NVIDIA announced the release of NVIDIA Omniverse Cloud Sensor RTX™, a suite of microservices designed to deliver physically accurate sensor simulation and accelerate the development of autonomous machines. The announcement was made at the Computer Vision and Pattern Recognition (CVPR) conference, according to the NVIDIA newsroom.
A revolution in sensor simulation
Omniverse Cloud Sensor RTX aims to revolutionize the sensor industry, a rapidly growing, multi-billion dollar sector. These microservices enable developers to extensively test sensor recognition and associated AI software in realistic virtual environments before deploying them in the real world. This approach not only increases safety, but also reduces the time and costs associated with physical prototyping.
“Developing safe and reliable autonomous machines powered by generative physics AI requires training and testing in virtual, physics-based worlds,” said Rev Lebaredian, vice president of Omniverse and Simulation Technologies at NVIDIA. “NVIDIA Omniverse Cloud Sensor RTX microservices enable developers to easily build large-scale digital twins of factories, cities or even the planet, accelerating the next wave of AI.”
Enhancement of large-scale simulation
Built on the OpenUSD framework and powered by NVIDIA RTX™ ray tracing and neural rendering technology, Omniverse Cloud Sensor RTX combines real-world and synthetic data from a variety of sources, including video, cameras, radar, and LIDAR, to create highly accurate simulated environments. These capabilities are especially useful in scenarios where real-world data is limited.
Microservices can simulate a wide range of activities, such as checking if a robotic arm is functioning correctly, validating the operation of a baggage claim at an airport, detecting if a tree branch is blocking the road, or ensuring that a conveyor belt in a factory is running smoothly.
Research Milestones and Real-World Applications
The announcement comes just as NVIDIA won the CVPR Autonomous Grand Challenge for End-to-End Driving at Scale. The winning workflow can be replicated using Omniverse Cloud Sensor RTX, enabling autonomous vehicle (AV) simulation developers to test self-driving scenarios in a high-fidelity simulated environment before deployment in the real world.
Additionally, companies such as CARLA, Foretellix, and MathWorks will have first access to the Omniverse Cloud Sensor RTX for AV development, enabling sensor manufacturers to validate and integrate the digital twin of their sensor in a virtual environment, dramatically reducing the time required for physical prototyping.
Availability and Early Access
NVIDIA is now accepting registrations for early access to its Omniverse Cloud Sensor RTX, available later this year, which will enable groundbreaking advancements in the development of autonomous machines and AI-driven applications.
Image credit: Shutterstock
