Group Project EMOVERSE

This project explores an integration of culinary decision-making with emotional reflection and AI-generated poetry. By transforming a user’s meal choices, hunger levels, and emotional states into personalized haiku, we aim to inspire a thoughtful pause before everyday eating habits with Arduino-based sensors and a GPT-4 language model.

Instead of being a calorie calculator or an AI interpreter, this project seeks to introduce a reflective moment before making decisions—specifically, before opening the fridge. Inspired by the simplicity of haiku poetry, the system converts user inputs into short, elegant verses that offer insight, amusement, or comfort. The ultimate goal is not to restrict behavior but to provide a gentle companion that celebrates personal choice, highlights emotional states, and elevates the act of nourishing oneself.

As the user approaches the fridge, the rotary knob sensor tracks their hunger level via a numeric input, and the rotary encoder sensor records their emotional state through an emotion wheel (e.g., "Happy" or "Tense"). The user will be required to select from a series of meal-oriented data fridge magnets: "Meal," "Snack," or "Drink" for the purpose of consuming food.  After a short pause, the input is sent to a GPT-4 model and generates a personalized haiku through a text-to-speech (TTS) engine. The system also logs the generated haiku daily, allowing users to trace their history, gain self-understanding, and observe patterns over time.

Medium of physical object: 
Custom electronics and a 3D printed enclosure.