Anima Anandkumar of Caltech and NVIDIA. This talk was given on April 1, 2022. Autonomous robots need to be efficient and agile, and be able to handle a wide range of tasks and environmental conditions. This requires the ability to learn good representations of domains and tasks using a variety of sources such as demonstrations and simulations. Representation learning for robotic tasks needs to be generalizable and robust. I will describe some key ingredients to enable this: (1) robust self-supervised learning (2) uncertainty awareness (3) compositionality. We utilize NVIDIA Isaac for GPU-accelerated robot learning at scale on a variety of tasks and domains. 00:48 Generlizable Learning for Robotics 01:53 Trinity of Generalizable AI 05:08 Physical World if Continuous 07:08 Motivating Problem in Robotics 08:35 State estimation through PDE observer 09:59 Grid-free learning for continuous phenomena 12:32 Neural Operator 15:02 Fourier Transform for global convolution 16:06 FNO: Fourier Neural Operator 18:34 First ML method to solve fluid flow 25:42 Nvidia Modulus 29:08 Operational Space Control (OSC) 38:54 Reducing Supervision and enhancing robustness 47:06 Conclusion Learn more about Stanford Online’s Robotics Program and courses here: