Reducing Labor and Improving Efficiency with Volume Difference Comparison Between As‑Built Point Clouds and Design Data

On construction sites, it is critically important to verify that the completed terrain and structures have the shapes and dimensions as designed. However, conventional methods have relied heavily on manual measurements with tape measures and leveling instruments, making detailed inspection over wide areas difficult. A method that has attracted attention in recent years is to compare as‑built point cloud data (a collection of 3D points obtained by scanning the site) with design data (the planned geometric information) and visualize and quantify differences in volume and shape from the discrepancies. This article explains in detail what this “as‑built point cloud × design data volume difference comparison” entails, how it is performed, the outcomes that can be obtained, and the advantages that lead to greater efficiency in field operations. Finally, we also introduce efficient point cloud acquisition methods using the latest tools for simple surveying.

Contents

• What is a volume difference comparison between as‑built point clouds and design data?

• Data and preparations required for volume difference comparison

• Outputs and analyses obtained from volume difference comparison

• On‑site benefits of volume difference comparison

• Simple on‑site surveying: easily acquiring high‑accuracy point clouds with LRTK

• FAQ

What is a volume difference comparison between as‑built point clouds and design data?

First, let’s clarify the meaning of comparing “as‑built point clouds” and “design data.” As‑built point cloud data are three‑dimensional survey data of the site obtained by laser scanners, drone photogrammetry, mobile LiDAR, and so forth. They represent the ground surface and structures as a multitude of points—essentially a “digital replica” of the site. On the other hand, design data refers to the modeled or drawing information of the intended finished form prepared before construction. Examples include 3D design models created using BIM/CIM, civil engineering design drawings (CAD data), or numerical data indicating alignments and ground elevations (e.g., CSV files).

In quality control of construction (as‑built control), it is necessary to verify that the finished work conforms to this design data. Traditionally, heights and thicknesses at key locations were measured manually and differences from the design values checked. However, manual methods limit the number of measurable points, creating a risk that defects that “differ from the design” will be discovered only later in locations that were not measured. Organizing large numbers of measurement results and compiling reports manually was also a heavy burden. As a new approach to as‑built control, comparing point cloud data and design data has emerged. By overlaying and comparing the as‑built point cloud and the design drawings (or models) on a surface-by-surface basis, you can detect large‑area shape deviations without overlooking them.

In earthworks in particular, calculating the soil volume difference between the as‑built terrain and the design surface allows you to determine the excess or shortage of fill or cut. For example, you can easily see with point cloud analysis whether filling in an embankment has reached the planned elevation, or whether a road base has the required thickness. The point cloud × design data volume difference comparison is thus an approach that analyzes the 3D differences between as‑built conditions and design, and presents them numerically and visually, streamlining quality verification and construction management.

Data and preparations required for volume difference comparison

To perform a volume difference comparison between as‑built point clouds and design data, several data types and preparations are required. Below are the main steps in sequence.

• Acquiring the as‑built point cloud: First, obtain point cloud data of the site to be compared. Use high‑precision laser scanners, drone photogrammetry, or other methods to scan the construction area thoroughly. Recently, methods for easily acquiring point clouds have become more widespread, such as LiDAR on iPhones and iPads or smartphone‑connected simple surveying devices.

• Preparing the design data: Next, prepare the design data that serves as the reference for the finished form. For civil engineering works, this may include road or site development design drawings (plan and longitudinal/cross sections) or 3D design models (LandXML or CAD data). Even if the design data are only 2D drawings, you can assign elevation information in software to generate a design surface model and compare it with the as‑built point cloud.

• Unifying coordinate systems: To correctly compare the as‑built point cloud and the design data, both must be aligned in the same coordinate system. Usually, control points or known points are established on site and used to georeference the point cloud survey data. The design model should also be aligned to the same geodetic system or georeferenced based on known reference points. This coordinate unification enables the point cloud and design drawings to be spatially overlaid correctly.

• Performing difference analysis: Using dedicated point cloud processing software or analysis tools, compare the as‑built point cloud data with the design data (surface). There are two major comparison methods. One is a cross‑section comparison, where cross‑sections are extracted from the point cloud along arbitrary section lines and overlaid with design sections to check for differences. The other is a more comprehensive 3D difference comparison, where the vertical difference between each point in the point cloud and the design surface is automatically calculated and the height deviations across a wide area are visualized as a heat map (color‑coded map). With the 3D difference method, you can understand over the entire surface how much the as‑built finish deviates above or below the design.

• Output and saving results: Save and output the analysis results as needed. For example, you can compile difference heat maps as screenshots or PDF reports, export height differences and volume values as CSV files, or export a triangular mesh model showing the difference from the design surface as CAD data. These deliverables can be submitted as inspection materials or stored as records of construction management.

The above summarizes the typical workflow for volume difference comparison. The key point is that inspections formerly conducted with paper drawings and discrete points can now be performed based on digital surface information. The design data used for comparison may be a 3D model or based on 2D drawings, but it is important to apply appropriate processing according to the data format (e.g., generating a TIN surface or extracting sections). Also note that the accuracy of the as‑built point cloud acquisition and the precision of coordinate alignment determine the reliability of the difference results, so take care at these stages.

Outputs and analyses obtained from volume difference comparison

Difference analysis between point clouds and design data yields various useful pieces of information. Below are examples of the main analyses and deliverables.

• Height difference heat maps: With the aforementioned 3D difference comparison, you can generate color‑coded maps showing vertical deviations between as‑built and design. The color gradient makes it immediately clear which areas are “higher than design” and “lower than design,” enabling visual evaluation of the finish accuracy over a wide area.

• Cross‑section comparisons: For roads, embankments, and similar structures, checking sectional shapes at key locations is important. You can extract longitudinal and cross sections from the point cloud at arbitrary positions and create drawings overlaid with the design sections. Showing the difference between the as‑built profile and the design line on section drawings makes local excesses and shortages clear.

• Volume differences and earthwork quantities: One major objective of difference analysis is the calculation of earthwork volumes. From the gaps between the as‑built ground surface and the design surface, you can automatically compute the excess and shortage of soil. For example, you can quantify how much fill is needed where excavation was excessive, or how much cutting is required where fill exceeds the design. These quantities can be output as CSV lists or visualized on a colorized 3D model showing surpluses and deficits.

• Automated pass/fail judgments: Recent point cloud analysis software includes functions to automatically judge whether measurements are within specified tolerance ranges based on difference calculations. If you set allowable tolerances in advance, the software can determine whether each deviation falls within that range and automatically mark pass/require correction. This eliminates the need for inspectors to check an enormous number of measurement points one by one and contributes to streamlining as‑built inspections.

• Report and data output: The difference comparison results can be compiled into final reports or delivered as electronic data. Items previously submitted as paper documents with site photos and hand‑drawn sketches can now utilize digital evidence from point cloud analysis. For example, attaching heat map images and section drawings to a report with corresponding numeric tables enables the automatic generation of convincing as‑built reports. There is also a growing trend to electronically deliver the point cloud data itself or CAD data of the difference results (for example, DXF files showing the as‑built and design 3D shapes).

By performing volume difference comparisons, you can extract a variety of concrete quality assessment data from point clouds obtained on site. Small errors that were missed by manual inspection become visible in the data, making objective evaluation of construction accuracy easier. In public works in particular, the Ministry of Land, Infrastructure, Transport and Tourism recommends as‑built control using 3D data such as point clouds as part of the *i‑Construction* initiative. The results obtained from volume difference comparisons form the foundational data for this surface‑based as‑built control.

On‑site benefits of volume difference comparison

So, what specific labor‑saving and efficiency effects can be achieved by utilizing difference comparisons between as‑built point clouds and design data? Below are the main benefits.

• Prevention of measurement omissions and oversights: Because point clouds measure the entire site as surfaces, there is no need to “infer the whole from a few points” as with conventional methods. You can comprehensively verify every corner of the finished shape, greatly reducing the risk of overlooking local defects. As a result, the occurrence of rework due to “parts being different from the design” discovered later is also suppressed.

• Speeding up as‑built inspections: The ability to automate pass/fail judgments through difference comparison directly improves inspection efficiency. For large structures in particular, the number of measurement points can be enormous, but if the software automatically sorts measurements into within/outside tolerance, inspectors can focus only on abnormal areas. Aggregation of inspection results can be completed with the push of a button, significantly reducing the total time required for inspections.

• Reducing effort in report creation: If you can automatically generate an as‑built report summarizing the point cloud × design comparison results, administrative workload is reduced. Tasks that used to require manually pasting photos and drawings and compiling tables can be completed succinctly through the report output functions of analysis software. Objectively supported reports with charts and numerical data are also effective explanatory materials for the client and reduce the stress of document preparation.

• Optimizing construction processes: Using volume difference comparisons during construction enables early detection and correction of issues. For example, by regularly scanning the site and comparing it with the design, you can notice deviations early and perform course corrections. Making minor adjustments early is preferable in cost and time to large-scale rework after completion. In other words, difference comparisons accelerate the PDCA cycle and support both quality improvement and cost reduction.

• Improving safety and productivity: Non‑contact, rapid surveying with point cloud technology contributes to worker safety. Dangerous slopes or high areas can be scanned without personnel entering hazardous zones, reducing safety risks. Furthermore, as‑built measurement and aggregation work that previously took days can be dramatically shortened, allowing personnel to be allocated to other important tasks and improving overall site productivity.

As described above, the as‑built point cloud × design data volume difference comparison is a solution that can transform how work is done on site, not just verify accuracy. It reduces surveying and inspection labor while improving quality and decreasing rework risk, resulting in enhanced project efficiency and lower costs. As reflected in Ministry guidelines, this method will become an important piece of construction DX and ICT construction going forward.

Simple on‑site surveying: easily acquiring high‑accuracy point clouds with LRTK

Now that the usefulness of volume difference comparison is clear, we will touch on how to easily acquire point cloud data on site. Conventional 3D scanners and drone surveys required expensive equipment and specialized skills, but simpler solutions have recently emerged. A representative example is a simple surveying system called LRTK.

By using LRTK (※), anyone can easily measure point clouds on site by combining a smartphone and a small GNSS receiver. The operation is simple: walk around the site and capture images with your smartphone, much like recording a video, and obtain surrounding 3D data. The key point is that LRTK assigns high‑precision position coordinates (absolute coordinates via RTK‑GNSS) at the time of capture, so you can obtain geodetically referenced point clouds immediately without post‑processing. Even with about 5 minutes of on‑site work, you can acquire high‑density point clouds over a wide area, and display and compare multiple point clouds and design data in a cloud‑based 3D viewer.

On LRTK cloud services in particular, simply uploading the captured point cloud data allows volume calculation and difference analysis to be performed with one click. There is no need to master complex software: colorized heat maps that visualize finish deviations and automatic calculation of fill/cut volumes relative to design drawings are performed instantly. The workflow of scanning on site → cloud processing → confirming results makes it possible to decide on the spot whether additional fill is required and immediately reflect that in the work, enabling a rapid PDCA cycle.

In this way, simple surveying with LRTK greatly lowers the barriers to point cloud utilization and brings 3D technology within reach as a daily construction management tool. To maximize the efficiency gains from volume difference comparison, consider adopting the latest on‑site technologies that are easy to use.

*(※LRTK: A name for an innovative mobile 3D measurement solution that leverages high‑precision real‑time positioning (RTK) technology.)*

FAQ

Q: What software is required to compare point clouds and design data? A: Software that supports point cloud processing and earthwork calculations is required. Examples include optional functions in civil engineering CAD software and dedicated point cloud analysis software. Recently, cloud services that allow point cloud upload and difference analysis (e.g., LRTK cloud) have become more common, enabling processing in a browser without installing software.

Q: Can I acquire point clouds without a 3D laser scanner? A: Yes. Point clouds can be obtained by drone photogrammetry, iPhone/iPad LiDAR, or simple surveying devices that combine a smartphone with GNSS (such as LRTK mentioned above). Choose the method that suits the required accuracy and coverage: use a smartphone for small surveys and a fixed laser scanner for large, high‑precision surveys.

Q: Can comparison be performed if the design data exist only as paper drawings? A: Yes. You can extract necessary section and base terrain information from paper drawings and recreate a digital design surface in point cloud software. For example, you can manually enter key elevations and generate a TIN surface or draw a design profile polyline to model sections. Although more laborious, it is feasible to compare point clouds with design data even when original 3D design data are unavailable.

Q: How reliable are the difference comparison results in terms of accuracy? A: The results fundamentally depend on the accuracy of the point cloud measurements. If you survey with high‑precision equipment and perform accurate coordinate alignment, comparisons can be made with errors of less than a few centimeters or less (a few inches or less). However, if the point cloud density is too low or there are errors in coordinate transformation, the results may be affected. Perform quality checks on the point cloud before analysis and, if necessary, calibrate with ground control points to obtain highly reliable difference results.

Q: Will volume difference comparison between as‑built point clouds and design data become mainstream in the future? A: Yes, that trend is already underway. Under the national promotion of *i‑Construction*, as‑built control using three‑dimensional data is being standardized. As point cloud technology becomes more widespread, workflows for comparative analysis are being established across companies. In the future, it will become commonplace even on small‑ and medium‑scale sites to acquire point clouds with a tablet on hand and immediately check deviations from the design. Advanced companies and local governments are already actively implementing these practices.