On May 15, 2026, the Japan XGRIDS Users Community 2026 was held at the Osaki Bright Core Hall in Tokyo. Nearly 100 attendees gathered, including not only XGRIDS users but also people from the construction/surveying industry and creators involved in VR/digital twins. The new handheld 3D scanner, “Lixel K2,” was also on display, and the event served as a platform to convey to many people the expectation that its use will further increase productivity.
Lixel The K2 can perform wide-area and high-resolution scans simultaneously.
The Lixel K2 is a professional-grade handheld 3D scanner developed by XGRIDS. It features a compact body weighing approximately 1,200g, integrating a 360-degree LiDAR, a fisheye camera, a high-definition camera, and an RTK module. It is the successor to its predecessor, the LixelKity K1 (hereinafter referred to as K1).
The general specifications are as follows:
| Lixel K2 | |
| Main unit weight | Approximately 1.2kg |
| LiDAR method | 360-degree and diagonal arrangement |
| scan distance | 40m or more (at 10% reflectivity) 100m (cutoff distance) |
| Scan speed | 200,000 points/second |
| Camera configuration | Fisheye panoramic x 2 + fixed front camera x 1 (total of 3 cameras, effective pixels 4000 x 3000 pixels each) |
| RTK | Standard built-in |
| Relative precision | 1cm |
| Absolute precision | 3cm |
| Real-time motion removal | Yes (World’s first) |
| storage | 512GB eMMC |
Diagonal placement of LiDAR sensors improves work efficiency.
The previous model, the LixelKity K1, had its LiDAR sensor mounted horizontally. Therefore, in order to perform an all-around scan including the ground in a single pass, it was necessary to hold the device tilted forward.
The LiDAR’s scanning range is understandable from its appearance, and for operators with a high level of hardware proficiency, the required tilt is immediately obvious. It’s easy to use. However, people unfamiliar with digital devices tend to hold it straight, which creates a problem of narrowing the LiDAR’s scanning range.
Therefore, the Lixel K2 changed the LiDAR sensor’s orientation to an angled position. Even when held straight, it has a -7 degree angle, allowing it to scan from the ground to the sky.
This isn’t just a spec upgrade; it directly leads to improved operations. Even when it’s difficult to dispatch surveyors or operators with specialized knowledge, it allows for less problematic scanning by on-site staff.
Addition of a fixed front camera – improves detail rendering in 3DGS.
In addition to LiDAR, the LixelKity K1 is equipped with two fisheye lens cameras on both sides, enabling the creation of point clouds, meshes, and 3D GS (Gaussian splatting). While it has the advantage of easily scanning entire spaces, scanning fine details such as inscribed text on monuments, signs, and posters required the cumbersome process of separately photographing these details with a digital camera or smartphone and combining the data.
Therefore, the Lixel K2 added a camera with a 50mm equivalent telephoto lens. With a total of three cameras, it became possible to acquire photorealistic high-resolution images simultaneously in a single scan. This not only improves the detail quality when generating 3D GS, but also solves the problem of the sun’s angle and color temperature changing over time during shooting.
RTK, which was previously an option, is now included as standard.
To provide highly accurate positional information to scanned point cloud data, RTK (Real Time Kinematics) is required, which uses GNSS (Global Navigation Satellite System, such as GPS and Michibiki) for positioning.
While the LixelKity K1 required a separate module to use RTK, the Lixel K2 has RTK built into the main unit, making it more cost-effective.
Real-time motion object removal improves the quality of on-site inspections.
The most distinctive feature of the Lixel K2 is its ability to perform real-time dynamic object removal. In public places such as tourist spots and busy streets where people are constantly coming and going, people, vehicles, and animals that pass by during scanning remain as noise in the point cloud, requiring post-production on a PC.
The Lixel K2 detects and removes moving objects during scanning. Immediately after the scan is complete, a color, denoised point cloud preview can be viewed on the smartphone app.
While the quality is said to be about 80% of the finished product after post-production using a PC, the ability to immediately check data on-site, which previously required waiting several hours, is expected to improve the yield rate of the process.
Post-production software receives major update
LixelStudio 4.0, the point cloud processing software released simultaneously with the Lixel K2, significantly improves alignment and leveling accuracy when used with the Lixel K2. It also offers more stable integration of multi-SLAM data, reducing drift and inconsistencies. Post-production can be performed with almost a single click, and it supports cloud sharing and data/VR output for web browsers.
How is XGRIDS’ 3D scanning technology being utilized?
In his keynote speech, Takuya Ii, Director Sales JP of XGRIDS JAPAN Co., Ltd., first introduced the establishment of the Japanese subsidiary in April 2026. He stated that there is growing demand in the Japanese market for digitizing spaces in construction, surveying, plants, infrastructure, video production, and content creation, and that XGRIDS aims to provide 3D data that can be used on-site in an easy-to-use manner for anyone.

The company’s technology involves acquiring real-world data using handheld 3D scanners such as the Lixel K2, processing it at high speed with AI, and then deploying it to applications for various industries. Compared to conventional fixed laser scanners, the combination of handheld scanners and multi-SLAM technology improves acquisition efficiency on-site. In addition to LiDAR SLAM, the company also aims to improve accuracy by combining visual SLAM and RTK information.
According to Ii, the K2 is a lightweight scanner that even a woman can carry, while still achieving stable engineering quality.

In addition, during the speech by JP Xu Fan, Technical Director of XGRIDS JAPAN Co., Ltd., examples of its use overseas were presented.
First, in the field of robotics, the importance of 3D data that accurately reproduces the real world is increasing as a learning foundation for AI and robots to understand the real world. Specifically, it seems that a large amount of virtual simulation learning is required for physical AI being developed by NVIDIA and Tesla.
Therefore, Tsinghua University in China is using XGRIDS’ 3D GS data to conduct simulation learning for robots. They introduced how they have built a cycle that performs thousands of simulations in a digital space with a scale and structure close to reality.

In the fields of video production and XR, South Korean broadcaster SBS is reportedly scanning buildings and landmarks and importing them into Unreal Engine for use as backgrounds in its programs. A key feature is that it enables the creation of non-existent spatial effects within the studio through real-time rendering. In the United States, FOX News has adopted it for the production of documentaries recreating historical locations. One or two staff members measured 12 locations with a handheld scanner, and projected the data onto a large screen, allowing for high-quality video production with a small team and low cost.
Mr. Xu also commented, “XGRIDS aims to provide end-to-end support, not just for scanning space, but also for subsequent simulations and application development. We will continue to improve our product and platform while listening to feedback from people in Japan.”
In a session by Nojiri Tomoya, Manager Sales of the Twin Digital Business, Sales Third Department, ActiveRetech Co., Ltd., he presented case studies based on his experience handling XGRIDS products in the Japanese market for about a year and a half.

Shintani summarized the changes that have occurred in the Japanese market into three points. First, 3D capture can now be performed not only by specialist technicians but also by on-site personnel. Second, scanning and processing, which used to take several days to a week, has been shortened to about 30 minutes to an hour from shooting to processing. Third, wide-area scanning, such as large outdoor sites or entire city blocks, has become easier.

XGRIDS products are initially being adopted in the Japanese market, primarily in the surveying and construction design fields, where they are being used as a replacement for stationary laser scanners and for creating current condition drawings, floor plans, and elevation drawings. In the manufacturing industry, they are being used for factory layout management and digital twin construction. In the XR and gaming fields, they are being used for 3D asset creation for Meta Quest and Apple Vision Pro, as well as for VTuber live streaming. Universities are using them for research, digital archiving, and urban modeling. Furthermore, their applications are expanding to various fields such as railways, plants, power, automobiles, real estate, and television broadcasting.

What they all have in common is that they acquire real-world space as photorealistic 3D data and combine it with CG, game engines, and VR/AR to create professional data and spatial content in less time and at a lower cost than before.
Nojiri commented that with the launch of the new Lixel K2, the Japanese 3D scanning and modeling market will undergo even more significant and dramatic changes from here on out.
“There isn’t actually a single person who doesn’t use this technology at all.”
The discussion, moderated by Nakamura Kaoru, CEO of HoloLab Inc., will cover three topics: “What kind of user tools will XGRIDS products become?”, “Will ‘capturing the entire scene’ become the norm?”, and “How far will the Lixel K2 spread?”. First, Nojiri will answer the first topic.
Nojiri: To be honest, if I had to name an industry where this technology wouldn’t be suitable, I thought it would be companies that provide SaaS that are entirely online and don’t have any physical assets. However, even those companies often have stylish and well-designed offices. XGRIDS can be used to scan those offices with high precision and present them to job seekers as virtual office tours during online company information sessions. Considering this, I think there isn’t actually a single person in the world who doesn’t use this technology at all.
In response to the topic, “It will become commonplace to be able to photograph the entire site,” Ii discusses the advantage that even people without specialized skills can perform a certain level of surveying.
Ii: “Traditionally, even simple surveys required the services of a surveyor. However, surveyors were often fully booked, and long waiting periods would halt construction. But with the XGRIDS scanner, even new employees who don’t have a deep understanding of the subject matter can now conduct basic on-site surveys.”
The question, “How far will the Lixel K2 extend?” includes the topic of how to choose and use the Lixel K2 compared to other XGRIDS products, and Xu will answer this question.
Xu: “First, the Lixel L2 Pro (a high-end model with a long effective range for LiDAR) is mainly used by people who do surveying. It’s not just a handheld scanner; it also has a wide range of modules for mounting on vehicles and drones. The Portal Cam (suitable for 3D GS production) is ideal for creating visually-oriented content. The Lixel K2 is recommended for those who haven’t yet tried a handheld 3D scanner and want both point cloud and 3D GS data.”
A new job function: “3D scanning”
At the event, there was also discussion about the value of the scanner for creating virtual twin content to be deployed on social VR platforms such as VRChat and cluster. Traditionally, backgrounds for VTuber live streams and spatial materials for immersive XR content had to be constructed entirely with CG or required the introduction of multiple scanners and personnel to capture reality. However, with the Lixel K2, a small team can go to the site and acquire both survey-grade point cloud data and high-quality 3DGS at the same time with a short scan.
On the other hand, just because it’s easier to use doesn’t mean that everyone will get the same results. Understanding the scan path, lighting environment, and distance to the object will make a difference in the appearance of the 3D GS and the usability of the point cloud data. Just as the skill of the photographer is behind high-quality photos and videos, spatial scanning also requires the skills of the person performing the scan.
If the Lixel K2 becomes recognized as a field camera that unfolds real-world spaces as XR content, I feel that this scanner could potentially become an entry point for a new profession that handles spatial capture.












