Nvidia’s robot simulator now includes human characters

Nvidia has announced updates to its robotic simulation tools that will allow companies to build and test virtual robots in a variety of realistic environments and operating conditions, all securely in a cloud environment.

Revealed at CES, enhancements to Isaac Sim, Nvidia’s robotic simulation toolkit, include the ability to simulate the behavior of people in environments such as warehouses or manufacturing plants, to help collaborative robots (cobots) or autonomous mobile robots understand and identify common behaviors and potential obstacles they will face in the real world.

According to Nvidia, improvements to the robotics platform will accelerate the development and deployment of autonomous robots by making it easier to develop, validate and deploy the artificial intelligence (AI) that powers them, ensuring they can operate successfully and safely in various environments.

“Simulation is the essential technology that will enable the development of the complex software systems that will power the next wave of smarter, more autonomous robots. In simulation, virtual robots have a proving ground for their complex software stacks and multitude of AI models,” said Gerard Andrews, senior product manager at Nvidia.

Accuracy and safety

By adding simulations of human behavior and interaction in environments, such as picking up and moving objects, pushing carts and moving to new locations, it is possible to test how adding robots to the environment might play out, without endangering people.

Common events – such as people interacting with industrial environments or moving around a warehouse – and less common events – such as unexpected emergencies and other scenarios – can be simulated to help build robots that can react in different ways. .

To this end, using Nvidia RTX technology, Isaac Sim has improved sensor support, allowing him to render physically accurate data from sensors in real-time, including ray tracing to provide more accurate sensor data under various lighting conditions or in response to reflective materials.

This allows simulated worlds to be based on physically accurate sensor models, minimizing differences between the simulation and the real environment, to ensure robots are trained as accurately as possible.

Virtual robots accessible on the cloud

The new version of Isaac Sim also provides many new simulation-ready 3D assets, including warehouse parts and popular robots, so developers and users can start building quickly.

Powered by Nvidia Omniverse, the company’s platform for building and running applications from the metaverse, Isaac Sim is accessible via the cloud, giving teams working on robotics projects the ability to collaborate with accessibility, agility, and scalability to test and train virtual robots.

“With access to the cloud and its vast set of physically accurate and photorealistic simulation capabilities, Isaac Sim is poised to establish new methodologies for developing intelligent robots,” said Gerard Andrews.

Source: ZDNet.com

Leave a Comment