LRTK Connecting 3D CAD/BIM Data and the Field: Construction DX with 3D Measurement

Introduction: Challenges in Connecting Construction DX and 3D CAD

In recent years, promotion of DX (digital transformation) has been accelerating in the construction industry as well. Use of 3D CAD and BIM data during the design phase has advanced, and creating detailed three-dimensional models of buildings and infrastructure has become common. However, the reality is that the 3D design data created in this way is often not being sufficiently utilized on actual construction sites. A large gap still exists between the digital models refined in the office and outdoor fieldwork, and this gap has become a barrier to construction DX.

For example, construction plans that look perfect on the design model often do not proceed exactly as drawn in the field due to terrain, surveying errors, and other influences. In addition, many onsite tasks still rely on paper drawings and craftsmen’s intuition, so even valuable 3D data can end up being “left unused.” The disconnection between design and construction leads to rework and mistakes, hindering productivity improvements.

The key to solving these issues and truly linking 3D CAD/BIM data with the field is 3D measurement technology. By leveraging the latest 3D scanners and positioning technologies, it becomes possible to capture the actual on-site conditions as precise digital data and directly compare and verify them against the design data. This article explains the forefront of construction DX enabled by 3D measurement and the solution that makes it easy to realize: LRTK.

The Disconnection between 3D CAD/BIM and the Field: Why the Shift? How to Compensate?

Why does the gap between digital and field arise? One major cause is misalignment of coordinates. No matter how precise the 3D model prepared in the design is, if alignment is insufficient when using it on site, the correspondence between the digital and the real will not match. For example, even if you try to display a 3D model on a tablet or smartphone at the site, ordinary GPS can have errors on the order of several meters, so the model will be displayed slightly offset from its actual position. Conventional AR (augmented reality) apps often require placing markers and manually aligning the model, and the larger the site, the greater the errors and effort, making accurate overlay difficult.

From the perspective of construction accuracy, “shift” also occurs. Even when construction is performed according to drawings, small cumulative errors in setting out or marking can result in the position or dimensions of structures deviating from the design values. Discovering and reworking these deviations later leads to significant costs and time loss.

So, how can we compensate for this “shift”? The key point is to treat design data and on-site measurement data in a common coordinate system. Specifically, it is necessary to introduce technology that can measure positions on site with accuracy on the order of centimeters, and to align the designer’s coordinates with the field coordinates. A strong candidate for this is the RTK-GNSS (Real-Time Kinematic satellite positioning) described below. Using high-precision positioning technology can minimize the discrepancy between the model and the actual object in advance, making it possible to link the digital and the physical without misalignment. In the next chapter, we will look in detail at the importance of high-precision 3D measurement and coordinate alignment.

High-Precision 3D Measurement and Coordinate Alignment Hold the Key

What matters most in connecting the field and design data is high-precision 3D measurement and coordinate alignment (unification). 3D measurement means capturing the site’s terrain and structures as digital three-dimensional data (point clouds or 3D models), allowing detailed recording of the site’s current state. However, no matter how detailed the point clouds are, if they are in a different coordinate system than the design data, matching the two requires extra work. This is where the power of positioning technology for coordinate alignment comes into play.

By employing satellite positioning technologies like RTK-GNSS, point clouds and photos acquired on site can be assigned absolute coordinates such as latitude, longitude, and elevation. Ordinary smartphone GPS can have errors of several meters, but with the RTK method, corrections can reduce errors to around 1-2 cm (0.4-0.8 in). For example, in Japan, using the centimeter-class positioning augmentation service (CLAS) provided by the Quasi-Zenith Satellite System “Michibiki” can supply high-precision correction signals. Attaching a high-precision GNSS receiver to a smartphone and performing RTK positioning allows acquisition of measurement data on site in the same coordinate system as the design drawings.

The point cloud data obtained this way with absolute coordinates can be overlaid immediately with BIM/CIM models created in the office for comparison. Because manual alignment is unnecessary, you can simply overlay the measured data on the design 3D model, greatly reducing back-and-forth between the field and design. Also, coordinate agreement significantly improves the accuracy of AR model projection described later. High-precision 3D measurement and coordinate alignment form the foundational technology that acts as the “pipe” connecting the digital and the field.

Requirements for AR Construction Management and 3D Data Alignment Technologies

To realize AR-based construction management, which has attracted attention in recent years, accurate alignment of 3D data is indispensable. When a smartphone or tablet held up at the site displays the design 3D model overlaid on the real object on the screen, what technical elements are required to enable such AR construction management?

• Smartphone and AR apps: As a convenient platform for field use, smartphones and tablets are the primary tools. Recent iPhones and iPads are equipped with high-performance cameras and displays and have sufficient processing power for AR rendering. Using dedicated AR apps (e.g., those utilizing Apple’s ARKit), design 3D data can be loaded, rendered, and alignment calculations executed on the device.

• LiDAR scanner: Some of the latest smartphones are equipped with LiDAR (light detection and ranging) sensors. LiDAR scans the surrounding environment with laser and acquires point cloud data, allowing the device to spatially understand site terrain and structures. This enables the model placed in AR to be aligned correctly in height as well, and to represent occlusion by existing objects. Furthermore, point clouds acquired by LiDAR can be used for as-built measurement, allowing comparison of terrain changes before and after construction and keeping records.

• Visual tracking technologies (ARKit, etc.): AR apps use VIO (Visual-Inertial Odometry) techniques leveraging camera images and device gyros to track the device’s pose in real time. This allows the virtual model to remain fixed in place even while the user moves. However, usual ARKit-based AR manages positions in relative coordinates, so errors accumulate as the user moves longer distances and the model can drift. Initial alignment also needs to be done manually, and especially on large outdoor sites there are limits to how accurately the model can be placed.

• High-precision GNSS (RTK): To add absolute-coordinate alignment to the visual stability of ARKit, RTK high-precision GNSS is used. As mentioned above, if RTK-GNSS can accurately determine the smartphone’s position in a global coordinate system, BIM/CIM models prepared with positioning information can be overlaid on site at the correct location without markers. This greatly reduces the effort of initial alignment, and because the model’s relationship with the real world does not collapse even while the user walks around, it is possible to maintain high-accuracy AR display across wide sites.

By combining the above technologies, a single smartphone can realize “accurately overlaying design data on the actual site.” With this high-precision AR construction management, anyone can intuitively share completed images on site before construction, check as-built conditions on the spot during construction, and more. It is a concrete on-site practice of construction DX.

Using Measured Point Clouds to Detect Design Deviations

High-precision measured point cloud data is powerful for comparison verification with design models. For example, scanning a completed structure in 3D and overlaying the acquired point cloud with the design 3D model makes it immediately clear which parts match the design and where there are excesses or deficiencies. By analyzing the differences between the point cloud and the design data and visualizing areas that exceed tolerance with color coding, you can digitally detect construction errors or finishing defects that would be missed by the naked eye or 2D drawings and correct them early.

For example, in earthwork, comparing the design ground model with the point cloud of the actual formed terrain shows whether embankment or cutting has been performed to the specified elevation using a colored heat map. Missing areas are shown in blue, and excessive piles are shown in red, so overall trends can be grasped at a glance. Previously, this information required many on-site height measurements and manual comparison with drawings; with point cloud data you can overview the as-built condition of the entire space. Utilizing measured point clouds makes QA inspections more efficient and comprehensive, directly reducing rework and ensuring quality.

From Point Clouds to As-Built Assessment, Volume, Area, and Cross-Section Calculations

Point cloud data can be used for many aspects of site management beyond difference checking with the design. From acquired point clouds, detailed quantity information that could not be obtained by manual measurement can be extracted. For example, the following analyses and calculations can be performed automatically from point cloud data:

• As-built assessment: Check whether the as-built shape of structures and terrain meets the specified dimensions and form. Measurements on the point cloud and difference checks against the design model by cross-section enable inspection results in accordance with as-built management procedures.

• Volume and earthwork calculation: Calculate volumes of embankment and excavation from point cloud data. For example, by comparing terrain point clouds before and after construction, you can immediately determine backfill or removal volumes on site. Tasks that used to require taking survey data back to the office for volume calculation software can now be completed on site.

• Area measurement: Surface areas such as road or floor surfaces and slope face surfaces can be measured from point clouds. Even irregularly shaped sections yield accurate area calculations when the contour is specified on the point cloud, improving inspection efficiency.

• Cross-section generation: Cut the point cloud at arbitrary locations to create longitudinal and cross-sectional drawings. This allows immediate on-site confirmation of excavation slopes, tunnel sections, etc., and smooth comparison with design sections.

By obtaining point cloud data, virtually all site measurements and quantity estimations can be performed digitally. With cloud-based point cloud viewers or dedicated software, distance and area measurements and section cuts can be done via a browser without specialized CAD software, enabling quick consensus building among stakeholders from site managers to clients by sharing data.

Real-Time Connection of BIM and the Field via Cloud Integration

To maximize the effectiveness of 3D measurement data and BIM models, cloud integration for information sharing is indispensable. Uploading point clouds, photos, and positioning data acquired on site directly to the cloud allows office personnel to confirm site conditions in 3D without being on site. For example, a field worker can perform a point cloud scan or survey with a smartphone and press a cloud sync button; minutes later a designer or client can view the latest site data from a browser on a PC. Not only planar survey point information but the acquired point clouds and photos are also plotted on a map, allowing immersive situational awareness even without visiting the site.

When cloud integration enables data to flow instantly between field and office, a “digital site” where BIM and the field are connected in real time is realized. Previously, survey results had to be brought back on a USB memory stick and loaded in the office, or drawings sent by email for review, but centralizing models and point clouds in the cloud greatly reduces such waiting times and effort. Design changes and sharing of construction status are conducted swiftly, and because all stakeholders can refer to a single source of the latest data, discrepancies in understanding are prevented. The combination of cloud and 3D data strongly supports collaboration across distances and drives construction DX.

“One-Person 3D Surveying” Realized by Smartphone × LRTK

A solution that integrates high-precision positioning, 3D measurement, and AR utilization as described above is LRTK. LRTK is a next-generation surveying system that mounts a compact RTK-GNSS receiver called the LRTK Phone on a smartphone and uses a dedicated iOS app and cloud service in combination. With this system, the smartphone you normally use instantly becomes a centimeter-precision surveying instrument with cm level accuracy (half-inch accuracy). From position measurement (positioning) to photo recording, point cloud scanning, and AR 3D model projection, the necessary functions are all-in-one, enabling a single worker to complete field tasks.

The biggest advantage of LRTK is that high-precision measurement can begin immediately without complex on-site setup or calibration. The high-performance GNSS antenna attached to the iPhone automatically receives satellite correction information, and the app performs real-time coordinate correction, so users obtain high-precision position information without special effort. For instance, simply walking around the site with LRTK records the trajectory at the user’s feet as a log with cm level accuracy (half-inch accuracy), while the surrounding terrain is successively converted into point clouds by the LiDAR scanner. Pointing the camera at a structure yields a detailed 3D point cloud model, and the acquired point clouds are stored in the cloud with coordinates in the World Geodetic System. Tasks that previously required skilled 3D surveying or as-built management can now be performed by anyone with just a smartphone in their pocket.

Moreover, using LRTK, you can upload BIM models of planned structures to the cloud in advance and call them up on site for AR display. The 3D models are automatically placed at the correct location and elevation, so they appear as if the completed objects exist there. Workers can view the smartphone screen and confirm virtual columns and walls rising from the ground while proceeding with pile-driving or foundation layout. The model follows without shifting even when moving to another location, allowing a single worker to sequentially confirm positions at multiple spots. One-tap point cloud scans of terrain, automatic calculation of differences between that data and the design model, and color-coded display of results on the smartphone are also possible. These functions can easily produce deliverables in line with the Ministry of Land, Infrastructure, Transport and Tourism’s 3D as-built management guidelines, contributing to streamlining paperwork and reporting.

Starting Construction DX with LRTK: The First Step to Building a Digital Twin

High-precision 3D measurement that a single person can complete with LRTK is a powerful first step toward construction DX. Accurately digitally recording site conditions and merging them with design data is essentially the process of building a digital twin (a digital twin of the site). Traditionally, digitizing and accumulating as-built data required considerable effort, but with LRTK, everyday patrol inspections and post-construction checks can easily accumulate 3D data. Fusing that data with BIM models forms a digital twin that links design and construction in real time, dramatically improving the accuracy and efficiency of construction management.

Once a digital twin is established, simulation and prediction for various problems occurring on site become possible, accelerating decision-making. Also, in the operation and maintenance phase after completion, the detailed 3D records obtained during construction are useful. As the foundation for promoting construction DX, LRTK plays the role of an entry point.

“Everyone inspecting construction while holding a smartphone and checking designs in AR” — that future site image is just around the corner. The tool making that familiar now is LRTK. By leveraging LRTK, which realizes simplicity, high precision, and all-in-one functionality without complicated procedures, please take your company’s field operations to the next stage. We expect LRTK to become your strong partner in taking the first step toward construction DX.