Hands-on
Hard-skill tutorials with a check every three steps
Some skills cannot be learned just by 'understanding' them; someone has to put every command, every expected result, and every failure cause in front of you. This tutorial set is written to that standard: every step has a verifiable result, and if it doesn't run, that step is the one that is wrong — no guessing. Tutorial 0 is the zero-experience starting point, and the three after it are ordered by ever more expensive hardware and higher entry barriers.
The tutorials were written against the versions current in November 2026. Version numbers and script paths in commands change upstream; when a path does not exist, first ls the actual directory of your version, then follow the approach described here.
From zero to your first Python script: terminal lesson one
For anyone who has never installed an environment or opened a terminal, this is the first step up. No Linux, no GPU, nothing to buy throughout — just an ordinary laptop.
zero-experience · 1–2 h
Build a six-axis arm in Gazebo from scratch and make it obey you
URDF → gz-sim → gz_ros2_control → MoveIt 2, one chain walked end to end, about 6 hours
basic Linux commands required; a little ROS 2 exposure is enough · 6–8 h
Train a grasping policy from scratch: MuJoCo + SAC + HER
First get the arm to reach the target, then learn to pick it up. Training runs with a single command, about 5 hours
Knows Python and has installed pip packages; RL concepts can be zero · 5–6 h
Isaac Lab parallel training: 4096 environments at once
From environment setup to exporting a deployable policy, about 4 hours (excluding training time)
Comfortable with Linux and Python; has an RTX GPU (VRAM ≥ 12 GB) · 4 h + training
How to use the tutorials
Every step writes three things down: what to type, what you should see, and the most likely cause when you do not see it. When stuck, do not guess and do not dig through search results — check the «stuck» table. After one tutorial runs, the derivations track explains why it works: start here.