Chenjing Technology officially launched a new brand for embodied intelligence and a full-stack technology matrix, breaking through the boundaries of spatial intelligence technology.

This release encompasses the Insight fully autonomous spatial intelligent camera, the TinyNav high-performance navigation algorithm library, and the RoboSpatial spatial editing toolchain. It marks significant breakthroughs in the fields of key perception hardware, underlying algorithms, and application toolchains for embodied intelligence, providing solid support for promoting robots to become the next physical task infrastructure after automobiles and mobile phones.

I. Defining the "Spatial Intelligence" Foundation for Robots in the Complex Physical World

Currently, artificial intelligence is accelerating its advancement from the digital world to the physical world. As a typical representative of physical AI, embodied intelligence serves as a bridge connecting the digital and physical entities and is also a key means to empower the transformation and upgrading of the manufacturing industry. However, how to enable robots to have high-precision perception and decision-making capabilities in the complex and changeable real physical environment and achieve large-scale implementation from laboratories to workshops and households remains a common challenge faced by the industry.

The "LooperRobotics" brand system launched by Chenjing Technology this time aims to create a standardized perception base for driving general robots through independent innovation in software and hardware integration.

The Insight fully autonomous spatial intelligent camera launched at the press conference is a major highlight of this release.

As the core data entry point for embodied intelligence, the Insight fully autonomous spatial intelligent camera has reshaped the industry's perception boundaries with technological leaps in four dimensions:

Firstly, the closed-loop of computing power at the edge: It is equipped with a 10TOPS NPU computing power, enabling the cooperation between the SLAM algorithm and AI binocular ranging at the camera end, ensuring the data processing efficiency at the source of perception.

Secondly, the full-range vertical field of view: It adopts a 188° ultra-wide-angle lens and a vertically enhanced layout to eliminate perception blind spots and achieve panoramic coverage of information.

Thirdly, millisecond-level real-time response: Relying on the edge computing architecture, AI inference can operate in milliseconds, endowing robots with a rapid obstacle avoidance response similar to biological instincts.

Fourthly, industrial-grade environmental robustness: It has a 24g anti-seismic capability and strong light suppression, ensuring the accuracy of the robot's pose output under extreme working conditions such as bumpy roads and complex lighting.

This high-performance and low-power integrated solution provides a standardized perception source for embodied intelligence.

As the "navigation soul" of Looper Robotics, TinyNav endows robots with robust local perception and autonomous positioning capabilities, supporting safe exploration, real-time mapping, and long-term memory management in unknown environments, ensuring the continuous and stable operation of robots under unsupervised conditions. The core of its technology lies in a generative world model for robot training. With a small amount of real-scene data, it can generate an infinite amount of high-fidelity synthetic data, fundamentally improving the generalization ability and robustness of the robot's decision-making system.

The RoboSpatial spatial perception software platform has strong software and hardware compatibility. Through modular design, it creates an off-the-shelf software and hardware integrated solution, allowing robot manufacturers to skip the complex underlying development process and quickly empower the core capabilities. This innovation not only fundamentally solves the industry problems of high development threshold, long cycle, and difficult adaptation in traditional robot development but also realizes the structured storage and intelligent call of environmental information by constructing a "spatial knowledge base" exclusive to robots.

II. The Innovative Model of "Spatial AI as a Service" Demonstrates Broad Application Potential

Hu Wen, the co-founder and CEO of Chenjing Technology, said that in the framework of physical AI, intelligent agents need to have a high-dimensional reconstruction ability of the physical world. Chenjing Technology has built a technological closed-loop centered around the Insight camera as the perception entry point, the RoboSpatial platform as the spatio-temporal base, and the TinyNav engine as the cognitive core, and put forward the new concept of "Spatial AI as a Service". Through the coupling of software and hardware, this matrix transforms fragmented physical signals into structured spatial intelligence, endowing robots with the "spatial intuition" to understand physical logic.

If automobiles are mobile infrastructure and mobile phones are information infrastructure, then robots are expected to become the future "physical task infrastructure". The software and hardware integrated "spatial intuition" solution and the innovative model of "Spatial AI as a Service" of Chenjing Technology will provide standardized infrastructure support for the upstream and downstream of the industrial chain.

At the press conference, Chenjing Technology demonstrated the in-depth cooperation results with leading enterprises in the industry such as Unitree and Diguaji. After being equipped with the full-stack spatial intelligent solution, humanoid robots and quadruped robots demonstrated excellent autonomous navigation and interaction capabilities, verifying the broad application potential of this technology in the fields of industrial inspection, logistics distribution, and public services.

The release of this full-stack solution not only fills the technological gap in the relevant fields but also reflects the positive actions of Chinese technology enterprises in promoting the collaborative innovation of the industrial chain.

The matrix products launched by Chenjing Technology this time have solved the three core pain points of "seeing clearly, moving smoothly, and using effectively", and are expected to promote the accelerated integration of embodied intelligence into the real economy and become an important force in promoting the digital and intelligent transformation of the economic society.

For more information about the products at the press conference, please visit the official website:

LooperRobotics official website: https://www.looper-robotics.com

TinyNav high-performance navigation algorithm library: https://github.com/UniflexAI/tinynav