Is there any tutorial of training a physical robot to walk?


I am helping to design a new university course using nano.
I wonder if there is any tutorial of using machine learning to make a physical robot to walk? (four legs or bipedal using cheap servo)
I assume the following procedure?

  1. Draw the 3D model in Omniverse
  2. Do the reinforcement learning
  3. Feed the model to the physical robot
  4. If it does not work, then how to improve it?

It will be good if there is a video tutorial about this.


You may google it or check this projects sharing at Latest Jetson & Embedded Systems/Jetson Projects topics - NVIDIA Developer Forums to see if can gain some ideas.

I would check this out, is accessible enough to get started, from sentdex:

But yes creating an accurate enough model of the robot and then train a neural network on that model is the way to go.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.