Streamline Slope As-Built Management: Improve Accuracy and Simplify Documentation with 3D Point Clouds

Introduction: Current Challenges in Slope Works and As-Built Management

In slope works (such as framed slopes and sprayed concrete slopes), controlling the final shape (as-built management) is essential for quality assurance. On site, measurements are taken after completion to confirm whether the slope has been constructed with the design’s prescribed gradients and thicknesses. However, measuring on steep slopes presents problems in both safety and workload. Conventionally, survey staff climbed the slope to measure lengths with tape measures or leveling rods, or picked up points with a total station. Manual measurement is hazardous and tends to limit the number of measurable points. In addition, compiling the collected data into drawings and tables for documentation is time-consuming, placing a burden on construction managers.

As-built management standards require grasping the shape of the entire slope. For example, the Ministry of Land, Infrastructure, Transport and Tourism (MLIT) standard specifies that the entire slope surface, measured away from the slope crest and slope toe, must be checked for vertical deviation from the design surface at all points (the allowable error is generally about ±5 cm (±2.0 in))[^1]. In other words, you must prove that no location on the slope deviates from the design by more than 5 cm (2.0 in). In practice, though, dangerous conditions and time constraints often lead to measuring only a few representative points. As a result, there is a risk that undetected irregularities exceeding the tolerance may be hidden in unmeasured areas.

Furthermore, preparing documentation for submission to the client consumes time. Organizing measurement results and creating as-built management charts (comparison tables of design and measured values, cross sections, etc.) was traditionally done manually. As the number of measurement points increases, the amount of input into spreadsheet software and plotting on drawings grows, increasing the chance of human error. Thus current slope as-built management faces a dilemma: “We want accuracy, but safety and effort are concerns,” and “We want efficiency, but the documentation must be properly prepared.”

Balancing “Accuracy,” “Workload,” and “Documentation” in As-Built Management

For construction managers, it is important to balance accuracy, workload, and documentation in as-built management. Prioritizing accuracy requires detailed measurements across the entire slope, which demands significant time and manpower. Conversely, reducing effort by decreasing measurement points risks missing defects in unmeasured areas, as mentioned above. No matter how accurately you measure, inadequate paperwork will fail inspections. To satisfy clients and inspectors, documentation that conforms to standards (for example, as-built management charts and photographs) must be prepared, which also requires effort.

At many sites, experienced technicians select measurement points based on their judgment and strike a pragmatic balance between accuracy and efficiency. However, methods that rely on individuals are prone to subjectivity, making it difficult to maintain the optimal balance consistently. Especially in slope works, check points vary with topography and construction methods, so standardized simplification is difficult. Consequently, practices such as “measuring more than necessary to avoid rework” or “working overtime because paperwork isn’t finished” are commonly observed.

What is needed is a system that increases efficiency while ensuring accuracy and automatically produces the required documentation. Rather than indiscriminately increasing measurement points, an approach that innovates the measurement method itself to cover the whole area while reducing labor is necessary. This is where 3D point clouds for as-built management are gaining attention. Point cloud measurement technology can record slope shapes with a volume of information that is in a different dimension from conventional methods, offering the potential to reconcile accuracy and efficiency.

TS As-Built vs Point Cloud As-Built: Which Suits Your Site?

As-built measurement methods include the conventional method using TS (total station) and the increasingly common point cloud scanning method. Understand the characteristics of each and consider which is more appropriate for your site.

TS as-built management is familiar to experienced personnel. Using a total station or leveling instruments, coordinates and elevations are measured at specific points. For slopes, transverse survey lines are drawn at regular intervals, and representative points on each line (slope toe, slope shoulder, several intermediate points, etc.) are measured for position and elevation. TS provides very high accuracy, with single-point errors on the order of millimeters, so it excels at ensuring point accuracy. However, TS acquires “points” at a time, and covering a large surface requires measuring many points one by one. For wide slopes or slopes with many irregularities, using only TS to understand the entire surface is extremely labor-intensive. TS surveys are usually conducted by teams of two or more (operating the survey instrument and the prism rod), which also poses efficiency issues. Although robot total stations that enable one-person surveying and reflectorless modes have increased the possibility of single-operator measurement, the basic procedure of “pointing the instrument at each location to be measured” remains the same.

Point cloud as-built management, on the other hand, uses datasets consisting of vast numbers of measurement points obtained by laser scanners or photogrammetry. A single scan can capture millions to tens of millions of points, enabling surface-based surveying of the entire slope. Whereas TS “measures with points,” point clouds “measure with surfaces” [Schematic diagram of conventional TS vs point cloud]. For example, placing a terrestrial laser scanner (TLS) on a tripod and projecting lasers onto the slope can non-contact capture extensive three-dimensional data in minutes. A drone equipped with a camera or LiDAR can fly over and image the area, allowing measurement of hazardous steep slopes without human entry. Increasingly, devices operable by a single person are available, making them attractive in labor-short sites.

However, point cloud measurement has caveats. Historically, there were high barriers such as the cost of dedicated scanners and drones, the PC environment needed for data processing, and operator proficiency. Because results are a large collection of points, specialized software processing is required to extract needed dimensions or produce drawings. Recently, these challenges have been mitigated: services enabling point cloud acquisition with inexpensive mobile devices and advancements in point cloud processing software have created environments where initial investment and specialist knowledge are less prohibitive.

Choosing the method that suits a site depends on slope scale, geometry, and desired deliverables. For small, simple slopes, TS measurements at representative points may suffice. For large, complex terrain or sprayed surfaces with many irregularities, point cloud scanning offers significant advantages. In practice, hybrid operations combining TS and point clouds are common: TS measures control points and check points with high accuracy, and those points are used to georeference point cloud data, allowing point clouds to achieve accuracy comparable to conventional methods. This hybrid approach enables point cloud use even in tunnels where GNSS is unavailable, complementing each method’s weaknesses. In short: “Point clouds excel at high-density coverage over wide areas, while TS excels at precise accuracy for specific points.” Leveraging each method’s strengths makes it possible to grasp slope conditions more efficiently and accurately than before.

Strengths of 3D Point Clouds: Measuring Surfaces Captures Irregularities

The greatest advantage of using 3D point clouds is that the entire slope can be measured without omissions. Conventional methods inevitably leave “gaps between points,” but high-density point clouds fill those gaps with detailed data. For sprayed concrete slopes, for example, variations in workers’ spray thickness can create subtle surface irregularities, but point clouds can record those irregularities in three dimensions. Changes that TS surveys might miss become obvious when generating color-coded maps or cross sections from point cloud data. You can check across the entire surface whether “there are any depressions where thickness is insufficient?” or “are there any protrusions?”—enabling early detection and correction of construction defects.

Point cloud measurement’s non-contact, remote nature is also important. There is no need for people to enter steep or freshly sprayed weak slopes; shapes can be captured safely from a distance. This reduces the risk of occupational accidents and cuts costs related to scaffolding or assigning safety observers. There are reports of steep near-vertical rock slopes and high-mounted framed slopes—previously difficult to measure manually—that were safely and quickly surveyed with laser scanners or drones.

Point clouds also offer clear efficiency benefits. A single scan can cover a wide area in a short time, drastically shortening total work time. In one earthwork site, a TS survey took three days to measure the current topography of several hectares, whereas TLS took two days and drone photogrammetry completed it in half a day. In another case, surveying with a laser-scanner-equipped drone took one-sixth the time of the traditional method, reducing overall workdays by more than half. Thus point cloud technology dramatically increases surveying productivity, directly contributing to shortened schedules and reduced labor costs.

Point clouds also provide confidence in accuracy. Some may worry that measuring a wide area at once increases error, but advances in laser scanners and photogrammetric analysis now allow point cloud surveys to achieve errors on the order of several centimeters to several millimeters (i.e., a few centimeters (a few in) to a few millimeters (a few in)). With proper control point corrections, the ±5 cm (±2.0 in) criterion required for as-built management can be comfortably met. Comparative studies of earthwork volume calculations and as-built measurements versus conventional methods have reported quantity differences within about 1%. In short, point cloud scanning is a technology that can reconcile accuracy and efficiency.

Workflow of As-Built Management Using Point Clouds and Simplified Reporting

Let’s look at the actual workflow for as-built management using point cloud data. The key is that once digital three-dimensional data are captured, subsequent processing and report generation can be largely automated. A typical procedure is as follows.

• Measurement planning and control point placement: First, plan how to acquire point clouds according to the slope’s shape and extent. For laser scanners, consider scanner locations and scan ranges; for drones, plan flight paths; for mobile devices, plan walking routes. If necessary, install and survey known control points so the acquired point cloud can be assigned accurate coordinates (existing construction control points may be used).

• Point cloud data acquisition: Perform 3D scanning on site according to the plan. For TLS, set up a tripod at a vantage point and emit laser; for drones, capture photos or laser scans from the air; for mobile devices, walk along the slope while LiDAR scans beneath the device—methods vary by equipment. The important thing is to acquire a point cloud that covers the entire slope in a short time. For example, using an iPhone’s LiDAR, you can gather a point cloud covering tens of meters (tens of ft) around you simply by pointing the camera and walking in front of the slope.

• Data processing and comparison with design data: After acquisition, process the point cloud on a PC or in the cloud. Filter out unwanted points (pedestrians, machinery, trees, etc.) and keep only the terrain and structure points. Then overlay the point cloud with the design geometry. If a 3D model or design cross-section data created from drawings are available, align them in the same coordinate system and compute deviations for each point. Control point information is crucial here. With properly georeferenced point clouds, you can accurately evaluate whether areas conform to design by comparing coordinate differences.

• As-built evaluation (pass/fail judgment): Analyze the differences between the point cloud and the design and determine whether values are within the specified tolerances. Specifically, check for each point whether the vertical deviation (or horizontal deviation) from the design surface is within ±◯ cm. Specialized software can display the differences as color-coded heat maps. For example, areas within design tolerance can be shown in blue–green and areas exceeding tolerance in red. This makes it intuitive to identify which parts of the slope are within tolerance and which are out of tolerance. If everything is blue/green, it’s a pass; if there are red areas, those parts can be corrected.

• Automated report generation: Once the as-built evaluation is complete, summarize the results in the required format. The strengths of digital measurement are evident here as well. Point cloud software or linked tools can automatically generate as-built management charts. For example, lists of design values, measured values, and differences for representative points, or cross sections with pass/fail indications, can be output with a single click. Tasks that previously required manual entry into Excel can now be exported directly from point cloud data to Excel or PDF, reducing human error and shortening processing time. MLIT’s “Guidelines for As-Built Management Using 3D Measurement Technologies (draft)” defines evaluation methods and submission formats for using point clouds, and software that follows these guidelines can automatically create compliant reports[^2]. Increasingly, submissions also include the point cloud or 3D model itself as electronic deliverables. Delivering digital data allows clients to inspect details anytime, covering aspects that paper drawings may not convey. This reduces document volume and streamlines inspection exchanges.

As shown above, point cloud-based as-built management digitizes the entire process from field measurement to report creation. Because on-site data are directly used for analysis and output, manual data entry is eliminated. This provides major labor savings and improves both measurer workload and data reliability.

Case Study: Halved Workdays and Automatically Generated Reports on One Site

Here is a real example showing how introducing point cloud as-built management dramatically improved work efficiency. On a certain mid-sized slope project, measuring the as-built of a sprayed slope several tens of meters long used to require a three-person team working two full days. On the first day, heights and lengths at various locations were measured with TS, and on the second day data were compiled and as-built charts prepared. After trialing 3D point cloud measurement, on-site measurement was completed in half a day and the remaining half-day of office work produced the reports. Specifically, one worker scanned the entire slope with a laser scanner (about 1 hour), processed the acquired point cloud in the office, compared it with the design data, and output a heatmap-equipped as-built report the same day. Workdays were reduced by about half compared to the conventional method, and because only one person was required, total man-hours were drastically reduced.

On that site, full-coverage checks with point clouds also allowed preemptive detection of thin areas. A slight spray variation that could have been missed by conventional methods (thickness shortfall of about 3 cm (1.2 in)) appeared red on the heatmap, and re-spraying brought it into the allowable range before inspection. The construction manager commented, “Thanks to point clouds, we have no lingering concerns and can submit confidently to the client.” Regarding as-built charts, they said it was much easier because “there’s no need to manually write measurement point names and values; we just print the automatically generated Excel sheet.”

Other MLIT examples report that drone aerial surveys halved the time required to compile as-built from 12 days to 6 days, and that laser scanner use completed surveys in one-sixth the manpower time compared to conventional methods[^3]. These successes demonstrate the substantial time-saving potential of point cloud technologies. Importantly, increased data coverage also reduces the risk of rework. Avoiding re-measurement and remedial work due to missed measurements or errors contributes to overall schedule stability. Point cloud as-built management follows the idea that “hasten slowly”: investing effort in digital measurement upfront can prevent downstream trouble and labor.

Client and Inspector Reactions and the Reality of Digital Inspections

When introducing new technology, a common concern is how clients and inspectors will react. Initially, some inspectors questioned whether point clouds were truly accurate or whether differing formats from drawings would cause problems. However, with national promotion of ICT construction (i-Construction), understanding and trust in 3D point cloud use have been gradually increasing.

Clients often comment that point clouds are “visually easy to verify.” Previously, only a few measured points appeared on paper drawings, but a point cloud heatmap shows the slope’s quality over the entire surface in color, making it easy to grasp at a glance. If you explain, for example, “The point with the largest deviation from the design is only about +3 cm (1.2 in), which is within tolerance,” experienced inspectors tend to respond, “I see—this shows overall good workmanship,” and accept the results. In many cases, digital data are actually more persuasive.

On-site inspections still sometimes require paper forms. Older inspectors may prefer printed charts, so attaching paper output in the prescribed format is prudent. Since the source data are derived from point cloud analysis, paper outputs are free from transcription or calculation errors. If an inspector asks, “Where and how did you measure this time?” you can show the measurement points and cross sections positioned on the point cloud, or display the 3D model on a tablet. Some clients now accept electronic submission of as-built management data, and some prefectures support electronic deliverables (digital submissions)[^4]. The trend toward digital inspection is likely to accelerate.

At MLIT-supervised projects, the principle application of CIM (use of 3D models) is advancing, and attempts are underway to perform supervision and inspection by checking point clouds and BIM models on tablets. Local governments are also conducting training on maintenance and inspection using point clouds as part of infrastructure DX, and exemplary ICT utilization cases are being recognized. Given this background, concerns that “point clouds will be shunned in inspections” are diminishing. Well-captured and well-processed point cloud deliverables are welcome to clients because they provide abundant evidence to support as-built claims and facilitate accountability for construction quality.

Overall, client and inspector responses are becoming increasingly positive. However, it is important to remember that new technology does not automatically guarantee acceptance: deliverables must comply with existing standards. Using point clouds does not exempt you from submitting the prescribed charts and numeric reports. With the MLIT guidelines (draft) clearly positioning point cloud use, field engineers and inspectors can now handle digital as-built artifacts according to common rules. With rule-making progressing, the environment for adopting new technologies on site has become safer.

Conclusion: Accuracy and Efficiency Can Coexist

The introduction of 3D point clouds is an effective solution to the challenges of accuracy, efficiency, and documentation in slope as-built management. The key point is that “you can increase efficiency while maintaining accuracy.” Where practitioners once had to choose between accuracy and efficiency, point cloud scanning allows both to be achieved at a high level. Wide areas can be measured in a short time while capturing fine details, reducing previously overlooked defects. This reduces rework and re-measurement, leading to overall time savings.

Digitalization of data also dramatically improves the efficiency of report creation and information sharing. If field-acquired point cloud data can be directly analyzed and reports output automatically, the labor required for manual documentation decreases. From a work-style reform perspective, this is a significant advantage. Reducing the long overtime associated with producing as-built reports will certainly reduce the burden on construction managers. Moreover, electronic deliverables simplify record-keeping for clients as well, benefiting both parties.

The technical barriers to entry are falling daily. What used to be expensive 3D laser scanners are now available for rent or as services, and small drones and smartphones can produce point clouds. Under MLIT’s i-Construction policy, ICT construction standardization is progressing and support systems for trying new technologies are being established (for example, as of 2024 the as-built management guidelines (draft) have been expanded to cover all work types, and a “slimmed-down” version is planned to further promote field application). In short, once psychological hurdles like “it looks difficult” or “inspections will be problematic” are overcome, the benefits are substantial.

Slope works will remain a vital part of infrastructure development and disaster prevention. 3D point cloud technology, which can both streamline and elevate quality control, is a powerful ally for construction managers. Consider adopting this as-built management method that reconciles accuracy and efficiency at your site—you will likely experience convenience and assurance that change the conventional wisdom.

Bonus: Smart As-Built Management Linking Point Clouds and Reports Using LRTK

If you’re interested in point cloud as-built management but unsure what to prepare, here is an easy-to-start solution: LRTK. LRTK is a modern 3D measurement system using smartphones; by attaching a small dedicated GNSS receiver to a smartphone, anyone can easily obtain high-accuracy point cloud data.

For example, attaching an LRTK device to an iPhone equipped with a LiDAR sensor lets you create a point cloud model simply by holding the smartphone and walking along the slope. LRTK supports RTK-GNSS for centimeter-level positioning, so the acquired point cloud is assigned global coordinates (absolute coordinates) and can be aligned with survey maps with comparable accuracy. Where 3D scanning once required devices in the millions of yen and specialist skills, LRTK can start with a single smartphone, making initial adoption barriers very low.

LRTK offers cloud services and dedicated apps that enable seamless workflows from point cloud acquisition to report creation. Specifically, smartphone-acquired point cloud data can be uploaded to the cloud on site, overlaid with design drawings in a browser to check deviations, and used to measure necessary dimensions. Heatmap displays are available with one click, allowing immediate judgment on whether the work matches the design. Templates for MLIT-compliant as-built management formats are also provided, enabling automatic output of as-built management charts (Excel or PDF) from point cloud data. For framed slope works, for example, items such as frame width, frame height, and slope length can be listed with differences from design values in the prescribed format. These outputs are in formats suitable for electronic deliverables, so there is no need to recreate documents.

In other words, LRTK makes “measure with a smartphone and submit the report as-is” smart as-built management possible. Because results can be checked and shared in the cloud immediately after acquisition, interim reports to clients and internal inspections are expedited. The app is intuitive, so even those unfamiliar with 3D or ICT can start using it after a few hours of training. LRTK also offers a photogrammetry mode, allowing point cloud generation from photos even on phones without LiDAR. Its flexibility to adapt to site conditions is another advantage.

Thus, LRTK is a groundbreaking tool that realizes point cloud × report linkage and brings efficiency to your smartphone. If you are considering improving accuracy, shortening time, and simplifying documentation for slope as-built management, consider starting with LRTK on your familiar smartphone. Leveraging cutting-edge digital measurement can help you take the first step toward smart as-built management that breaks with past conventions.

[^1]: MLIT “As-Built Management Standards” requires that all measured points on slopes be within ±50 mm (±1.97 in) of the design values, and that measurements be conducted over the entire slope surface (including top surface and small slope benches). [^2]: National Institute for Land and Infrastructure Management “Guidelines for As-Built Management Using 3D Measurement Technologies (draft) R7.3” details procedures from point cloud acquisition to representative value calculation, evaluation, and chart creation. Using software compliant with these guidelines makes specification judgment and report creation easier. [^3]: Reference cases: MLIT Kanto Regional Development Bureau i-Construction case collection (2019) reports efficiency improvements in as-built measurement by UAV photogrammetry. Lefixea Inc.’s comparative experiments also introduce examples where drone LiDAR surveying completed surveys in one-sixth the time of conventional methods. [^4]: Example of electronic deliverables: Niigata Prefecture “Reference Materials for Electronic Deliverables in ICT Utilization” (2020) shows procedures for submitting as-built management charts as PDFs and delivering point cloud data as electronic deliverables. Practices vary by local government, but digital submission of as-built data is gradually expanding.