Sensor technology is a multi-billion dollar growth industry that provides autonomous vehicles, humanoids, industrial manipulators, mobile robots, smart spaces and more with the data they need to understand the physical world and make informed decisions.

NVIDIA announced the NVIDIA Omniverse Cloud Sensor RTX on the 2nd.

Omniverse Cloud Sensor RTX is a set of microservices that can accelerate the development of fully autonomous machines of all kinds through physically accurate sensor simulation.

This enables developers to test sensor perception and related AI software at scale in physically accurate and realistic virtual environments before actual deployment. This allows developers to save time and money and improve security.

“Developing safe, reliable, autonomous machines powered by generative physics AI requires training and testing in a physics-based virtual world,” said Rev Lebaredian, vice president of Omniverse and simulation technologies at NVIDIA. “With NVIDIA Omniverse Cloud Sensor RTX microservices, developers can easily build large-scale digital twins of factories, cities — even the planet — accelerating the next wave of AI.”

■ Large-scale simulation support

Omniverse Cloud Sensor RTX is based on the OpenUSD framework and powered by NVIDIA RTX ray tracing and neural rendering technology. It allows real-world data from video, cameras, radar, and lidar to be combined with synthetic data to rapidly create simulation environments.

Microservices enable simulation of a wide range of activities even in scenarios where real-world data is limited. For example, it is possible to simulate whether a robotic arm is operating properly, whether an airport baggage conveyor belt is working, whether a tree branch is blocking a road, whether a factory conveyor belt is moving, and whether there are robots or people nearby.

■ Actual deployment case

The announcement of Omniverse Cloud Sensor RTX coincides with NVIDIA taking first place in the Autonomous Grand Challenge for Large-Scale End-to-End Driving at the Computer Vision and Pattern Recognition (CVPR) conference.

Omniverse Cloud Sensor RTX enables autonomous vehicle (AV) simulation developers to test autonomous driving scenarios in physically accurate environments before deploying AVs in the real world.

NVIDIA emphasized that this allows NVIDIA researchers to implement their successful workflow in a high-fidelity simulation environment.

■ Sensor digital twin verification and integration

Foretellix and MathWorks are among the first software developers to whom NVIDIA is providing Omniverse Cloud Sensor RTX access for AV development.

Additionally, Omniverse Cloud Sensor RTX enables sensor manufacturers to validate and integrate digital twins of their sensors in a virtual environment, reducing the time required for physical prototyping.