LG Innotek and NVIDIA develop robot vision sensing with Holoscan Sensor Bridge

LG Innotek has announced robot sensing development with NVIDIA, following demonstrations at GTC Berlin on 20 October. The component supplier showed a three-megapixel camera module and a multimodal vision system using NVIDIA's Holoscan Sensor Bridge to connect sensing hardware with AI computing.

Holoscan Sensor Bridge sends sensor information directly to the AI computer while bypassing CPU handling of that data. LG Innotek is applying the connection to cameras and systems that combine different sensor inputs. Its multimodal vision architecture brings together cameras, lidar and radar to support perception of a machine's surroundings.

The collaboration is planned to extend into vision sensing for humanoids and robot hands. LG Innotek says it will develop sensing solutions compatible with NVIDIA's next-generation platforms and present further systems at CES and GTC in 2027. That work combines its optical design and camera-module production with the computing interfaces used by robot developers.

LG Innotek's earlier robotics work includes a May 2025 vision-sensing collaboration with Boston Dynamics and camera-module mass production for a North American robotics customer in Paju from July 2026. It also signed a physical-AI cooperation agreement with TDK in July. The NVIDIA work adds a sensor-to-compute interface to that developing robot-component business.