iuliaferoli/backtoengineering

Back to Engineering is a learning website that organizes robotics and physical AI topics into a visual tech tree, similar to the skill trees found in games like Civilization or Satisfactory. Instead of a traditional linear course, learners click on topic cards to explore concepts, watch tutorials, and find resources, with arrows showing which topics need to be understood before moving on to the next one. Star-shaped nodes mark hands-on projects where the theory gets applied to something real, and topics are color-coded by area, such as electronics, mechanical engineering, programming, data science, and AI. The stated goal is to take someone from basic curiosity all the way to building intelligent robots, starting with simple exercises like blinking an LED or building a cardboard robot arm, and gradually working up to more advanced topics like ROS 2 and AI-driven manipulation using vision-language-action models. A featured example is a seven-episode robot arm project that follows this progression from a scratch-built cardboard prototype to an AI-powered robotic arm trained on custom data. Under the hood, the site itself is static and built with MkDocs Material for the pages, Cytoscape.js and a custom renderer for the interactive tree graph, and a Python script that builds the tree's underlying graph data. It automatically deploys to GitHub Pages whenever changes are pushed. Adding new content means creating a Markdown file with specific frontmatter fields, like a unique ID, category, and any prerequisite topics, then regenerating the graph data before pushing. The project uses two separate licenses: the site code and build scripts are under the Apache License 2.0, while the actual learning content, meaning the topic and project write-ups, is under a Creative Commons Attribution-NonCommercial license, which allows free reuse and adaptation for non-commercial purposes with credit, but does not allow commercial use of that content.