Utilization of Point Cloud Scanning for As-Built Records

Table of Contents

• Introduction

• What is point cloud scanning? Overview and acquisition methods

• Characteristics of point cloud scanning and its use in as-built records

• Comparison with conventional methods: advantages of point cloud scanning

• Procedures and points to note when using point cloud scanning

• Field use cases - Use in disaster recovery - Use in as-built / conformity management - Use in periodic infrastructure inspection

• Recommendation for simplified surveying with LRTK

• FAQ

Introduction

As-built records before and after construction or during disasters are important in civil engineering and municipal surveying work. Recently, three-dimensional measurement using point cloud scanning has attracted attention as a recording method. Because point cloud scanning can digitally preserve site conditions in detail, it enables more accurate and comprehensive records compared to conventional photographs or manual surveying. This article explains point cloud scanning—from the basics to how to use it for as-built records—in a way that is easy for beginners to understand. We introduce the benefits compared to traditional methods, practical implementation steps, points to watch for, and concrete field examples. Finally, we touch on simplified surveying using the latest solution, LRTK, to provide tips for introducing point cloud scanning.

What is point cloud scanning? Overview and acquisition methods

Point cloud scanning is a measurement technique that acquires the shape of objects or terrain as a multitude of points and reconstructs them in three-dimensional space. Each point contains X, Y, and Z coordinate values (and sometimes color or intensity information), and this vast collection of points is called point cloud data. In other words, point cloud data is a high-precision digital copy of the real world. Because it can survey wide areas at once and record shapes in detail down to millimeter- to centimeter-level resolution, it can digitally preserve the three-dimensional “true as-built” conditions that drawings and photos cannot capture.

The main methods for acquiring point cloud data include the following:

• 3D laser scanners (terrestrial LiDAR): A laser scanner mounted on a tripod emits laser light at high speed and measures the distance to objects from the time difference of the returned reflection. It can acquire millions of coordinates per second, enabling high-density surveying of buildings and terrain in a short time. Laser scanners are widely used in civil surveying today and can acquire data with millimeter-level accuracy.

• Photogrammetry: This method uses a drone or DSLR camera to photograph the target from various angles and specialized software to analyze the images and convert them into a point cloud. Feature points appearing in multiple photos are matched and 3D coordinates are determined by triangulation. Point clouds generated from photos include color information, resulting in 3D models that closely resemble the real scene. Aerial photogrammetry with drones is effective for grasping wide-area terrain and calculating earthwork volumes, allowing rapid acquisition of as-built topography (improving accuracy requires ground control points and sufficient photo overlap).

• Mobile scanners and others: Recently, mobile mapping systems that mount equipment on vehicles to measure while driving, and handheld scanners (with SLAM technology) that allow a person to walk and capture point clouds, have emerged. Also, simple LiDAR sensors built into smartphones have made casual point cloud measurement possible. However, standalone smartphone measurement accuracy and reliability are limited, and high-performance equipment or RTK corrections are essential for precise surveying.

Characteristics of point cloud scanning and its use in as-built records

The greatest characteristic of point cloud scanning is that it can digitize the site’s shape as-is into three-dimensional data. The acquired point cloud data is essentially a full-scale digital archive of the site. Unlike conventional photo records or flat drawings, point clouds allow you to measure any dimension later or create arbitrary cross-sections. For example, if after surveying you decide “I want to check the height in another area,” you can perform additional measurements in the digital space without returning to the site if you have the point cloud data. Therefore, keeping point clouds as as-built records is extremely useful for future review or evidentiary purposes.

Specifically, an increasing number of projects store pre-construction terrain and structures as point cloud data. Point clouds can faithfully record minute irregularities and tilts, so they are useful when “previous conditions” are needed later. For instance, during planning for an additional construction, opening stored point clouds will reproduce the past condition as an accurate 3D model, allowing you to check for discrepancies with the design in advance. Because point clouds are digital data, they are easy to share among stakeholders and gain value as materials for client explanations or maintenance records. As-built conditions that used to be recorded by photos and notes can now be preserved as precise and comprehensive digital records by utilizing point cloud scanning.

Comparison with conventional methods: advantages of point cloud scanning

Conventional as-built records have been created by combining photos and survey data from a limited number of points. Manual surveying can only confirm dimensions point by point, and there is always a risk of “missed measurements” or human error. Point cloud scanning, on the other hand, offers the following advantages:

• Capture shapes with high density and high accuracy: Point cloud data composed of millions to hundreds of millions of points allows detailed measurement of the subject down to every corner. With proper equipment, shapes can be recorded with millimeter-level accuracy, enabling detection of tiny deformations or steps that were often overlooked by traditional methods.

• Non-contact and safe: Because measurement is non-contact via laser irradiation or remote photography, measurements can be taken without approaching dangerous areas. Even unstable slopes or high structures at risk of collapse can be safely recorded from a distance, improving worker safety.

• Greatly improved efficiency: Wide-area surveys and recording of complex shapes can be completed in a short time. Work that used to take two people several days may be done by one person in a few hours with point cloud scanning. This significantly reduces labor and time and enables rapid understanding of site conditions even with limited personnel.

• Versatile use of 3D data: Acquired point cloud data allows free measurement of dimensions, generation of cross-sections and plans, comparison with design data, and deviation checks for as-built verification. Overlaying with BIM/CIM 3D models or GIS maps enables advanced analysis for construction planning and maintenance. Because various deliverables can be derived from a single as-built point cloud, it reduces re-measurement and streamlines report preparation.

• Improved reliability of records: Point clouds can be stored as digital data on cloud services or servers, avoiding degradation or loss like paper drawings or photo albums. If stored with date and coordinate system information, they can serve as convincing official evidence. They also prevent human omissions or reporting errors, making point cloud scanning a highly reliable method for quality control and later verification.

Procedures and points to note when using point cloud scanning

In practice, follow these steps when using point cloud scanning, and keep the accuracy-related points in mind at each stage.

• Measurement (data acquisition): First, plan the survey and decide the equipment and methods to be used. Select terrestrial laser scanners or drones according to the subject and site conditions, and install known points (control points) or target markers in advance as needed. Set the equipment correctly during measurement and acquire point clouds from multiple directions to avoid blind spots. Especially for large areas, divide the area and scan with overlapping coverage to facilitate later merging. Also pay attention to weather conditions and avoid measuring in rain or dense fog (because raindrops or fog become noise points and degrade accuracy). For photogrammetry, ensure sufficient brightness and adjust camera settings and flight altitude to achieve the required point density. Careful planning and measurement at the acquisition stage greatly reduce later work and errors.

• Data processing (point cloud generation and editing): Bring the acquired data back to the office and generate and process point clouds on a PC. For laser scans, align (register) point clouds from multiple scan positions in specialized software to integrate them into a single coordinate system. For photogrammetry, import captured images into software and perform feature matching to generate a point cloud. Raw point clouds often include unwanted noise points (from passersby, moving machinery, or scattered raindrops), so use filtering to clean the data. Using control points or RTK positioning data to assign an absolute coordinate (geodetic coordinate) to the entire point cloud makes it easier to overlay with other drawings or maps. In this processing stage, check for misalignments between point clouds and correct errors, and, if necessary, decimate the point cloud to reduce data size. Point cloud datasets often reach millions to tens of millions of points and can heavily tax a PC, so consider high-performance PC environments or cloud processing services.

• Data sharing: The value of point cloud data is realized when it is shared and used among stakeholders. Because files are large, you can hand them over with USB memory or external HDD, but cloud services accessible over the Internet are more convenient. Uploading point clouds to the cloud enables remote technicians or clients to view the same 3D data via a browser. With a dedicated viewer, supervisors or partner companies who cannot come to the site can confirm the as-built 3D model in real time. When sharing externally, be sure to include metadata such as coordinate system, units, and acquisition date. For sensitive sites, implement access restrictions and consider information security. Establishing a smooth sharing system makes communication and consensus-building using point cloud data more efficient.

• Data storage: Store point cloud data obtained as as-built records appropriately so it can be used in the future. Create project folders on internal servers or cloud storage and organize point cloud data along with generated models and drawings. Including the site name and acquisition date in file names and leaving survey conditions (equipment used, coordinate system, accuracy information) as metadata will help future reuse. If file sizes are large, save in compressed formats (e.g., LAZ, the binary-compressed form of LAS) to reduce size and improve portability. The most important point is to take reliable backups so the detailed as-built data you acquired is not lost. Regularly replicate data to different media and locations to guard against data loss. Properly stored point cloud data becomes a valuable digital asset that vividly restores past site conditions even after many years.

Field use cases

Below are representative examples of how point cloud scanning assists in actual fieldwork. Point cloud data supports a wide range of fields from disaster response to construction management and infrastructure maintenance.

Use in disaster recovery

At sites affected by landslides from heavy rain or earthquakes, initial as-built surveying of damage is essential. Traditionally, personnel had to enter unstable slopes and observe survey points one by one with total stations, which was extremely dangerous and time-consuming. Recently, drone aerial photography and terrestrial LiDAR point cloud measurement have been active in such situations. For example, at large-scale slope collapse sites, capturing the collapse area rapidly from the air with a drone and converting it to a point cloud allows the entire damaged topography to be recorded safely and remotely (non-contact and safe). Using the acquired 3D point cloud, the volume of displaced soil can be calculated instantly, or by overlaying with pre-collapse topography or design drawings you can immediately see which parts have collapsed and by how much. This enables rapid estimation of required fill or removal volumes and streamlines consideration of recovery methods. Additionally, sharing point clouds via the cloud allows multiple experts, including those remote, to discuss recovery policies while viewing the same as-built model. The main advantage of point cloud scanning in disaster recovery is that it enables safe and rapid understanding of site conditions without wasting the precious time immediately after a disaster.

Use in as-built / conformity management

As-built (conformity) management is the construction management process of confirming and recording whether completed structures and formed ground match the design in shape and dimensions. Traditionally, thickness, width, and height at key points were measured manually with tape measures and staffs, but the number of points humans can measure is limited and there is a risk of missing subtle unevenness. With support from the Ministry of Land, Infrastructure, Transport and Tourism’s *i-Construction* initiative, as-built management using point cloud data is rapidly spreading (efficiency and comprehensiveness). By non-contact high-density scanning of completed objects with 3D scanners or drone photogrammetry, you can obtain a full-scale 3D copy of the finished work. On the point cloud, you can measure member dimensions and slopes, or overlay with the design 3D model and display differences as color-coded maps (heat maps) to intuitively check deviations from design values. For example, checking the thickness of tunnel lining concrete around the entire circumference with point clouds can detect subtle thin or thick areas that are difficult to find by hand measurement. In one case, introducing point clouds for as-built inspection of reinforced concrete construction reportedly reduced work time and costs by more than 70% compared to traditional manual measurement. As-built point cloud data can be submitted directly as electronic deliverables and construction records, helping prepare for inspections and preventing future disputes. Point cloud scanning, which achieves both quality assurance and efficiency, is becoming the new standard for as-built management.

Use in periodic infrastructure inspection

Periodic inspections are conducted for aging mitigation of social infrastructure such as bridges and tunnels, and point cloud technology is being introduced here as well. Traditionally, inspections relied on visual and sounding tests, but by combining point cloud scanning it becomes easier to quantify deterioration and compare changes over time. For example, for road bridges, photographing girders from an aerial work platform and converting them to 3D point clouds records crack locations and member deformations. In tunnels, scanning the internal cross-section with a laser scanner and comparing re-scans after several years allows detection of slight displacements or sagging in the cross-sectional shape. By overlaying point clouds, you can quantify changes such as “this part settled by ◯ mm (◯ in) compared to the previous time,” providing numerical evidence to prioritize repairs and assess structural health. In addition, applying high-resolution photographic textures to point clouds enables precise 3D localization and management of concrete surface cracks and spalling. With concerns about a shortage of inspection technicians due to a declining birthrate and aging population, point cloud scanning—capable of obtaining vast amounts of information in a single measurement—is expected to be an efficient support technology for infrastructure inspection. Note that point cloud use does not completely replace human diagnosis; it is a tool to complement and enhance inspections. Still, recording and accumulating periodic inspections digitally is significant and will form the foundation for future DX (digital transformation) in maintenance management.

Recommendation for simplified surveying with LRTK

As discussed so far, point cloud scanning is very useful for as-built records, but traditionally it required expensive equipment and specialized skills. Attention is now being paid to simplified surveying using the smartphone-based solution LRTK. LRTK is a device that attaches a small RTK-GNSS receiver to a smartphone, turning the phone into a high-precision surveying instrument. For example, when measuring point clouds with a smartphone’s built-in camera or LiDAR, LRTK can provide centimeter-level position information (half-inch accuracy) to each point, enabling acquisition of an absolute-coordinate-attached 3D point cloud on the spot. A major advantage is that you can obtain point cloud data that always aligns with the public coordinate system without complex control point setups or post-processing coordinate adjustments.

Using LRTK allows in-house staff to perform 3D surveys that previously required specialized contractors. The operation is simple: attach a dedicated small GNSS antenna (LRTK device) to the smartphone and launch a surveying app. Real-time high-precision positioning is fed to the phone, and then by walking the site while performing the smartphone’s LiDAR scans or taking photos, you can create high-precision point cloud models. Acquired point clouds can be automatically uploaded to the cloud and instantly shared with supervisors or clients in the office. You can also use AR overlays on the smartphone screen to display measured data and design drawings together and confirm the completed image on-site for explanations. The fact that this entire workflow can be realized with just a smartphone without specialized expertise is a major attraction for field use.

Point cloud scanning and LRTK are highly compatible and can dramatically improve the efficiency and accuracy of as-built records. Real-time RTK positioning and cloud integration greatly reduce the effort from measurement to sharing, lowering the barrier to adopting 3D technology in daily work. As digitalization and DX are increasingly demanded, consider adopting simplified surveying with LRTK at your sites. It will help you fully leverage the benefits of point cloud scanning and enable rapid and accurate as-built recording even with small teams.

Frequently Asked Questions (FAQ)

Q1: What equipment is needed to start point cloud scanning? A: Generally, a 3D laser scanner for measurement or a drone equipped with a camera for photogrammetry is required. Choose appropriate equipment depending on what you want to record. Drone aerial surveys are effective for surveying entire terrains, while terrestrial laser scanners are suitable for recording structural details at high accuracy. Recently, solutions combining high-precision GNSS and smartphones (such as LRTK) have appeared, enabling convenient point cloud scanning with a phone. For small sites or simple uses, a smartphone plus LRTK may suffice. For full-scale operation, you will also need PCs and software for point cloud processing in addition to measurement equipment.

Q2: Is the accuracy and efficiency of point cloud scanning comparable to conventional surveying? A: When performed appropriately, point cloud scanning accuracy can match or even exceed conventional surveying. Laser scanners have distance measurement errors on the order of a few millimeters, and photogrammetry can achieve accuracy within a few centimeters if adequate control points are used. Combining with RTK-GNSS can also ensure absolute accuracy and alignment with official control point coordinate systems. In terms of efficiency, point cloud scanning is excellent; a survey that once took a day using traditional methods can sometimes be completed in minutes with drone point clouds. Automatic acquisition of large numbers of points reduces human measurement errors, and easier detection of differences in post-processing helps prevent rework. However, achieving high accuracy requires proper equipment calibration and control point corrections as needed. With sound procedures, point cloud scanning is a very effective surveying method in both accuracy and efficiency.

Q3: How should point cloud data be shared and stored? A: Because point cloud data are large and difficult to handle, using cloud services or dedicated viewers is recommended. For example, upload point cloud data to an internal cloud and share a link so recipients can view and measure the 3D display in a browser. Several services, free and paid, exist that allow viewing without local software. Email attachments are usually impractical due to size limits, so online storage or physical media are more reliable for transfer. For storage, organize project folders and keep raw and processed data separate. Regular backups in locations different from the primary storage are advisable. If you plan to reuse data later, save in common formats (LAS, PLY, etc.) so they can be opened in other software. In short, establish a system to ensure your important as-built data are not lost and can be used long-term.

Q4: Can point cloud scanning be done in rain or at night? A: It is generally best to avoid measuring in rain. For laser scanners, raindrops or snowflakes can be captured as points, resulting in noisy data and reduced accuracy. In photogrammetry, wet lenses or subjects can impair image analysis, causing lower accuracy or missing data. In light rain some teams may shield equipment with plastic and proceed, but this will require extensive post-processing to remove noise. When possible, choose fair-weather days for measurement. For night work, laser scanners can measure in darkness because the laser itself emits light; however, darkness can hinder equipment setup. Photogrammetry requires illumination to photograph subjects. If night measurement is unavoidable, consider using site lighting or cameras with flash. In any case, weather and lighting conditions directly affect point cloud quality, so select measurement days carefully.

Q5: What is LRTK and what does it enable? A: LRTK is an innovative device and solution for performing high-precision positioning and point cloud measurement using a smartphone. A small RTK-GNSS receiver is attached to a phone and used with a dedicated app. This transforms the smartphone into a versatile surveying instrument with centimeter-level positioning capability (half-inch accuracy), enabling accurate georeferencing of 3D point clouds and photos on-site. In other words, precise positioning and point cloud scanning that previously required expensive surveying equipment can now be achieved with palm-sized devices by introducing LRTK. For example, simply walking a site with a smartphone equipped with LRTK allows immediate acquisition of high-precision point cloud models aligned to geographic coordinates. The acquired data sync to the cloud so you can review the site’s 3D remotely from the office. Although LRTK is a product brand name, its concept is a “real-time position-corrected simplified surveying system.” It is easy to use even for those without specialized training, making it suitable for first-time point cloud scanning in the field. By using LRTK, you can quickly obtain high-precision as-built data even for small sites or emergency responses and easily reap the benefits of point cloud scanning.