ROSCon China 2026 recently concluded successfully at the Midea Global Innovation Park in Shanghai, bringing together robotics developers, research institutions, and industry leaders to explore the open-source robotics ecosystem and the latest technological advances.
As a Gold Sponsor of this year's event, SLAMTEC made a strong appearance with its core product portfolio, including the Aurora series of spatial perception solutions, the Poseidon omnidirectional mobile robot chassis, and the RPLIDAR series of LiDAR sensors.
As embodied AI continues to advance, robots are no longer limited to performing tasks at fixed locations. Instead, they must perceive space, understand their surroundings, navigate autonomously, and interact with the world in real-world, dynamic, and unstructured environments. Spatial perception and autonomous action are the fundamental building blocks that make these capabilities possible.
Shikai Chen: Perception Understands the World; Action Brings Intelligence to Life
Large language models give robots cognitive capabilities, but spatial perception and autonomous mobility provide the bridge that brings intelligence into the physical world.
At the conference, SLAMTEC CEO Shikai Chen delivered a keynote titled "Building a Multimodal Spatial Perception and Autonomous Action Platform for the Era of Embodied AI."
He highlighted that, in embodied AI applications, LiDAR-based perception is evolving from 2D to 3D, while motion and pose estimation capabilities are advancing toward six degrees of freedom (6DoF). Perception systems must go beyond recognizing spatial geometry to understanding the semantics of their surroundings.
Two years ago, SLAMTEC introduced its vision-only Aurora product line. Powered by deep learning and onboard computing, the system integrates feature extraction and loop closure detection capabilities to address longstanding challenges in traditional visual SLAM (vSLAM), including dramatic lighting changes, dark environments, and multi-floor navigation.
The solution integrates stereo depth estimation on a consumer-grade chip, enabling rendering at over 15 frames per second and real-time semantic recognition. The entire device is approximately palm-sized, consumes as little as 5 watts of power, and requires no external computing hardware.
Real-World Performance
A device consuming just 5 watts can map all floors of Shenzhen Bao'an International Airport's Terminal 3 in a single mapping run. The vision-only solution achieves a repeat localization accuracy of 3 cm. With tightly coupled RTK/GNSS integration, positioning error with a single antenna can be kept within 0.5 m, supporting outdoor mapping across areas of nearly 5,000 m².
Using a single sensor, the solution can generate dense TSDF maps with both texture and semantic information in real time. An integrated 3D navigation version is also coming soon. SLAMTEC further previewed a 3D LiDAR-vision fusion solution capable of achieving 2 cm positioning accuracy in environments without light.
For mobile robotics, SLAMTEC launched the Poseidon four-wheel omnidirectional chassis this year. Powered by the third-generation SLAMWARE NX, it achieves a positioning accuracy of 10 mm and provides an open ROS 2 interface.
On the Show Floor | Three Flagship Product Lines on Display
At ROSCon China 2026, SLAMTEC showcased three core product lines at its booth, providing ROS developers with ready-to-use, easy-to-integrate foundational capabilities for robotics development.
Numerous ROS developers and industry partners stopped by to explore the solutions, exchange ideas, and experience SLAMTEC's underlying perception and mobility technologies firsthand.
✅ Multimodal Spatial Perception: Aurora Series Spatial Perception Solutions
By fusing data from LiDAR, vision, and IMU sensors, and powered by SLAMTEC's proprietary AI-vSLAM technology, the
Aurora series enables robots to build 3D spatial models and understand scene semantics, delivering precise spatial perception in complex and dynamic indoor and outdoor environments.
Aurora S Lite: Lightweight All-in-One AI Spatial Perception System
Aurora S Lite integrates SLAMTEC's proprietary AI-vSLAM, stereo depth estimation, semantic recognition, and dedicated computing hardware. It delivers high-precision 3D localization, real-time dense 3D mapping, and environmental perception outputs without requiring additional computing hardware or in-house algorithm development.
Designed for quadruped robots, outdoor robotic lawn mowers, humanoid robots, and autonomous mobile robots (AMRs), Aurora S Lite simplifies integration and accelerates the deployment of intelligent robotic applications.
✅ Embodied AI Mobility Platform: Poseidon Omnidirectional Mobile Robot Chassis
The modular chassis lowers the barrier to robot development, providing the AI systems running on top with a highly maneuverable, high-precision mobility platform. Developers can focus on algorithm development without having to build low-level motion control systems from scratch.
✅ Core Sensors for Localization and Navigation: RPLIDAR Series LiDAR Sensors
Designed to support the evolution of embodied AI, the RPLIDAR series enables LiDAR-vision fusion for 3D SLAM, supporting online map correction and relocalization. This helps maintain continuous localization in changing environments and addresses common challenges such as localization drift and tracking loss in real-world applications.
Advancing Together with the ROS Ecosystem, Empowering Developers
The ROS ecosystem serves as a proving ground for robotics innovation. Close collaboration between foundational hardware and open-source software is essential to accelerating the evolution of embodied AI.
SLAMTEC has long been committed to ROS compatibility, continuously improving its SDKs, drivers, and developer documentation. These efforts enable LiDAR sensors, Aurora perception systems, and Poseidon mobile robot chassis to integrate seamlessly into ROS-based environments, reducing the time and cost of low-level development for robotics developers.
At ROSCon China 2026, SLAMTEC participated as a Gold Sponsor, engaging directly with developers, listening to real-world development challenges, and sharing insights into emerging technology roadmaps.
Looking ahead, SLAMTEC will continue to focus on multimodal spatial perception and autonomous mobility platforms, continuously advancing its products and working alongside the ROS community to bring embodied AI robots into real-world applications.
Gold Sponsor of ROSCon China 2026 | SLAMTEC CEO Shikai Chen: Perception Understands the World, Action Brings Intelligence to Life
Gold Sponsor of ROSCon China 2026 | SLAMTEC CEO Shikai Chen: Perception Understands the World, Action Brings Intelligence to Life
