Key investment points

3D perception creates a platform for the AI vision industry. The company's main business is the design, R&D, production and sales of 3D visual perception products. The main products include 3D vision sensors, consumer application equipment and industrial application equipment. The company relies on 3D visual perception integrated scientific research and production capabilities and innovation platforms to continuously incubate and expand new 3D visual perception product lines, and has achieved many representative commercial applications in biometrics, robotics, AIoT, and 3D scanning markets. A scientific and reasonable technical system is an important guarantee for the company's technological advancement. The company has built a 3D visual perception technology system with “full-stack technology research and development capability+full-field technology route layout”, which integrates many complex interdisciplinary technologies such as optics, mechanics, electronics, chip design, algorithms, SDK, and firmware development, covering the six major fields of structured light, iTOF, dTOF, binocular, lidar, and industrial 3D measurement in the horizontal span of technology.

AIGC has entered the 3D era, and the AI space intelligence industry integrates development. Spatial intelligence (Spatial Intelligence) is a multi-dimensional concept. It generally refers to an individual's cognitive and reasoning abilities in three-dimensional physical space and four-dimensional space-time, including aspects such as perception, reasoning, and decision-making. The company's 3D vision sensors can collect real 3D data on the human body, objects, and space in real time, giving various terminals the ability to sense, which is a key part of spatial intelligence on the sensing side. Unlike 2D content such as images or videos generated by generative AI tools, content generated in 3D has better control and consistency. As we elevate AI content to the 3D level, imagination and operability will be unified, and the 3D world generated will be interactive, which will accelerate the birth of more application scenarios and requirements, and I believe it will also change the way we currently produce movies, games, simulators, and other digital expressions of the physical world.

NeuralFeels opens up new market space. Meta FAIR team's NeuralFeels technology combines visual and tactile functions. Through multi-modal fusion, RGBD cameras (or 3D vision sensors) are used to enable robotics/dexterous hands to continuously 3D model unknown objects and more accurately estimate the posture and shape of objects during gesture operation. Through a combination of vision and touch, this technology enables more accurate posture tracking and shape reconstruction, providing new possibilities for robots to operate in complex environments. This year, the company integrated a multi-modal large model to launch the latest 2.0 robotic arm solution, which can automatically perform a series of complex tasks based on voice commands, and significantly improved recognition accuracy and operation efficiency. Up to now, the company has implemented a baseline for the application and deployment of two generations of multi-modal large models on robotic arms, and initially explored the adaptation and integration of hardware terminals such as 3D vision sensors, large models, and robotic arms. In the era of artificial intelligence, the company is committed to building a “robot and AI vision industry center”, and attaches great importance to the development and innovation of various AI technologies.

In the future, the company will continue to improve the ability to understand large model robotics/dexterous hands, etc., the ability to generalize in complex or dynamically changing environments, and the self-learning ability of human-computer interaction, optimize solution capabilities according to different application requirements, and accelerate their implementation in various scenarios.

Investment advice

We expect the company to achieve revenue of 0.55/0.85/1.2 billion yuan in 2024/2025/2026, and realized net profit of -0.08/0.01/0.1 billion yuan, respectively. The current stock price corresponding to 2024-2026 PS is 28 times, 18 times, and 13 times, respectively. For the first time, it is covered and given a “buy” rating.

Risk warning

Unprofitable risk, risk of iterative innovation in 3D visual perception technology, risk of core technology leakage, risk of loss of core technical talent, macro-environmental risk.