B.1: Isaac Sim

• Short Description: Nvidia Isaac sim can be used for domain randomisation, which is a technique for training AI algorithms to be robust and generalise well to new environments. In domain randomisation, the training data is generated by simulating various variations of the environment, such as changes in lighting conditions, object appearances, and other factors. This can help the AI model learn to be robust to these variations and perform well in a range of different environments. Isaac sim provides tools for easily creating and manipulating these variations in the simulated environment, making it a useful tool for domain randomisation. Additionally, it provides tools for generating realistic sensor data from the simulated environment, which can be used to train and evaluate AI algorithms.

• Reference, URL: https://developer.nvidia.com/isaac-sim

• Applicable Business Category: WEEE & Battery

• Application in relevant Projects/Initiatives: N.A.

• Asset Type: Other, Robotics simulator for synthetic data and domain randomisation

• AI Breadth: ML, Computer Vision

• Learning Ability: Reinforcement Learning, Synthetic Data Generation

• Related technologies: Python API, Universal Scene Description

• Applicable Research Area: Collaborative AI

• Applicable Technical Category: Machine learning

• License Information: Proprietary license

• Related to circularity and sustainability: Not directly

• Audience: Robotics researchers/practitioners, Developers

Go back to the AI portfolio of reference implementations