10 Technologies for Autonomous Vehicle, Robotics Developers From NVIDIA’s GTC 2021

Click arrow or on image to advance.
9 of 10

NVDIA Modulus learns the laws of physics

One of the challenges in using simulation for training robots and other engineering modeling is realistic physics. NVIDIA this week said developers can use its Modulus framework to build models that can learn and obey the laws of physics.

Modulus, previously known as SimNet, includes a data preparation module to help manage observed or simulated data. It also accounts for the geometry of the systems modeled and the parameters of the space represented by the input geometry, said NVIDIA.

The GPU-accelerated toolkit offers rapid turnaround complementing traditional analysis, enabling faster insights. Modulus allows users to explore different configurations and scenarios of a system by assessing the impact of changing its parameters.

Modulus is designed to be customizable and easy to adopt, said NVIDIA. It offers application programming interfaces (APIs) for implementing new physics and geometry. The free download is intended to help users get started with AI-driven digital-twin applications.

Return to article.

1. Jetson AGX Orin adds processing horsepower

2. NVIDIA adds acceleration libraries, SDKs

3. Omniverse Replicator for Isaac Sim, robot manipulation

4. NVIDIA DRIVE Sim for autonomous vehicles

5. ISAAC Sim includes ROS support

6. NVIDIA DRIVE Orin supports all kinds of vehicles

7. DRIVE and DeepMap enable crowdsourced maps

8. Orin supports Clara Holoscan for surgical robots

9. NVDIA Modulus learns the laws of physics

10. NVIDIA Omniverse Avatar enables AI assistants

Robot Technologies