AI Lab Simulator

A hackathon-built MVP for running virtual physics experiments and generating AI explanations of the results. Pick an experiment, run it, and get a plain-language breakdown of what happened and why — powered by OpenAI, with a built-in fallback so the demo never breaks if the API key is missing.

Features

• Three interactive simulations: Ohm's Law, Projectile Motion, and Hooke's Law
• Flask REST API with a dedicated endpoint per simulation
• AI-generated explanations of results via OpenAI, with adjustable difficulty level
• Graceful degradation: works fully offline with a fallback explanation if no API key is set
• Lightweight vanilla JS frontend with charts and loading states — no framework overhead

Tech Stack

• Backend: Flask 3, Flask-CORS, Gunicorn
• AI: OpenAI API (optional)
• Frontend: HTML/CSS/vanilla JavaScript

API Reference

Endpoint	Method	Body	Description
/simulate/ohms	POST	{ "voltage": number, "resistance": number }	Runs Ohm's Law simulation
/simulate/projectile	POST	{ "angle": number, "velocity": number }	Runs projectile motion simulation
/simulate/hooke	POST	{ "force": number, "spring_constant": number }	Runs Hooke's Law simulation (spring constant defaults to 10)
/explain	POST	simulation result payload	Returns an AI-generated explanation at the requested difficulty level

Project Structure

backend/
├── app.py                    # Flask app entry point
├── routes/
│   ├── simulations.py        # Simulation endpoints
│   └── explain.py            # AI explanation endpoint
└── services/
    ├── simulations.py        # Physics calculations
    └── ai_explainer.py       # OpenAI integration + fallback logic
frontend/
├── index.html
├── app.js
└── styles.css

Run Locally

1. Create and activate a Python virtual environment.

2. Install dependencies:

   pip install -r backend/requirements.txt

3. (Optional) Set your OpenAI API key for live AI explanations:

   export OPENAI_API_KEY=your_key_here

4. Start the Flask app:

   python -m backend.app

5. Open http://127.0.0.1:5000

If OPENAI_API_KEY isn't set, the app still works end-to-end and returns a built-in fallback explanation, so the demo never breaks.