting photos into point clouds without using lasers—is also widely used. This method takes photos of the site from various angles using drones or single-lens cameras and uses software to analyze and create 3D models. Advances in SfM (Structure from Motion) technology and increased computing power have made it possible to generate high-precision point clouds from photos. For example, hundreds of drone-acquired photos can produce point clouds on the order of tens of millions of points to generate detailed terrain models. Photogrammetry is relatively low-cost and easy to start with, so it is increasingly combined with laser scanning to leverage the strengths of both methods.

Key points in processing and analyzing point cloud data

Raw point cloud data acquired at the site typically require several post-processing steps before they can be used. First, point clouds often include unnecessary points such as tree leaves, passing vehicles, and people; these noise points must be filtered out. When combining multiple acquisition methods—such as drone and terrestrial laser scans—accurate alignment (registration) of individually acquired point clouds is also important. These processes unify the entire point cloud into a single coordinate system, producing high-accuracy “as-built 3D data.”

Moreover, raw point clouds are collections of countless points, and in that state they can be difficult to directly compare with design data expressed as surfaces and lines like CAD drawings or BIM models. Therefore, as needed, meshes or surfaces are generated from point clouds to extract terrain surfaces or structural contours as polygons or NURBS surfaces. For analyses such as volume calculations or cross-section generation, subsets of the point cloud may be extracted or converted into gridded elevation data (DEM). Recently, software and cloud services for point cloud processing and analysis have become more robust, with tools that automatically remove noise and perform comparative analyses without specialized expertise. These processing advancements make it possible to smoothly integrate site-acquired point clouds into design and construction management.

Benefits of using point cloud data: dramatically improve productivity and quality

Introducing 3D point clouds brings various benefits to civil engineering site operations. Below are the main effects.

• Efficiency through labor savings and time reduction: Point cloud surveys significantly reduce work time and effort. For example, site surveying that used to take two days can, in some cases, be completed in about half a day with a one-shot 3D survey using a drone. Replacing manual, point-by-point measurements with comprehensive scanning can dramatically shorten processes from surveying to drawing creation and quantity calculation. In addition, as-built inspections can shift from manual checks with rulers and gauges to automated comparisons using point cloud data, reducing the days required for inspection. With limited on-site personnel, more tasks can be accomplished.

• Improved accuracy and quality: Point clouds from laser scanning or photogrammetry are very dense and capture fine site undulations. Errors that would be missed by sparse point surveys can be fully captured by point clouds. Reports indicate that with appropriate control points, drone photogrammetry can achieve accuracy of a few centimeters or less (a few cm or less, a few in or less), meaning even simple methods can secure sufficient precision. This contributes to strict verification of as-built conditions and improved accuracy in quantity calculations, reducing rework and raising construction quality.

• Recordkeeping and data assetization: Point clouds can store a full digital record of the site at a given time, becoming an asset that can be reviewed or reused later. If pre-construction terrain point clouds are preserved, they can be compared in detail with post-construction conditions. For example, in the 2021 debris flow disaster in Atami City, Shizuoka Prefecture, differences between pre- and post-event point clouds were used to quickly calculate the range and volume of collapsed sediment, aiding damage assessment. Point clouds captured at project completion are also useful for future maintenance, enabling monitoring of long-term changes. The ability to accumulate a “comprehensive site history” that cannot be preserved with paper drawings or photos is a major advantage.

• Safety and work-style reform: Surveying at dangerous heights or steep slopes can be made safe by replacing human entry with drones or remote measurement. Data can be collected non-contact in areas where people cannot safely enter, reducing worker risk. Labor-intensive as-built measurements can also be automated, reducing physical burden and overtime. Even veteran technicians unfamiliar with ICT are increasingly able to use these tools with the support of younger staff, making them accessible regardless of age or experience. Utilizing point cloud technology contributes to improving working conditions and the image of the construction industry, which can positively affect recruitment and retention of workers.

Use in as-built surveys: quickly and accurately grasp large areas

Point cloud data proves powerful in the initial civil engineering step of as-built terrain surveys (surveying). Traditionally, surveyors used total stations and levels to painstakingly measure many elevation and distance points. This approach requires extensive time and manpower to survey large sites and cannot capture fine undulations between measurement points. Introducing point cloud surveys using drone aerial photography or terrestrial laser scanning allows the entire site to be measured “as an area” in a short time. For example, a single drone flight can acquire surface data numbering in the millions of points in about 30 minutes to one hour. The survey results can be visualized immediately as a 3D model, and back in the office analysts can measure elevations and cross-sectional shapes at any point. Surveys that once took two days can now be finished in less than half a day, delivering dramatic efficiency gains directly tied to reduced survey costs and shorter construction schedules.

High-precision as-built point clouds also improve the accuracy of the design phase. By conducting design reviews on 3D models that reflect terrain details, planners can detect planning issues (such as clashes or insufficient retaining wall height) that might be overlooked on planar drawings. Remote measurement of cliffs, rivers, and aging infrastructure that are difficult to access in person also enhances the safety of preliminary surveys. In this way, point cloud-based as-built surveys form a foundation that raises the efficiency and accuracy of the entire project, including subsequent design and construction.

Use in the design phase: improve planning accuracy with 3D as-built models

Point clouds from surveying are also highly useful in the design phase. Traditionally, designers planned based on 2D drawings and a limited number of survey points, imagining the site conditions. By using detailed 3D terrain models generated from point clouds, designers can recreate the site in virtual space and perform design reviews. For road design, planning alignments and vertical profiles along terrain derived from point clouds enables accurate calculation of cut-and-fill volumes. For reconstruction of tunnels or bridges, point clouds of existing structures allow designers to derive interference-free design dimensions. Precisely grasping the as-built conditions during the design phase prevents mistakes like “built to the drawing but didn’t fit the site,” reducing the risk of rework during construction.

Design reviews using point clouds as a background also aid stakeholder consensus building. 3D proposals are easier for owners and contractors to intuitively understand, shortening the time required to reach agreement. In some cases, VR or AR is used to overlay design models on point clouds and share the completed image. For example, using AR glasses on-site to overlay a future structure model on the as-built point cloud can facilitate smoother meetings with owners. Incorporating 3D point clouds into design dramatically enhances planning accuracy and communication efficiency.

Design verification during construction: catching construction errors with point clouds

During construction, measuring completed structures or partial earthworks and comparing point clouds to design data strengthens quality control. Overlaying point clouds of completed portions with the design 3D model (BIM data or 3D-converted design drawings) allows detailed verification of construction accuracy. For instance, acquiring a point cloud after concrete placement and comparing it with the design model can reveal slight dimensional discrepancies or missed elements at a glance. A major general contractor used this method to detect dimensional errors earlier than the traditional final inspection, enabling early correction and significantly reducing rework. Incorporating point cloud measurements and design comparisons during construction enables early detection and correction of mistakes, resulting in shorter schedules and lower costs.

Using point clouds for as-built inspections also allows comprehensive checks rather than spot inspections. Traditional as-built management often relied on verifying elevation at a few specified points, but with point clouds the entire structure’s shape can be thoroughly examined. For contractors, objective 3D data serve as evidence, smoothing on-site inspections with owners.

As-built evaluation with heat maps: visualizing the finish with color

When comparing point clouds to design data, using heat maps to visually display deviations is effective. A heat map colors points according to deviation from the design surface, intuitively showing finish precision. For example, by setting colors such that areas within design tolerance are blue or green and areas with excessive fill or over-excavation beyond thresholds are red, one can instantly identify “too high” or “too low” spots. With heat maps, site agents and inspectors can grasp the finishing status of the entire space without comparing to drawings.

Heat maps are easy to create in software and can be taken to the site on tablets for inspection. Recently, advanced approaches have appeared that project heat maps onto actual structure surfaces using AR to check as-built conditions. Color-based visualization is also useful as explanatory material for owners, being easier to understand than traditional numerical tables and enabling faster judgments on acceptance or rework. Introducing heat map evaluations advances as-built management into a more reliable and efficient process.

Earthwork volume calculation using point clouds: accurately compute fill and excavation quantities

In civil engineering, earthwork volume calculations—how much soil to fill or excavate—determine project cost and schedule. Point clouds dramatically streamline and improve the accuracy of these calculations. Traditionally, a method involved extracting several cross-sections from design drawings and estimating total volume from those sections. Using point clouds, the volume difference between the as-built terrain and the design terrain can be integrated on a computer to obtain the total earthwork volume in a single operation.

For example, at one site a point cloud acquired by drone photogrammetry in about 30 minutes was used to calculate fill volumes on the spot and revise construction plans the same day. Previously, volume calculations were taken back to the office and required hours to days, but point cloud utilization made immediate on-site calculations possible. This immediacy allows quick decisions such as changing construction approach or arranging the number of dump truck trips earlier. In as-built inspections, accurate measurement of completed fill or excavation volumes from point clouds clarifies surpluses or shortages and smooths settlement with the owner. Point cloud-based volume calculations are a powerful tool for both cost control and schedule management.

Use for progress management: visualizing construction with 3D data

Point cloud data are also used for construction progress management. Regular site scans accumulate 3D data of construction status at specific times. Comparing these with the project schedule makes it clear which areas are on schedule and which are delayed. For example, weekly drone flights and point cloud generation can visualize terrain and structure progress over time, integrating quantity (as-built) management with schedule management. Spatial progress that was difficult to understand with paper schedules or photos can be easily visualized using point clouds.

Furthermore, shared point cloud data among stakeholders enables remote monitoring. If site-acquired 3D data are shared via the cloud with headquarters or the owner, remote parties can check site details. In one tunnel project, point clouds acquired by robots and drones were transmitted via satellite communications to headquarters in Tokyo for real-time construction management in a demonstration. Using point clouds this way, headquarters managers and clients can check progress and quality without visiting the site, accelerating reporting and approval processes. This reduces the burden on site agents and enables faster responses when anomalies occur.

Cloud-based point cloud sharing: check 3D data from remote locations

Cloud platforms are further promoting effective use of point cloud data. Traditionally, handling large point cloud datasets required high-performance PCs and dedicated software, but cloud services that allow viewing and sharing point clouds on the web have become widespread. With these services, large point cloud files acquired at the site can be uploaded to a server and all stakeholders can view the same 3D data via a browser. For example, owners, design consultants, and partner companies can discuss while viewing point clouds in real time, enabling efficient communication with shared spatial understanding.

Cloud-based point cloud management offers other benefits. Centralized, always-up-to-date data prevent mistakes like “working from outdated drawings.” Because point clouds can be accessed from tablets or general laptops without dependence on high-end PC environments, access from site offices or field locations is easy. Some services provide collaboration features such as attaching comments or sketches to point clouds, and examples of using point clouds like drawings in remote meetings are increasing. Cloud utilization is transforming point clouds from mere survey deliverables into an information-sharing infrastructure for the site.

Applications in maintenance: infrastructure inspection and digital archives

Point clouds are becoming indispensable in the maintenance phase after completion. If point clouds of structures are saved as a digital ledger at completion, re-scanning the same location years or decades later allows comparison of changes. In tunnel and bridge condition surveys, acquiring point clouds at regular inspections and overlaying them with previous data enables quantitative detection of crack progression and changes in cross-sectional shape. Point clouds can precisely capture ground settlement and subtle road deformations, aiding early repair planning. Digital difference analyses make it easier to detect abnormal signs that might have been missed by traditional visual inspections or partial measurements.

Local governments are also assetizing point cloud data. For example, Shizuoka Prefecture conducted aerial LiDAR for the entire prefecture under the “Virtual Shizuoka” project, acquiring and publishing 15 TB of high-precision 3D point cloud data. This data is available as open data and is used for disaster prevention, infrastructure inspection, urban planning, and even tourism VR. In the aforementioned Atami debris flow incident, this base data greatly contributed to understanding the affected area. Tokyo has also developed detailed point clouds for the 23 wards and released them in 2024. That administrations are positioning 3D data as the basis of a city digital twin indicates that point clouds will play an important role in maintenance.

Simple surveying with smartphones + GNSS: LRTK supports the field

Even with the clear utility of point clouds, some sites may worry that expensive equipment or specialized skills are required. Indeed, high-end 3D laser scanners once cost millions of yen, but recently methods for obtaining high-precision point clouds using simple surveying with smartphones combined with GNSS have emerged. A representative example is the LRTK series. LRTK consists of a small high-precision GNSS receiver that attaches to a smartphone and a dedicated app, providing a surveying solution that enables anyone to perform cm-class positioning (cm level accuracy, half-inch accuracy) easily. By utilizing smartphone-built-in LiDAR or cameras and simply walking around, users can acquire high-precision 3D point clouds with absolute coordinates, allowing accurate digitalization of site conditions without special training or large equipment.

This smartphone surveying system has the major advantage that surveys and point cloud scans can be carried out immediately when needed on site. For example, when sudden design changes or as-built confirmations arise, site personnel can instantly acquire point clouds and assess conditions without dispatching specialized survey teams. Real-time GNSS corrections provide consistently high-precision positioning, so acquired point clouds can be immediately compared with design drawings or existing survey coordinates. Tools like LRTK are compatible with the Ministry of Land, Infrastructure, Transport and Tourism’s i-Construction initiatives, and the acquired point clouds have sufficient accuracy to be used as as-built management deliverables. Affordable equipment makes adoption easier, and use is spreading among small and medium-sized enterprises and local governments, making it a trump card for dramatically improving on-site work efficiency and surveying accuracy. For more details, please refer to the [LRTK official site](https://www.lrtk.lefixea.com/).