Tutorials

These tutorials walk through EmbodiChain from simulation fundamentals to task authoring, environment workflows, and learning. Code-oriented tutorials include runnable scripts; deployment tutorials include complete configuration examples.

Suggested learning path

Work through the categories in order when you are new to EmbodiChain:

1. Simulation and Control — create scenes, control robots, use Atomic Skills, and author a Task Program in Python.

2. Environments and Data — build Gym environments, deploy a configured Task Program, and generate demonstration data.

3. Learning — train reinforcement-learning policies.

To add a new robot after completing the core tutorials, continue with Adding a New Robot.

• Simulation and Control
  • Creating a simulation scene
  • Visualizing a Point Cloud
  • Creating a soft-body simulation
  • Creating a cloth simulation
  • Rigid object group tutorial
  • Rigid constraint tutorial
  • Loading an Articulation
  • Simulating a Robot
  • Simulating a Camera Sensor
  • Create a solver
  • Motion Generator
  • Opening a Drawer with Contact-Rich Motion
  • Atomic actions
  • Authoring a Task Program in Python
  • Interactive Robot Control with Gizmo
• Environments and Data
  • Creating a Basic Environment
  • Creating a Modular Environment
  • Configure and Run an Embodied Task Program
  • Expert Data Generation
• Learning
  • Reinforcement Learning Training