Solar PV DX with AR × Point Cloud Scanning: How **LRTK** Is Changing **PVsyst** Design

Introduction

Digital technology is rapidly advancing in the field of solar power plant design. Among the widely used software for plant layout design and energy yield simulation is PVSyst. Recently, new technologies such as AR (augmented reality) and point cloud scanning have been introduced, accelerating digital transformation (DX) in design and construction processes.

Particularly noteworthy is the use of LRTK, which enables centimeter-level high-precision positioning and 3D scanning using an iPhone. This makes it possible to integrate precise on-site 3D data and AR visualization into design workflows centered around PVSyst.

This article explains in detail how point cloud scanning and AR display using LRTK contribute to design accuracy, design efficiency, and on-site visualization in solar power plant design with PVSyst. We will dive into topics such as centimeter-accurate positioning with an iPhone, 3D model creation from point cloud data, construction support using AR, and shadow analysis—targeting specialist readers like renewable energy developers and surveying technicians. Finally, we will naturally introduce the convenience of simplified surveying with LRTK.

Solar Power Plant Design with PVSyst

PVSyst is a solar power system design and simulation software developed in Switzerland and used worldwide. By inputting project location data, weather data, and module and inverter specifications, it can predict annual energy production. It also accounts for various factors such as shading from surrounding obstructions, temperature rise, and cable losses to calculate a realistic expected energy output. Using PVSyst in the design phase supports layout and equipment configuration decisions and the formulation of revenue plans through generation simulations.

In such simulations, the accuracy of input data greatly affects the reliability of the results. In particular, accurately modeling the effects of shadows from terrain undulation and surrounding buildings or trees is crucial. For example, underestimating a slight slope or the height of nearby trees can lead to overlooking actual energy losses. Conversely, overly conservative assumptions can negatively affect space utilization in the design and revenue forecasts.

Traditionally, designers obtained site information from field surveys, public topographic maps, and aerial photos and reflected that information in PVSyst’s 3D models. However, manual surveying and plan-based information have limitations: it’s not easy to comprehensively capture the exact heights of scattered trees or subtle elevation changes within a site. Detailed surveying in the early design stage also requires time and cost, so designers often must rely on approximate information. To address these challenges, the use of on-site information acquired via point cloud data has recently gained attention.

What Is LRTK?

LRTK is a next-generation positioning system composed of an ultra-compact RTK-GNSS receiver that can be mounted on a smartphone and a dedicated app. By attaching it to an iPhone or iPad, an ordinary smartphone becomes a surveying instrument with centimeter-level accuracy. The device integrates the antenna and battery and mounts on the top of the phone, connecting via Bluetooth, so no cumbersome wiring is required.

LRTK supports the centimeter-level positioning augmentation service (CLAS) provided by Japan’s Quasi-Zenith Satellite System (QZSS), enabling high-precision GNSS positioning even outside cellular coverage. It also supports RTK correction information via mobile networks, providing stable positioning from urban to mountain areas. Combined with the phone’s built-in sensors (accelerometer, gyroscope, and magnetometer), the device can also measure the device’s orientation (bearing) with high accuracy.

The dedicated app allows geotagged photo capture, trajectory recording while moving (log measurement), and 3D scanning using the phone’s camera and LiDAR. Acquired point cloud and photo data can be uploaded to the cloud for storage and sharing and retrieved as needed. The app also includes navigation features that guide users to recorded reference points or targets (AR arrow guidance) and AR display capabilities for designed 3D models, enabling measurement, visualization, and on-site instructions to be completed with a single smartphone.

Centimeter-Accuracy Positioning and Point Cloud Scanning with an iPhone

Using LRTK makes on-site 3D surveying achievable with unprecedented ease. Simply walking around with a smartphone in hand enables acquisition of point cloud data that reflects the surrounding terrain and structures. This is realized by combining the shapes captured by the iPhone’s camera and LiDAR sensor with high-precision positioning information provided by real-time kinematic (RTK). Point clouds with latitude, longitude, and elevation in a global coordinate system are automatically generated for each point, digitizing the site’s 3D model.

This method allows detailed 3D data to be collected quickly even across large sites. Unlike traditional laser scanners or drone surveys, it requires no advanced training or complex equipment and is intuitive to operate. For example, on a site covering several hectares, walking around the main areas can capture point clouds numbering in the millions within minutes to tens of minutes, dramatically improving the efficiency of initial surveys. The acquisition range can extend to about a 60 m radius, enabling coverage of tree tops and high structures at a distance—an especially notable advantage.

The acquired point cloud data are in absolute (global) coordinates, so post-processing alignment is unnecessary. Even when scanning multiple sections, all point clouds overlay on a unified coordinate system, greatly reducing the effort needed to stitch separate datasets together. Color (RGB) information is also attached to each point, making it easier to visually distinguish terrain and object types. These point clouds can be imported into CAD or 3D modeling software to generate terrain models, used to cut cross-sections and take measurements, and applied in many other ways.

3D Modeling from Point Cloud Data and Use in PVSyst

Point cloud data are a collection of countless points, but they can be visualized and processed into terrain models or 3D objects as needed. For example, elevation information for the ground surface can be extracted from the point cloud to generate a digital terrain model (DTM) and converted into a mesh-based terrain map. Feature objects such as buildings or trees in the point cloud can also be extracted and modeled as individual 3D objects.

These 3D model data are highly useful in PVSyst design. PVSyst allows construction of 3D scenes for nearby shading analysis, but traditionally designers often approximated on-site buildings and trees as simple boxes or planes. Using models derived from point clouds enables incorporation of precise, measured objects into PVSyst scenes. For instance, if accurate height and position data for trees both inside and outside the site are available, they can be reproduced as obstructions in PVSyst and used to perform detailed year-round shadow impact simulations.

Moreover, importing terrain data (elevation data) generated from point clouds into PVSyst makes layout design on sloped land more realistic. PVSyst supports importing terrain data in CSV or GeoTIFF formats, allowing panel placement to be considered directly on the measured terrain. Reflecting actual undulations in placement decisions enables proposals that reduce the amount of earthwork required (cut and fill) or adjust racking heights according to slope angles, facilitating optimization during the design stage. Since cut-and-fill volumes can be easily calculated from point cloud data, comparing multiple earthwork scenarios and evaluating economic feasibility becomes straightforward.

Improved Design Accuracy and Enhanced Shadow Analysis

The detailed on-site data provided by point cloud scanning dramatically improve design accuracy. When measured terrain and obstruction information are reflected in PVSyst models, the reliability of simulation results increases substantially. A particularly significant benefit is the improved precision of shadow analysis. Accurate height and position data for trees and structures allow comprehensive evaluation of shading effects by time of day and season, minimizing errors in estimated generation losses.

This also reduces project risk. In the past, insufficient field surveys sometimes revealed unexpected shading after construction began, forcing layout changes. Conducting precise shadow analysis in advance can prevent such situations. Likewise, when overly conservative safety factors led to underestimating generation, accurate data enable appropriate evaluation and help avoid missed opportunities due to underestimation.

Improved Design Efficiency: Rapid Surveys and Data Processing with LRTK

Efficiency from acquiring on-site data to reflecting it in designs is greatly improved. Point cloud scanning with LRTK dramatically shortens surveying work that used to take days, completing it on-site in minutes to tens of minutes. Design staff can collect the necessary data with a smartphone without arranging specialized survey teams or heavy equipment, reducing scheduling overhead.

Post-acquisition data processing is also simple. Point clouds and high-precision photos are uploaded to the cloud immediately, allowing office PCs to check and download the data right away. This enables designers to review acquired data remotely and perform additional measurements immediately if something is missing. Previously, detailed design often began only after survey data arrived; with LRTK, on-site surveys and design work can proceed in parallel and iteratively, shortening overall project timelines.

Also, point cloud data contain far more information than traditional cross-sections or plan views, so designers can freely cut sections and take measurements, reducing the need for additional field checks. Ambiguities can be analyzed in detail at the desk, lowering communication costs. For example, sharing a point cloud 3D model with contractors during meetings helps them intuitively understand site conditions that are difficult to convey with drawings alone, facilitating clear communication of design intent and smoother revision requests.

On-Site Visualization with AR: Sharing Design Concepts

Using AR technology, the planned appearance of a solar power plant can be visualized realistically on-site. With AR that offers high-precision positioning via LRTK, designed panel layouts and equipment 3D models can be overlaid onto the actual site. Unlike conventional AR, models tracked against RTK-GNSS positioning remain stable even when moving significantly, allowing accurate registration without drift. Point your smartphone camera and you can see virtual rows of solar panels and racking standing on otherwise empty land, letting stakeholders intuitively grasp a finished image that is hard to perceive from paper drawings.

This on-site AR visualization greatly aids communication among stakeholders. For example, at explanatory meetings with landowners or local residents, you can overlay the planned plant onto the actual landscape to show what the completed facility will look like. Designers themselves can discover issues unnoticed on the desk by walking the site with AR. Problems such as planned panels overlapping a depression on a slope or insufficient clearance from adjacent trees can be identified and addressed in the design by reviewing them in AR.

Construction Support: Layout Marking and Installation Guidance with AR

The combination of LRTK and AR is powerful not only in the design phase but also in on-site construction management. Traditionally, marking panel positions on-site required surveyors to lay out coordinates using tape or total stations. With LRTK, this task can be performed intuitively via AR. The design baseline lines and pile locations can be virtually projected onto the ground through the phone’s screen; when workers reach the correct position, markers appear, allowing them to place piles precisely by following on-screen instructions.

This AR guidance enables workers without surveying expertise to identify installation positions with centimeter-level accuracy. Even on vast mega-solar sites, the time-consuming process of measuring and installing each point individually is reduced, leading to fewer personnel and shorter work times. Boundary markers hidden in brush or reference points under snow that are difficult to confirm visually can also be located using AR-guided cues.

Digital technology also aids post-construction verification. Re-scanning the completed installation and comparing it with the design data can visualize construction deviations. On the LRTK cloud, functions are provided to overlay the design 3D model and as-built point cloud, displaying matching areas in blue or green and discrepancies in red as a heat map. This makes quality checks and early detection of corrective actions easier. At the same time, point cloud data can be used for as-built drawings, streamlining the creation of handover documentation.

Conclusion

Combining AR and point cloud scanning with LRTK is transforming the design and construction processes of solar power plants. Integrating PVSyst’s advanced simulation capabilities with precise 3D data acquired on-site enables data-driven design decisions in areas that previously relied on experience or assumptions. Higher design accuracy and efficient, waste-free planning are being realized, and DX is elevating every project stage from on-site visualization and consensus building to construction management and verification.

LRTK, which allows anyone to perform precision measurements akin to professional surveyors simply by attaching a device to a smartphone, is truly a trump card for on-site DX. Its convenience for simplified surveying will likely expand its use beyond renewable energy development into civil engineering and construction as well. By actively adopting the latest technologies, planning and constructing solar power projects is becoming ever smarter and more reliable.