BrainrotGenerator Slot Machine

An interactive kiosk-style slot machine that generates AI art using OpenAI's DALL-E 3. Users stop three spinning reels (Animal, Fruit, Object) and the system generates a unique "brainrot" themed image combining all three elements.

Designed for Raspberry Pi with physical GPIO buttons, but also works on desktop for development and testing.

✨ Features

• 3D Spinning Reels: Smooth CSS-based cylinder animations with realistic physics
• Physical GPIO Integration: Works with switches or push buttons on Raspberry Pi
• AI-Powered Generation: Uses GPT-4o-mini for creative prompts and DALL-E 3 for images
• Real-time WebSocket Communication: Instant feedback between hardware and UI
• Gallery System: View all generated images with metadata
• Kiosk Mode Ready: Fullscreen, cursor-hidden interface perfect for exhibitions
• Error Recovery: Automatic reset on API failures, robust handshake mechanism

🎨 Design

• Custom typography using Adobe Typekit fonts (ironmonger-inlaid, freehouse)
• Golden yellow accent color (#ffcc33) with dark theme
• 3D CSS transforms for realistic reel effects
• Responsive animations optimized for Raspberry Pi performance

📁 Project Structure

slotmachine/
├── backend/
│   └── main.py              # FastAPI server with WebSocket & GPIO
├── frontend/
│   ├── index.html           # Slot machine interface
│   ├── gallery.html         # Image gallery viewer
│   ├── styles/
│   │   ├── common.css       # Shared styles
│   │   ├── slot.css         # Slot machine specific
│   │   └── gallery.css      # Gallery specific
│   └── generated/           # Generated images & manifest
├── setup/ (Still needs to be tested)
│   ├── DEPLOYMENT.md        # Complete Raspberry Pi guide
│   ├── setup_pi.sh          # Auto-setup script
│   └── setup_kiosk.sh       # Kiosk mode configuration
├── env/                     # Python virtual environment
├── requirements.txt         # Python dependencies
└── README.md

🚀 Quick Start (Development)

1. Install Dependencies

# Create virtual environment
python -m venv env

# Windows
env\Scripts\activate

# macOS/Linux
source env/bin/activate

# Install packages
pip install -r requirements.txt

2. Configure OpenAI API

Create a .env file:

OPENAI_API_KEY=sk-your-actual-key-here

3. Run the Server

uvicorn backend.main:app --reload --host 0.0.0.0 --port 8000

The server starts on http://localhost:8000

4. Test the Interface

• Slot Machine: http://localhost:8000
• Gallery: http://localhost:8000/static/gallery.html

Keyboard Controls (for testing without GPIO):

• A, S, D - Stop reels A, B, C
• R - Reset

🥧 Raspberry Pi Deployment

For complete Raspberry Pi setup with GPIO buttons and kiosk mode, see:

setup/DEPLOYMENT.md

Quick setup:

# Transfer project to Pi
scp -r slotmachine/ pi@192.168.1.XXX:~/

# On the Pi
cd slotmachine
chmod +x setup/*.sh
bash setup/setup_pi.sh

# Edit .env with your OpenAI key
nano .env

# Optional: Enable kiosk mode
bash setup/setup_kiosk.sh

GPIO Pin Mapping

• GPIO 23 → Reel A (Left - Animal)
• GPIO 27 → Reel B (Middle - Fruit)
• GPIO 22 → Reel C (Right - Object)

Works with both push buttons and toggle switches.

🎮 How It Works

1. Spin: Three reels start spinning automatically showing random items
2. Stop: Press GPIO buttons (or A/S/D keys) to stop each reel
3. Generate: After all reels stop, frontend sends handshake to backend
4. Prompt: GPT-4o-mini creates a creative Italian name and DALL-E prompt
5. Image: DALL-E 3 generates the image (1024x1024, vivid style)
6. Display: Image appears fullscreen with metadata
7. Gallery: All images saved with timestamps and accessible from gallery page

🛠️ Technology Stack

• Backend: FastAPI, WebSockets, gpiozero, pigpio
• Frontend: Vanilla HTML/CSS/JavaScript with 3D transforms
• AI: OpenAI GPT-4o-mini (prompts), DALL-E 3 (images)
• Hardware: Raspberry Pi with GPIO buttons/switches
• Deployment: systemd service, Chromium kiosk mode

🐛 Troubleshooting

Backend won't start

# Check Python environment
which python
python --version

# Check dependencies
pip list | grep fastapi

# View errors
cd backend
python main.py

GPIO buttons not working

# Check pigpiod service
sudo systemctl status pigpiod

# Test GPIO manually
python3 -c "from gpiozero import Button; Button(23).wait_for_press(); print('Works!')"

Images not generating

• Verify OpenAI API key in .env
• Check network: ping api.openai.com
• View backend logs: sudo journalctl -u brainrot-backend -f

📝 Notes

• Uses handshake mechanism to prevent frontend/backend desync
• Supports both switch state changes and button presses
• Auto-reset on API errors with user-friendly messages
• Images saved to frontend/generated/ with JSONL manifest
• CSS optimized for Raspberry Pi performance (no heavy animations)

🔐 Security

• Keep .env file private (contains API key)
• Change default Pi password
• Use static IP for stable kiosk deployment
• Consider firewall if exposing to network

📜 License

MIT