Skip to main content

Build the Body. Code the Brain.

Welcome to the world's first spec-driven course on Physical AI & Humanoid Robotics. This isn't just a textbook—it's your guide to building intelligent robots that perceive, think, and act in the real world.

Written by Sharmeen Fatima, this course bridges the gap between software engineering and embodied intelligence, teaching you how to create robots that don't just exist in code, but move, sense, and respond in physical spaces.

What You'll Learn​

This textbook covers the complete robotics stack:

ModuleFocusKey Technologies
Module 1The Robotic Nervous SystemROS 2, rclpy, URDF
Module 2The Digital TwinGazebo, Unity, Simulation
Module 3The AI-Robot BrainNVIDIA Isaac Sim, Nav2, SLAM
Module 4Vision-Language-ActionOpenAI Whisper, GPT-4, LLMs
Module 5Capstone ProjectFull autonomous system integration

Why This Course is Different​

1. Spec-Driven Development​

Every concept is built on:

  • ✅ Clear specifications: Know exactly what you're building
  • ✅ Testable outcomes: Verify your robot works at every step
  • ✅ Production-ready code: No toy examples, real-world patterns

2. Co-Learning with AI​

You're not alone! This course embraces modern tools:

  • Use ChatGPT and Claude to debug faster
  • Leverage LLMs for code generation
  • Learn to work with AI, not against it

3. Hands-On from Day 1​

No endless theory. You'll:

  • Write your first ROS 2 node in Chapter 1
  • Control a simulated robot by Module 2
  • Build a voice-controlled autonomous robot by the Capstone

Who This Course is For​

Beginners: If you know Python and want to enter robotics, start here. We use biological analogies to make ROS 2 intuitive.

Software Engineers: If you build web apps or AI models and want to work with physical systems, this course bridges the gap.

Robotics Enthusiasts: If you've tinkered with Arduino or Raspberry Pi and want to scale up to autonomous systems, this is your roadmap.

Prerequisites​

Required:

  • Comfortable with Python (functions, classes, async)
  • Basic command-line usage (cd, ls, pip install)
  • Familiarity with Git (clone, commit)

Recommended but not required:

  • Undergraduate-level physics (helpful for simulation tuning)
  • Experience with Linux (Ubuntu 22.04 preferred)

System Requirements:

  • Minimum: Ubuntu 22.04, 8 GB RAM, 50 GB disk
  • Recommended: Ubuntu 22.04, 16 GB RAM, NVIDIA RTX GPU, 100 GB disk
  • Alternative: Docker on Windows/macOS (with performance trade-offs)

Course Philosophy​

The Cyber-Physical Mindset​

Robotics isn't just software or hardware—it's both. You'll learn to:

  • Think in coordinate frames (not just variables)
  • Design for physics (not just algorithms)
  • Test in simulation before deploying to hardware

From Zero to Autonomous​

We follow a progression:

  1. Nodes (communication)
  2. Simulation (safe testing)
  3. Navigation (autonomous movement)
  4. Intelligence (voice + vision)
  5. Integration (full system)

Each module builds on the last, but is independently valuable.

How to Use This Textbook​

🎯 Linear Path (Recommended for Beginners)​

Work through modules 1-5 in order. Each chapter has:

  • Learning objectives: What you'll master
  • Code examples: Copy, paste, and run
  • Mermaid diagrams: Visualize architectures
  • Hands-on exercises: Test your understanding

🚀 Skip to What You Need (For Experienced Learners)​

  • Already know ROS 2? Jump to Module 2 (Simulation)
  • Want AI control ASAP? Skim Modules 1-2, focus on Module 4 (VLA)
  • Need Nav2 only? Read Module 3 (Isaac Sim & Nav2)

💡 Best Practices​

  1. Run every code example: Don't just read—execute
  2. Complete exercises: Passive learning doesn't work in robotics
  3. Break things: Debugging teaches more than perfect code
  4. Use AI assistants: ChatGPT, Claude, and Copilot are your friends

Visual Learning​

This course uses 30+ Mermaid.js diagrams to explain:

  • System architectures
  • Data flows
  • State machines
  • Dependency graphs

Every technical concept has a visual aid.

The Capstone Project​

Your final project: Build a voice-controlled autonomous humanoid that:

  • Listens to commands ("Go to the kitchen")
  • Parses with an LLM (GPT-4 or Gemini)
  • Navigates using Nav2
  • Executes in Gazebo or Isaac Sim

This isn't optional—it's the culmination of everything you've learned.

Getting Help​

  • Documentation: Every chapter links to official ROS 2, Nav2, and Isaac Sim docs
  • Community: Join the Physical AI Discord (coming soon)
  • Issues: Report errors at GitHub Issues

Ready to Begin?​

Let's start with the fundamentals. Click below to dive into Module 1: The Robotic Nervous System.

👉 Start with ROS 2 Nodes

Good luck, and welcome to the future of Physical AI! 🤖⚡

— Sharmeen Fatima