What is Point Cloud Data?

​

"Point cloud data" refers to a collection of numerous points in three-dimensional space, with each point having XYZ coordinates (and, in some cases, attributes such as color or reflectivity). By using technologies like 3D laser scanning or photogrammetry, real-world structures and terrain can be scanned and digitally recorded as this collection of points. Point cloud data is highly valuable as it can precisely replicate the shapes of structures and terrain, and its use is expanding across various fields, including civil engineering and construction.

​

The point cloud data with absolute coordinates is especially important. Absolute coordinates refer to coordinates based on unified standards, such as the global geodetic system, and are essential when comparing point clouds with map coordinates, existing drawings, or past and future data. For example, in bridge inspections, by managing the structural point cloud model with the same reference coordinates each time, even small changes in displacement or tilt can be detected. Point cloud data linked to absolute coordinates allows for the quantitative capture of structural movement and deformation, and is expected to be useful for long-term maintenance and management.

​

Using Point Cloud Data in Bridge and Tunnel Inspections

​

Traditional bridge and tunnel inspections involved technicians using elevated work platforms or scaffolding to visually inspect deterioration up close, manually measuring crack widths with rulers. These methods were time-consuming, dangerous, and limited in the areas that could be measured. The use of point cloud data in inspections is a significant departure from these traditional methods. By scanning the entire structure with drone-mounted cameras or ground-based LiDAR, high-density 3D point clouds and images can be obtained, digitally recording every detail of the site. The acquired point cloud data can be analyzed in the office, reducing the number of on-site visits while enabling a more detailed understanding of the condition.

​

One specific application is crack detection. By combining point cloud data with high-resolution photos, technicians can accurately identify where cracks or concrete delamination are located on the bridge and measure their size in real-world dimensions. With the use of the latest image recognition AI, photographs can be automatically mapped to corresponding locations on the point cloud, and the dimensions of cracks or delaminations can be measured on the 3D model. This helps prevent overlooking deterioration and accurately identifies repair locations.

​

Next, deformation measurement is another key application. For example, in tunnels, traditional methods involved measuring only pre-designated points at regular intervals to estimate internal displacement (such as cross-sectional contraction). However, by scanning the entire tunnel circumference with a 3D laser scanner, it is possible to understand which parts of the cross-section have deformed and by how much. By analyzing the acquired 3D point cloud data, deformation in the tunnel can be calculated at any cross-section, detecting even minor distortions that traditional methods would have missed. Point cloud data analysis for displacement measurement thus provides a powerful means of quantitatively assessing the structural health of the facility.

​

Additionally, applications in durability assessments are advancing. For concrete bridges, precise analysis of surface irregularities from point cloud data can detect deterioration phenomena such as delamination or exposed rebar. In fact, attempts to extract fine surface defects from LiDAR point clouds and create deterioration distribution maps have been reported. By comparing point cloud data over time, the degree of deterioration can be quantified, helping in the evaluation of a structure's durability and the progression of deterioration.

Examples of Efficiency Improvements Using Point Cloud Data

Example 1: Bridge Deterioration Diagnosis
In the case of a road bridge, drones were used to capture and create point clouds of the entire bridge, aiding in the diagnosis of deteriorated areas. Since drones can easily capture high and narrow areas, inspections of parts that were previously difficult to access, such as the bearing sections, could be safely conducted. A digital twin (3D model) of the bridge was created from the acquired point cloud data and images, allowing for detailed assessment of the location and extent of cracks and delamination. Inspectors used this model to manage deteriorated areas and prioritize repairs. As a result, the use of work vehicles was reduced, and high-altitude work was minimized, significantly easing the on-site workload. In fact, the introduction of autonomous drones resulted in reports of a 30-40% reduction in labor and efficiency compared to traditional methods. This example demonstrates how point cloud data-enabled precision diagnosis provides both efficiency and reliability.

Example 2: Tunnel Deformation Measurement
In tunnel inspections, 3D laser scanning for deformation measurement is gradually replacing traditional methods. Previously, only limited measurement points could be used to assess the changes in tunnel diameter, making it difficult to capture the overall picture. In one case, a vehicle-mounted laser scanner was driven through a tunnel, scanning the tunnel walls while moving. As a result, a 20-meter section was measured in approximately 6 minutes, and the point cloud data was displayed immediately. The point cloud captured the tunnel's internal surface with an accuracy of about 5 millimeters, revealing high-density deformation distribution that could not be obtained with traditional total station measurements. By using point cloud measurements, even the smallest deformations such as subsidence or contraction in tunnels can be detected, and measurement time is significantly reduced, making tunnel health monitoring more efficient and advanced.

Example 3: Application in Long-Term Maintenance Planning
Point cloud data is also powerful for long-term maintenance planning of structures. For example, by accumulating point cloud data during periodic bridge inspections, the progression of deterioration and deformation can be compared over time. In one municipality, a virtual 3D model (digital twin) was created from point cloud data acquired during the inspection of small bridges, and this model was shared among stakeholders for repair planning. On-site verification was conducted only during the inspection, and subsequent work, such as marking deteriorated areas and reviewing repair methods, was done on the digital twin, which streamlined the entire maintenance cycle. Since the 3D model accurately records the current condition of the bridge, it is also useful as an archive for reviewing the past condition during future renovations. In this way, continuous use of point cloud data supports long-term preventive maintenance planning and helps maintain infrastructure effectively within limited budgets.

​

Benefits and Challenges of Introducing Point Cloud Data

Benefits:
The use of point cloud data in bridge and tunnel inspections offers various advantages. First, there's a significant reduction in time. By using drones and laser scanners, the on-site inspection time can be dramatically shortened. In fact, there are cases where on-site inspections that typically took 4 to 6 hours using aerial work platforms were completed in 1 to 2 hours with drone inspections. Next, cost reduction is also a major benefit. With fewer personnel and less equipment operating time required for inspections, labor costs and equipment rental costs are reduced. Additionally, completing the survey in a short time leads to shorter road closures, reducing societal costs.

​

An important benefit is improved safety. Traditionally, it was dangerous for people to access areas such as the underside of bridge girders or tunnel ceilings. With remote measurements, there is no need for personnel to physically enter these areas. For example, combining autonomous drones with point cloud technology has allowed inspectors to check for deterioration in confined spaces beneath bridges without having to physically enter them, significantly improving safety. Furthermore, the objectivity and comprehensiveness of the data are also advantages. Since point cloud data records the entire structure without missing any details, even minor damages that might be overlooked with traditional methods are captured as digital data. This allows for flexible responses, such as reanalyzing the data or consulting with experts, which increases the reliability and reproducibility of inspection results.

​

On the other hand, there are several challenges that need to be recognized when implementing point cloud data. One of the biggest challenges is the volume of data and processing. High-density point cloud data results in very large file sizes, making storage and sharing difficult. For example, scanning a single span of a bridge in detail can produce millions of points, and the file size can reach several gigabytes. As a result, data transfer can take a long time, and uploading to the cloud may face communication bottlenecks.

Another challenge is the need for specialized software and hardware. Visualizing and analyzing point cloud data often requires high-performance workstations and expensive software, which can be a barrier to utilization for municipalities and small to medium-sized businesses. Although open-source point cloud processing software is emerging, there are still many cases where specialized knowledge is needed to handle the data.

Additionally, accuracy management is a critical issue. To obtain point clouds with absolute coordinates, GNSS positioning or alignment with known reference points is required. However, equipment calibration errors or positioning inaccuracies can cause misalignment between point clouds and drawings. Ensuring high-precision positioning (using RTK-GNSS or established reference points) and performing sufficient overlapping scans for post-processing adjustments (registration) are necessary to maintain quality management across the entire data acquisition and processing workflow.

​

Point Cloud Data Acquisition and Utilization with LRTK

​

In recent years, the LRTK series has gained attention as a solution that makes the use of point cloud data more accessible. LRTK (by Lefixea Inc.) offers devices and cloud services that allow anyone to easily acquire high-precision point clouds with absolute coordinates. For example, the LRTK Phone is a pocket-sized GNSS-RTK receiver that attaches to a smartphone, allowing easy point cloud measurements when combined with the built-in LiDAR of iPhone or Android devices.

By using RTK-GNSS, it provides centimeter-level positioning while the smartphone’s 3D scanning function captures the point cloud data. This process enables automatic assignment of coordinates to the 3D point cloud without the need for complicated target placement. Compared to traditional methods where markers must be set up, surveyed, and then aligned with the point cloud afterward, this revolutionary method drastically reduces preparation time before measurements.

​

On the other hand, for more extensive and high-resolution scanning, the LRTK LiDAR device can be used. The LRTK LiDAR is a device that combines GNSS-RTK technology with a high-performance 3D scanner, enabling precise and rapid point cloud acquisition from up to 200 meters away. Even for large structures like bridges, this single device can scan from a distance and is capable of surveying vast areas and measuring complex terrains and structures.

The acquired point cloud data is vast, but with LRTK, the smartphone can display up to 15 million points in just a few minutes, allowing for immediate checks on-site for missing or distorted data. If any data is missing, re-scanning can be done right away, ensuring that anyone can easily obtain high-precision data without errors.

​

Additionally, LRTK offers robust cloud analysis and sharing features. The scanned point cloud data is automatically uploaded to the cloud and centrally managed. Even without specialized software, point cloud data can be viewed via web browsers, and coordinate verification, distance and slope measurements, and other functions can be completed in the cloud. This allows teams in remote offices to discuss and collaborate on the same 3D data smoothly. Civil engineering-specific analysis functions, such as volume calculations and as-built management, are also available in the cloud, enabling the immediate use of on-site scanned data for civil design and maintenance. In short, by implementing LRTK, anyone can easily acquire and utilize high-precision point clouds with location data.

​

As LRTK-like methods become more widespread, the inspection and maintenance of bridges and tunnels will become even more efficient. The easy acquisition of detailed point cloud data with absolute coordinates will accelerate the creation of digital twins in the infrastructure sector, and integration with AI for automatic damage detection and deterioration forecasting is also expected. The use of point cloud data is truly a key technology for civil engineering DX (digital transformation). As infrastructure ages, utilizing point cloud data and its acquisition technology, such as LRTK, will help achieve safe, reliable, and labor-saving inspections. We highly encourage considering the use of point cloud data and LRTK technology for this purpose.