Drone × SfM Processing Opening the Way for Construction DX: Latest Cases of Rapid Point Cloud Generation

The construction industry urgently needs "Construction DX (digital transformation)" to improve productivity and reduce manpower. A key technology attracting attention is the combination of drone aerial photography and SfM processing (Structure from Motion) to rapidly generate 3D point cloud data. This article explains how SfM works, the advantages of using drones for large-area, high-speed surveying, implementation examples of autonomous flight, RTK, and the latest LRTK technologies, and use cases for as-built management and disaster recovery. We also introduce difference detection using point cloud data (heatmap visualization), the benefits of GCP simplification and smartphone integration with LRTK, and conclude with the prospects for operational efficiency and labor reduction brought by adopting these technologies.

1. What is SfM? Photogrammetry mechanics and 3D point cloud generation technology

SfM (Structure from Motion) processing is a photogrammetry technique that reconstructs the three-dimensional structure of an object or site from multiple photographs. Using computer vision algorithms, it detects feature points in overlapping areas of multiple photos, matches these common points, and automatically calculates the camera positions and orientations and the 3D coordinates of each point. As a result, the shooting location and pose for each photo are estimated, and a point cloud dataset (3D point cloud) representing the shape of the object is generated. Because computers automate the aerial photogrammetry steps that previously required craft skill, a major advantage is that high-density 3D models can be built efficiently from many photographs.

The generated point cloud represents the surface of objects or terrain with innumerable points, each carrying X, Y, Z coordinate information (and color information based on the photos). The site can be reproduced almost like a realistic 3D photograph, and it is easy to perform analyses such as measuring distances and volumes from the acquired point cloud or creating cross-sections. In addition, point clouds obtained from photos can be used to generate textured 3D models, which can be overlaid with design data or used for VR visualization. SfM processing can be automated by dedicated software (e.g., Metashape or Pix4D), so as long as images are captured with a camera, anyone can now realize high-precision 3D surveying.

2. Why drone (UAV) aerial photography + SfM enables fast, wide-area point cloud acquisition

The combination of drone (UAV) aerial photography and SfM processing is a powerful method for acquiring wide-area 3D point clouds in a short time. Shooting from the air allows a wide area to be covered at once, making it far more efficient than ground-based surveying on foot. For example, where traditional surveyors using total stations might spend days measuring a large earthwork site, a drone can capture hundreds of photos in just tens of minutes to half a day of flight and obtain a detailed point cloud model with centimeter-level accuracy. In fact, the Ministry of Land, Infrastructure, Transport and Tourism reported a case where terrain surveying of about 0.3 square kilometers, which previously took 45 days, was reduced to 1.5 days using UAV laser surveying. Similarly, photogrammetry dramatically shortens on-site work time compared to manual point measurements.

Another reason is that high-density photo data from drones and advances in SfM software have sped up data processing (office work) as well. By parallel processing on high-performance PCs or cloud services, point cloud generation from hundreds to thousands of photos can be completed quickly. This enables a rapid cycle from surveying → point cloud generation → analysis, making it realistic to scan sites weekly or even daily to monitor progress. Drones also allow safe overhead inspection of steep slopes or hazardous post-disaster areas where people cannot enter. Drone + SfM, which can digitize wide areas with high accuracy, is increasingly recognized as a foundational technology for digital twins of construction sites.

3. Implementation examples: autonomous flight, RTK correction, and LRTK-equipped drones

Various technical measures have been implemented in recent years to efficiently obtain accurate point clouds with drones. First, autonomous flight (autonomous navigation) enables stable aerial photography by anyone. With dedicated flight planning software, you specify the area to survey and the drone will fly a regular route and capture photos at an optimal overlap (a guideline is forward overlap of 80%+ and side overlap of 60%+). This provides uniform, high-quality photographic data, stabilizing the accuracy of point cloud generation by SfM. During autonomous flight, AI-based obstacle detection and altitude-hold functions operate so that even non-experts can safely and reliably capture wide-area imagery.

Next, high-precision positioning via RTK correction is indispensable. RTK (Real-Time Kinematic) is a technique that augments satellite positioning in real time, allowing GNSS receivers mounted on drones to obtain centimeter-level position information. Normally, the positions of drone photos are optimized later in software, but with RTK each photo’s geographic coordinates are recorded with high accuracy, dramatically improving overall model alignment accuracy. Specifically, single-receiver positioning errors of several meters can be reduced to a few centimeters with RTK, meaning it is possible to obtain accurate 3D models without having to set up many ground control points (GCPs) in advance. RTK corrections can be received either via a base station placed on site communicating by radio, or by connecting to network RTK (VRS) services provided by providers such as telecom operators or the Geospatial Information Authority. With the latter, correction information is obtained via the internet, so RTK positioning can be performed without dedicated radios or base stations.

More recently, LRTK-equipped drones have appeared. LRTK (Lightweight RTK) is a compact, integrated RTK-GNSS device and service developed by a startup originating from Tokyo Institute of Technology. While originally designed to be attached to smartphones, its compactness and versatility have raised expectations for drone surveying applications. By mounting a palm-sized, few-hundred-gram receiver on the aircraft and connecting to the onboard computer via Bluetooth or Wi-Fi, simple high-precision positioning can be achieved. If RTK, which used to require specialized and heavy surveying instruments, becomes usable on small drones, a highly mobile and cost-effective surveying system can be built. There are already trials where LRTK modules are retrofitted to non-RTK consumer drones to enable centimeter-level aerial photogrammetry on inexpensive platforms. The spread of autonomous flight + RTK as "precision autonomous drone surveying" is expected to make wide-area point cloud acquisition easier and more precise.

4. Use cases and outcomes of SfM in as-built management, progress management, disaster recovery, and terrain surveying

Point cloud technology from drone aerial photography and SfM is being applied across many civil engineering and construction sites. Here are representative use cases and observed outcomes.

• As-built management: Point clouds are powerful for verifying whether finished work matches design. For road construction, for example, the completed pavement and slopes are turned into point clouds via drone photogrammetry and overlaid with design 3D models or drawings for comparison. Because thickness and height differences can be analyzed across surfaces in the point cloud, what used to be judged from a few survey points can now be evaluated across entire areas, improving quality control accuracy.

• Progress management: SfM point clouds are also useful for tracking construction progress. By updating point clouds with regular drone flights (e.g., weekly or monthly), volume changes from embankment or excavation can be quantified. For example, point cloud comparisons can calculate how much soil was removed or brought in since the last survey, assisting objective assessment of work volume and schedule control. Scanning concrete immediately after placement to compare with design models to confirm required thickness and slope is another example. Progress management using point clouds advances site visualization and automates recordkeeping, streamlining report generation.

• Disaster recovery: Drone + SfM is effective for assessing landslides and debris flow disaster sites. Since detailed 3D surveys can be conducted from the air even in hazardous zones, collapsed soil volumes and affected areas can be rapidly mapped. In one earthquake-affected site, drone imagery was captured immediately after the event to create a point cloud model that quickly estimated the fallen soil volume, aiding emergency recovery planning. Moreover, by using LRTK as mentioned later, workers could measure and photograph precise locations with a smartphone and share them to the cloud, enabling remote headquarters to grasp the situation and make decisions instantly. Situational assessment that used to take days can in some cases be completed within hours, contributing to faster disaster response.

• Terrain surveying: SfM point clouds are used for topographic surveys needed in road and earthwork planning. Tasks that once required survey teams to traverse terrain and measure many points can now produce a high-density terrain model in a short time via drone imagery. Even in rugged topography, contours and longitudinal/transverse sections can be freely created from point clouds, allowing planners to incorporate detailed terrain information into designs. For projects like forest road maintenance or erosion control, acquiring up-to-date terrain point clouds in advance streamlines route selection and earthwork quantity estimates. Based on such outcomes, the Ministry of Land, Infrastructure, Transport and Tourism’s "i-Construction" initiative recommends drone photogrammetry, and many cases have reported shortened survey periods and cost reductions.

5. Difference detection (design deviations, work quantity, volume) and heatmap visualization using point cloud comparisons

Point clouds obtained by SfM can be compared with design models or past point cloud datasets to extract various difference information. This enables quantitative evaluation of site progress and quality and intuitive identification of issues. Typical difference detection and visualization uses are listed below.

• Detecting deviations from design models: Overlaying the completed point cloud with the design 3D model (or design terrain data) and computing deviations at each location allows as-built management to visualize these deviations as color maps (heatmaps) so that areas that are over- or under-built relative to the design can be seen at a glance. For example, in ground preparation work, generating a map that shows areas more than 5 cm above the design in red and areas below it in blue helps quickly identify spots that need rework. What used to be inferred from only a few measured points can now be detected across the entire surface with point cloud comparisons.

• Calculating work quantity and soil volumes: Comparing previous and current point clouds allows accurate calculation of soil volume changes due to construction. For instance, in levee works, one can verify whether the fill volume within a section matches the design by examining differences between before-and-after terrain point clouds. Automatically calculating removed or placed soil from point cloud differences gives credibility to work quantity management. Combining this with heatmap displays shows where and how much soil has increased or decreased, making such visuals valuable for progress meetings.

• Monitoring deformations and displacements: Comparing multiple point clouds acquired over time can capture subtle displacements of structures or terrain. For example, in routine bridge inspections, overlaying last year’s point cloud with the current one and showing differences in color can detect pier settlement or girder deflection changes at millimeter precision. Age-related changes that are hard to notice visually can be quantitatively understood through digital point cloud comparison. This technology contributes to advanced infrastructure maintenance and reduces inspection labor.

The results of these difference analyses can be exported as numerical reports in software or shared as colored heatmap diagrams. Visualization through point cloud comparison allows construction managers to quickly detect deviations from design or delays in progress and take corrective action at the right time. In short, point cloud data serves as a mirror reflecting the site’s "truth," driving site DX from intuition-based decisions to data-driven management.

6. Field benefits of correction observation with LRTK: GCP simplification and smartphone integration (complementing ground occlusions)

While RTK is important for efficiently obtaining high-precision point clouds, the recent emergence of LRTK (lightweight RTK solutions) is becoming a game-changer. LRTK consists of a small, lightweight RTK-GNSS receiver, a smartphone app, and cloud services, enabling centimeter-level positioning easily for anyone. By attaching the dedicated receiver to a smartphone and connecting the app to a network RTK correction service, even a single operator can start high-precision positioning immediately. The receiver is dramatically more compact than traditional RTK survey equipment (about 10 cm in diameter and a few hundred grams), so it can be carried on a survey pole or mounted on a helmet to keep hands free—making on-site handling extremely convenient. No complicated device operation is required, and positioning accuracy and satellite status can be checked at a glance on the smartphone screen, so technicians without specialized surveying knowledge can master it with short training.

Introducing LRTK greatly simplifies the installation work of GCPs (ground control points) needed for photogrammetry. Even when using an RTK-equipped drone for aerial photography, several control points are usually installed on site for accuracy verification or as a backup. With LRTK, a single person can efficiently measure those coordinates. Walking the site with a smartphone, tapping a button at the center of a marker or at a reference point instantly acquires high-precision coordinates and records them to the cloud. This converts a task that previously took two or more people half a day into a quick single-person job, enabling immediate transition to aerial photography. Furthermore, even when performing photogrammetry with small drones that do not have onboard RTK, having a few GCPs measured with LRTK ensures accuracy, making inexpensive platforms practically useful.

In addition, LRTK combined with smartphone cameras is powerful for complementing ground occlusions. Areas that drone aerial photography cannot capture—such as the undersides of bridges, backsides of structures, or under dense tree canopies—require ground photos to be added to SfM processing. If those ground photos are taken with a smartphone (or tablet) equipped with LRTK, accurate position coordinates are attached to each ground image, making integration with aerial photos smooth. For example, images of a bridge underside or complex piping around a plant facility taken from the ground with LRTK-derived position data can be easily aligned with other aerial images in SfM software. As a result, a full-coverage 3D model including occluded areas that drones cannot capture can be constructed. On site, supplementary photos can be shared to the cloud immediately and combined and analyzed centrally, enabling near real-time collaboration between field and headquarters. LRTK thus makes data sharing between site and office seamless and speeds up decision-making.

In this way, LRTK supports site DX as a drone-friendly "easy high-precision positioning tool." Domestic construction majors and surveying firms have already started pilot introductions, and the concept of a "GPS surveying device for every person" is expected to spread.

7. Conclusion: Operational efficiency and labor reduction brought by LRTK adoption

We have reviewed the latest cases and technology trends in point cloud generation using Drone × SfM, and the key technologies are RTK for real-time high-precision positioning and the new generation of LRTK solutions. Utilizing these technologies enables dramatic efficiency improvements from surveying to construction management and makes on-site labor reduction increasingly feasible. Autonomous drone flight enables anyone to acquire data, SfM processing automatically generates massive 3D information, and LRTK allows individuals to perform precise surveying even on sites lacking specialized personnel—digital technology is transforming construction sites and powerfully driving Construction DX.

Going forward, the spread of smart positioning technologies such as LRTK will usher in an era where site and office are directly connected by data. With point cloud data shared in the cloud, remote offices can grasp conditions and issue instructions, and new workflows such as AI-driven automated as-built checking will become possible. This means that even with limited personnel, many sites can be managed safely and accurately, which will be a major boon to the labor-short construction industry.

The effects of adopting LRTK go beyond improved surveying accuracy. They include shortened lead times for data acquisition and processing, reduced heavy labor, simplified reporting, and—most importantly—faster decision-making by closing the information gap between site and office. These cascading efficiencies naturally lead to on-site labor reduction and contribute to workstyle reform and improved safety.

Finally, the combination of Drone × SfM × LRTK is a powerful solution that realizes Construction DX. Establishing a system that can rapidly acquire and utilize high-precision 3D point clouds will surely take site operations to the next stage. If your site has not yet adopted these technologies, don’t miss the wave—experience the power of smart surveying. A step toward new technology can greatly enhance tomorrow’s site productivity and competitiveness.