Interactive textbook · embodied AI, from scratch

Read robotics like a book.Run every page like a lab.

Minimal, real, readable code — one runnable file per chapter, from a bare simulation loop to a policy you train, export, and drive in your browser. No framework to excavate, no black boxes. The code is the product, and it all runs on a free Colab T4 or a CPU laptop.

Start from the first page0.0 · quickstart.pyJump to the flagship chapter →
overall0%
start the first chapter · Your First Robot Policy in Twenty Minutes

Nothing completed yet — read a chapter, or commit a prediction in an exercise, and it shows up here. No account needed; this lives only in your browser.

  • Phase 0 · Foundations0/6 · 0%
  • Phase 1 · Imitation0/9 · 0%
  • Phase 2 · Reinforcement0/8 · 0%
  • Phase 3 · Advanced0/9 · 0%
  • Phase 4 · Capstone0/3 · 0%
  • Phase 5 · Practitioner0/8 · 0%

See it work

live · P2
PushT behavior-cloning policy — covariate-shift toyThe demonstrated-coverage region (dashed) is the set of block positions the training demos actually visited, derived from real scripted-expert rollouts. Inside it the policy pushes the block onto the target; dragged outside it, the policy makes small, confident, wrong movements.where the demos pushed the blocktargetdrag me out of distribution →
drag the block (or arrow-keys when focused) · poster reads with JS off
pusherthe agent that pushesT-blockthe thing being pushedtargetwhere it needs to end up

This is the real behavior-cloning policy from Chapter 1.1, running in your browser on a MuJoCo-WASM simulation — the same bc.py you'll train. Drag the block out of the region the demonstrations covered, and watch a confident policy fail. That failure has a name, and the whole course is built around understanding it.

Why this course is built the way it is

  • the code is the product

    One file you can read top to bottom

    Every chapter builds a single runnable script — a few hundred lines, no framework to excavate. You read it like a textbook, and every line you read is a line that runs.

  • runs on the free tier

    A laptop or a free Colab T4 is enough

    Every learner-facing path completes on CPU or a free T4. The wall-clock times on each page are measured on real hardware, never estimated; where a tier hasn't been measured yet, the page says so.

  • honest by construction

    Real sims, real policies, real numbers

    The browser demos run the same code you train locally. Seeds are mandatory, results reproduce within a recorded band, and every number you read traces back to a run — including the ones that show a method failing.

The arc

Six phases, one runnable file at a time.

43chapters are live and readable today — the whole arc from a bare simulation loop to a from-scratch practitioner's stack and a real-arm graduation. The only thing not yet linkable is the leaderboard capstone: it's graded against hidden seeds, so it opens when the grading server ships, and not before.

See the dependency map — how every chapter connects →

  1. Phase 0 · Foundations

    6 live

    Make the simulator behave. Step physics by hand, author a scene, get frames and rotations right, and record your first teleoperated dataset.

  2. Phase 1 · Imitation

    9 live

    Teach a policy from demonstrations. Behavior cloning, exactly why it breaks, and the models built to stop it from breaking.

  3. Phase 2 · Reinforcement

    8 live

    Learn from reward instead of examples — for when demonstrations run out and the robot has to try, fail, and improve.

  4. Phase 3 · Depth (optional)

    9 live

    Off the main line, taken when the itch strikes. Learn a world model, build a physics engine from scratch — dynamics, constraints, contact — compare it against the simulator you have trusted all along, then plan through your own engine with sampling-based MPC. The capstone needs none of it.

  5. Phase 4 · Capstone

    3 livecapstone gated

    Put every phase together into one project you can point at, measure honestly, and show — including the human-in-the-loop RL that turns a decent policy into a reliable one.

    coming · gatedThe leaderboard capstone is scored against hidden seeds, so those chapters open only once the grading server ships — they aren't linkable yet, and we won't pretend they are. The chapters that lead into them are already here.

  6. Phase 5 · Practitioner

    8 live

    Build the practitioner's stack from scratch. A ViT and contrastive vision-language for perception, the two-tower VLA shape and a FAST action tokenizer, LoRA and INT8 quantization by hand — then the real-arm teleop → record → train → deploy loop that graduates you off the simulator.