El Plan

Learning STEM at full throttle with the power of Formula 1!

Comment

Inspiration

STEM education can be dry and intimidating for many students. But what if learning math or physics felt like racing down a Grand Prix circuit? Inspired by the high-speed world of Formula 1, we wanted to fuse real racing data with engaging educational experiences to teach science and engineering in a way that’s fast, interactive, and fun.

What It Does

El Plan STEM is a Formula 1-themed educational chatbot platform designed to teach STEM subjects using real F1 data. It features:

  • An AI-powered chatbot that generates and evaluates science/math questions using real lap times, pit stop data, and driver stats.
  • A general F1 assistant that provides facts, insights, and answers about races, drivers, and circuits.
  • A note-taking system for users to save study points and answers.
  • A data graphing system that visualizes key racing metrics and trends.
  • Built-in machine learning models for predictive insights and analysis (e.g., turn severity, momentum shifts).

How We Built It

  • Backend: Django with MongoDB for flexible, schema-less F1 data storage.
  • Chatbot Engine: Google’s Gemini AI model, integrated with sentence-transformers for semantic vector search.
  • ML Models: Used scikit-learn and TensorFlow to train predictive models based on racing data.
  • Frontend: Django templates with JavaScript and CSS for interactivity and responsiveness.
  • Deployment Ready: Virtual environment setup, organized route structure, and API endpoints for chatbot interaction.

Challenges We Ran Into

  • AI Tuning: Crafting STEM questions that made sense contextually and aligned with racing data took extensive prompt engineering.
  • Data Handling: Real F1 data is complex and sparse — formatting it for meaningful use in chat responses was a balancing act.
  • Model Integration: Ensuring ML models performed consistently within chatbot interactions required careful optimization and testing.

Accomplishments That We're Proud Of

  • Built a working educational chatbot that teaches real science through real-world data.
  • Successfully integrated Gemini AI to simulate educational dialogue with context awareness.
  • Developed a system that makes learning more relatable and exciting using Formula 1.
  • Delivered a modular, scalable backend ready for future features like user tracking and progress analytics.

What We Learned

  • How to combine machine learning, AI, and real-time data into an educational workflow.
  • The importance of data preprocessing when using AI and ML on real-world datasets.
  • How to structure a Django project with multiple apps and custom routes cleanly.
  • That learning can be significantly improved with gamification and real-world context.

What's Next for El Plan

  • More subjects and grade levels.
  • Including speech-to-text features.
  • Have a progress tracker.
  • Interactive telemetry maps and sector analysis tools.
  • Account management.
  • Launch a Beta version.
Share this project:

Updates