RoboSense collaborates with NVIDIA to deepen physics AI cooperation, E2 and EM4 officially integrated into HSB

On September 28, 2026, RoboSense announced that its second-generation high-precision ultra wide angle all solid state digital LiDAR E2 and "Thousand Line" ultra long range vehicle grade digital LiDAR EM4 were officially connected to NVIDIA Holoscan Sensor Bridge (hereinafter referred to as "HSB"). As a result, RoboSense has become the world's first laser radar solution partner to achieve mass production delivery in the NVIDIA HSB ecosystem.

Previously, RoboSense had joined the NVIDIA Jetson, DRIVE, and Omniverse ecosystems. The integration of HSB marks a further extension of our cooperation from AI computing and simulation development to the integration of real sensor data, bridging the critical link from physical world perception to AI computing. Both parties will further assist developers in reducing the underlying integration complexity of multi-sensor systems, shortening application development cycles, and promoting the faster scaling of physical AI applications.

Physical AI is accelerating its entry into automotive, robotics, and complex industrial scenarios. Machines need to continuously obtain real-world information and complete perception, computation, decision-making, and execution in a very short period of time. With the rapid growth of sensor types, quantities, and data volumes, issues such as inconsistent interfaces and data protocols, complex development of underlying drivers, and high data transmission delays are becoming increasingly prominent, which constrain the development efficiency and real-time performance of physical AI systems. Developers often need to invest a lot of energy in interface adaptation and underlying development to connect different sensors to the AI computing platform.

NVIDIA HSB is an advanced sensor streaming technology based on Ethernet, aimed at achieving real-time data transmission and simplifying high-speed sensor fusion and actuator integration on NVIDIA edge AI platforms. With its core advantages of ultra-low latency, ease of use, scalability, and security, HSB can help developers shorten sensor driver and system integration cycles, and invest more resources in algorithm, model, and application innovation.

E2 is the first mass-produced product created by RoboSense's self-developed ultra large array SPAD SoC "Peacock" chip. It adopts a fully solid-state scanning architecture and has ultra-high precision, ultra wide field of view, lightweight and compact design, and high robustness. It can be applied to humanoid robots, quadrupedal robots, lawn mowing robots, drones, and space intelligent data acquisition devices. EM4 is the world's first mature mass production ultra thousand line digital laser radar, with a maximum of 2160 lines, 41.04 million points/second imaging capability and a maximum distance of 600 meters, which can provide ultra-high definition 3D perception capability for advanced intelligent driving and auto drive system systems.

After completing platform integration, E2 and EM4 will natively support NVIDIA HSB. Developers can easily integrate the two products into systems built on edge AI platforms such as NVIDIA Jetson and NVIDIA IGX based on HSB standard APIs and open source software. The high-quality 3D data generated by RoboSense digital LiDAR can be transmitted to the AI processing pipeline through Ethernet with low latency; Under hardware configurations that support corresponding acceleration capabilities, data can also be directly written into GPU memory, reducing data handling and CPU usage, providing support for multi-sensor fusion, environmental understanding, real-time inference, and application deployment.

Relying on self-developed full stack chips, digital technology platforms, and mass production capabilities, RoboSense will accelerate the construction of high-quality data entry points for physical AI, continue to promote the integration of innovative categories into the NVIDIA HSB platform, work together with global industry partners to continuously lower the threshold for physical AI development and application, and jointly accelerate the large-scale application of physical AI.

　　Original title: Industry's first! RoboSense collaborates with NVIDIA to deepen physics AI cooperation, E2 and EM4 officially integrated into HSB