LG Innotek is collaborating with NVIDIA to develop sensing solutions for physical AI systems, including robots and autonomous machines. At NVIDIA’s GTC event in Berlin on October 20, 2026, the South Korean components manufacturer unveiled a 3-megapixel camera module and a Multi-modal Vision System designed to support NVIDIA Holoscan Sensor Bridge (HSB) technology. The integration aims to streamline the transfer of sensor data to GPU-based computing systems, helping AI applications process real-world information with lower latency and support faster machine decision-making.
A sensor-to-GPU pipeline for physical AI
The collaboration targets a technical bottleneck in physical AI: moving large volumes of real-world sensor data into computing systems quickly enough for machines to interpret their surroundings and respond.
NVIDIA’s Holoscan Sensor Bridge is designed to stream sensor data over Ethernet to GPU memory, using FPGA-based interfaces and compatible networking hardware. This can reduce CPU involvement in the data path and support low-latency processing pipelines. NVIDIA documents the technology for applications involving cameras, radar, lidar and other sensor types.
LG Innotek’s newly unveiled portfolio includes a 3-megapixel camera module and a Multi-modal Vision System (MVS) that combines multiple sensing inputs. The company’s broader sensing expertise spans cameras, lidar and radar, technologies used in autonomous driving and robotics.
The distinction matters because a robot rarely operates using a camera feed alone. A machine may need to combine visual information with depth measurements, distance estimates and other sensor readings to understand its environment. A standardized, low-latency data path can help developers bring those inputs into a common AI processing pipeline.
However, faster data transfer is only one component of real-time intelligence. The performance of a complete system also depends on sensor quality, synchronization, GPU processing, AI models and the control software that translates predictions into physical actions.
Why NVIDIA’s Holoscan platform matters
NVIDIA is building an ecosystem around the computing infrastructure needed to develop physical AI systems. Its Holoscan platform provides tools for streaming sensor data, processing signals and running AI applications across compatible edge and cloud environments.
Holoscan Sensor Bridge is part of that stack. Rather than requiring every sensor manufacturer to develop an entirely separate integration approach, the technology offers a standardized interface for transferring data into GPU-based processing pipelines.
For robotics developers, this could reduce some of the engineering work involved in connecting sensors to AI computers. It may also help simplify the integration of multiple sensor types, an important consideration for autonomous machines operating in complex environments.
The opportunity extends beyond humanoid robots. Industrial inspection systems, autonomous vehicles and medical imaging equipment all rely on timely processing of sensor information. NVIDIA identifies these types of streaming and real-time applications among the use cases for its Holoscan ecosystem.
LG Innotek’s role is to provide sensing hardware that can fit into this broader architecture. Its combination of optical engineering, camera modules and multimodal sensing capabilities could give robotics manufacturers another option for building systems around NVIDIA’s computing platform.
Physical AI is creating a new hardware market
The announcement comes as technology companies increasingly focus on AI systems that interact with the physical world rather than simply generating text, images or software.
According to IDC, the global physical AI market is projected to exceed $40 billion by 2029. IDC’s market framing includes AI embedded in robots and machines that sense their surroundings, make decisions and act in real-world environments.
That market opportunity is drawing investment across the hardware stack, including processors, sensors, connectivity components, robotics platforms and AI software. NVIDIA supplies computing platforms and development tools, while companies such as LG Innotek are positioning sensing components as part of the infrastructure needed to bring physical AI into production.
LG Innotek is also pursuing partnerships across adjacent markets. The company has previously announced a strategic partnership with lidar technology provider Aeva and collaborations with autonomous-driving software and mobility companies. It has also described work on robotic components and camera modules for robotics customers.
These efforts suggest that LG Innotek is seeking a broader role in physical AI systems, rather than limiting its business to individual camera components. Its stated ambition is to combine sensing, communication modules and semiconductor substrates into a wider solution portfolio.
Humanoid robotics and the next development phase
LG Innotek and NVIDIA plan to expand their work into vision sensing and robotic hands for humanoid robots, with the goal of supporting integration with NVIDIA’s next-generation platform. The companies also plan to showcase further solutions at CES and GTC in 2027.
That roadmap points to a move from individual sensing components toward more integrated robotics subsystems. For humanoid machines, vision can help identify objects and navigate environments, while robotic hands require sensing and control capabilities to manipulate objects with precision.
Still, the announcement does not establish commercial deployment volumes, pricing or independently measured improvements in robot response times. Those details will be important in assessing whether the collaboration translates into scalable products.
For robotics developers, the practical test will be whether the combined sensing and computing architecture makes systems easier to integrate, more responsive and reliable under real operating conditions. Sensor compatibility, latency under load, power consumption and support for safety-critical applications will all influence adoption.
LG Innotek’s collaboration with NVIDIA reflects a broader shift in AI infrastructure: the sensor, data pipeline, processor and model increasingly need to be designed as parts of the same system. If the companies can deliver reliable, interoperable solutions, the partnership could help LG Innotek expand from component supply into a more integrated role in the emerging physical AI ecosystem.
Market Landscape
Physical AI is creating demand for tightly integrated sensing, computing and software systems that allow machines to interpret and respond to their environments. NVIDIA’s Holoscan platform addresses the sensor-to-GPU data pipeline, while LG Innotek is developing camera modules and multimodal sensing systems compatible with that architecture.
The market includes robotics manufacturers, autonomous-driving developers, industrial automation providers and suppliers of AI computing infrastructure. IDC projects that the physical AI market will exceed $40 billion by 2029.
LG Innotek’s opportunity is to expand beyond component manufacturing into integrated sensing solutions for NVIDIA-based robotics platforms. However, adoption will depend on system-level latency, sensor interoperability, reliability, production readiness and the ability to meet the cost and performance requirements of commercial robotics deployments.
Top Insights
- NVIDIA integration: LG Innotek unveiled a 3-megapixel camera module and multimodal vision system designed to support NVIDIA Holoscan Sensor Bridge technology.
- Faster sensor processing: HSB enables compatible sensor data to reach GPU memory with reduced CPU involvement, supporting low-latency AI processing pipelines.
- Multimodal sensing: Combining camera, lidar and radar inputs can help physical AI systems interpret complex environments for robotics and autonomous machines.
- Humanoid robotics expansion: The companies plan to explore vision sensing and robotic hands for humanoid robots, with further demonstrations targeted for 2027.
- Broader market opportunity: The collaboration positions LG Innotek to pursue integrated sensing solutions as physical AI expands across industrial automation, robotics and autonomous driving.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
