LG Innotek and NVIDIA Unveil Humanoid Robot Sensors
The companies showcased advanced vision sensing solutions for humanoid robots at the GTC conference in Berlin.
Updated on Oct. 22, 2026 in Robotics
Live Poll
Are you optimistic that the rapid development of physical AI robots will improve society?
LG Innotek and NVIDIA have introduced new humanoid robot sensing solutions at GTC Berlin. This collaboration focuses on integrating physical AI technologies to improve robot perception.
Why it matters
The partnership aims to enhance how robots recognize objects and spaces by minimizing processing delays. This technology is intended to improve performance for both humanoid robots and autonomous vehicles.
The new sensing system features a 3-million-pixel camera module and integrates NVIDIA's Holoscan Sensor Bridge, which transfers data directly to GPU memory to bypass CPU-related bottlenecks.
The players
LG Innotek
An electronics component manufacturer that specializes in optical solutions and automotive parts.
NVIDIA
A multinational technology company known for designing graphics processing units and AI computing platforms.
Boston Dynamics
An engineering and robotics design company that focuses on the development of advanced mobile robots.
The details
The multimodal vision system processes data from cameras, LiDAR, and radar simultaneously to achieve higher object recognition accuracy. The technology is designed to facilitate real-time decision-making for autonomous systems.
Timeline
May 2025: LG Innotek began a collaboration with Boston Dynamics on vision systems.
October 20, 2026: The GTC Berlin conference opened.
October 22, 2026: The GTC Berlin conference closed.
2027: The companies plan to exhibit sensing solutions at CES in Las Vegas.
The Tech Race
The shift toward using the NVIDIA Holoscan Sensor Bridge represents a move away from CPU-reliant processing in favor of high-speed GPU direct-memory access. This advancement pits manufacturers against one another in a race to build more responsive and accurate physical AI.
Users can expect future robots to navigate spaces more reliably and safely due to reduced processing delays. These improvements in vision sensing will likely influence the accessibility and capability of future service robots in commercial environments.
The takeaway
The integration of multi-sensor data processing signals a critical leap in how humanoid robots perceive the world around them. Developers and manufacturers are increasingly prioritizing direct GPU data streams to solve latency issues in real-time robotic motion.
Further reading
Learn more about the latest hardware developments in the Robotics sector.
Source note: This article includes information reported by 조선일보.
Live Poll
Are you optimistic that the rapid development of physical AI robots will improve society?
