Accelerating physical AI development with NVIDIA
Integrating ST’s comprehensive portfolio with NVIDIA’s Holoscan Sensor Bridge (HSB) and Isaac Sim ecosystem for robotics
ST, in collaboration with NVIDIA, is empowering developers by simplifying and accelerating the creation of physical AI solutions. Through strengthened collaboration and tighter ecosystem integration, we are making it easier to design, train, and deploy humanoid robots and other physical AI systems.
Integration with HSB + high-fidelity modeling = faster time to market
As part of the collaboration, ST is working with NVIDIA to integrate its comprehensive portfolio with NVIDIA Holoscan Sensor Bridge. This includes HSB running on STM32 microcontrollers, inertial measurement units (IMUs), image sensors, time-of-flight (ToF devices), and motor-control solutions.
From a software perspective, high-fidelity NVIDIA Isaac models of ST components are being integrated into both companies’ robotics ecosystem.
Here are the first two tangible results of this engagement:
HW – HSB depth camera
The camera integrates a comprehensive set of ST sensors: two VB1940 RGBIR 5.1 MP image sensors for stereoscopic depth, a VL53L9CX ToF LiDAR module for standalone 3D depth mapping, and an LSM6DSV16X IMU for accurate positioning. Leopard's depth camera for robots is plug-and-play compatible with the Holoscan SDK via a 10G Ethernet interface. The camera can generate accurate 3D streams in low-power mode, with the option to engage host processing for higher-resolution depth maps and visualization. It is designed to support a broad range of robotic applications across physical AI platforms.
This enables developers to train vision, balance, gait, and motion control algorithms in simulation with confidence that these behaviors will transfer reliably to real hardware.
SW – ST high-fidelity MEMS sensor modeling
ST provides a high-fidelity model of its ASM330LHH IMU within NVIDIA Isaac Simulator, built from real device measurements including component-specific noise characteristics. The model provides a number of benefits for developers:
- Enables accurate virtual testing through direct integration with the simulator.
- Ensures balance and motion‑control algorithms behave consistently between simulation and real hardware.
- Uses the digital twin in Isaac Sim to let teams iterate on gait, balance, and navigation, reducing R&D effort.
Provides low‑latency sensor simulation to support the fast, stable responses required for complex humanoid movements.
- "ST is well engaged within the robotics community, providing robust support and a well-established ecosystem. Our collaboration with NVIDIA aims to unleash the next wave of cutting-edge robotics innovation with developer and customer experience streamlined at every step, from the inception of AI algorithms to the seamless integration of sensors and actuators. This will accelerate the evolution of sophisticated AI-driven physical platforms.”
- “Accelerating the development of next-generation autonomous systems requires high-fidelity simulation and seamless hardware integration to bridge the gap between virtual training and real-world deployment. The integration of STMicroelectronics’ sensor and actuator technologies with NVIDIA Isaac Sim, Holoscan Sensor Bridge and Jetson platforms provides developers with a unified foundation to build, simulate and deploy physical AI at scale.”
ST’s portfolio for robotics
ST offers one of the broadest semiconductor portfolios in the industry of components applicable to robotics - covering every major subsystem from edge AI processing and motor control to distributed sensing, power management, and secure connectivity. This integration depth means fewer vendors, faster system-level optimization, and a single partner across every major subsystem.
ST sensors for humanoid robotics
Humanoid robots depend on a wide range of sensing technologies to understand their surroundings, interpret movement, and interact naturally with people and objects. ST offers inertial, imaging, and depth sensing and is collaborating to create a developer-friendly ecosystem that supports robust perception and reliable sim-to-real performance.
Beyond providing accurate and rugged sensor hardware integrated with the Holoscan Sensor Bridge NVIDIA architecture, ST enhances the developer workflow through high-fidelity sensor models integrated into platforms like NVIDIA Isaac, helping ensure that behaviors trained in simulation translate seamlessly to physical robots.
ST’s portfolio and roadmap of AI enabled sensors further empowers edge intelligence, enabling faster reaction times and more fluid, lifelike motion across all parts of a humanoid robot.