Welcome to the future of education. This isn't just a PDF; it's a living, breathing platform that adapts to the student. Built for the SpecKit Hackathon, this project demonstrates the power of Agentic Coding to create high-fidelity educational tools.
Embedded directly into the learning experience is an AI Assistant powered by Google Gemini 2.5 Flash.
- Context Aware: Select any text in the book, click "Ask", and the AI explains that specific concept.
- Personalization Engine: The AI knows if you are a "Python Expert" or "Hardware Novice" and adjusts its analogies accordingly.
- Multi-Lingual: Instantly translate technical documentation into Urdu (and other languages) while preserving code blocks.
- Deep Space Mode: A visually stunning dark mode with animated starfields.
- Holographic HUD: Glassmorphism effects on the navbar and chat widgets.
- Interactive Diagrams: Mermaid.js flowcharts for ROS 2 nodes and Control Systems.
From ROS 2 Humble internals to Vision-Language-Action (VLA) models, this textbook covers the full stack of modern physical AI.
- Multi-OS Labs: Dedicated tabs for Linux, Windows (WSL2), and macOS setup.
- Code-First Approach: Python labs for Inverse Kinematics and Reinforcement Learning.
The project follows the SpecKit Methodology: Specification → Implementation.
/ (Root)
├── sspecs/ # 📐 The Blueprints (SpecKit Plans)
├── course_outline.md # 📝 Curriculum Design
└── physical_ai_textbook/ # 🏗️ The Implementation
├── docs/ # - Markdown Content (The Book)
├── src/ # - React Frontend (ChatInterface.tsx)
├── static/ # - Assets (Images)
└── backend/ # - 🧠 The Intelligence Engine
├── app/main.py # - FastAPI + Gemini Logic
├── requirements.txt
└── Procfile # - Deployment Config
cd physical_ai_textbook
npm install
npm run start
# Opens at http://localhost:3000To enable the Chatbot and Personalization features:
cd physical_ai_textbook/backend
# Create a .env file with GEMINI_API_KEY=...
pip install -r requirements.txt
uvicorn app.main:app --reload
# Runs at http://localhost:8000Traditional textbooks are static. If a student doesn't understand "Kalman Filters", they are stuck. In this project, the student can click "✨ Personalize", and the book rewrites itself:
"Imagine a Kalman Filter like blending two eyes: one eye is your math model, the other is your sensor. Both are blurry, but together they see clearly."
This is the power of Physical AI + Generative AI.
- Spec-Driven: Built using SpecKit-Plus workflows.
- RAG Chatbot: Integrated via FastAPI + Gemini.
- Bonus: Personalization (Profile Context).
- Bonus: Translation (Urdu Support).
- Bonus: Reusable Intelligence (Subagents).
Created with ❤️ by Usama & Antigravity (Claude Code Agent)