Robot Rave

Meet Ravitto - The Dancing Robot!

Experience 3D Gaussian Splat Viewer

A music-reactive dancing robot built on Raspberry Pi that dances to the beat of your music with LED matrix eyes that express emotions.

Origin

This project started as a hackathon project at the SoTa (State of the Art) event with Cocoa called "The Robot Rave". What began as an experimental idea to make a robot dance to music has evolved into a full-featured music-reactive robot with personality!

Features

• Music-Reactive Dancing: Real-time beat detection and BPM analysis to sync robot movements with music
• LED Matrix Eyes: Freenove 8x16 LED matrix displays expressive eye animations that react to music and movement
• Multiple Dance Patterns: Various dance styles including bounce, sway, shuffle, pump, and more
• Autonomous Mode: Robot listens for music and automatically starts/stops dancing
• Web Interface: Beautiful responsive control panel accessible from any device on the network
• Audio Visualization: Real-time waveform and spectrum analyzer display
• Smart Silence Detection: Robot stops dancing within ~500ms when music stops
• 3D Gaussian Splat Viewer: Immersive WebGL visualization of Ravitto with 50k+ particles

3D Gaussian Splat Viewer

Experience Ravitto in a stunning 3D particle cloud visualization! The Gaussian splat viewer renders the robot as thousands of floating particles with realistic colors and lighting.

Features:

• 50,000+ Particles: High-fidelity point cloud forming Ravitto's shape
• Orbital Controls: Drag to rotate, scroll to zoom, double-click to reset
• Quality Levels: 1x (25k) to 8x (200k) particles for any device
• Color Modes: Normal, Neon, Thermal, and Golden rendering styles
• Explode Effect: Watch particles burst outward and reform
• Auto-Rotate: Cinematic orbiting camera animation

Browser Requirements:

• Modern browser with WebGL support (Chrome, Firefox, Safari, Edge)
• Hardware acceleration enabled
• For best performance: dedicated GPU recommended for 4x/8x quality modes

Access at /splat when running the backend, or view the live demo.

Troubleshooting:

• If the viewer shows a black screen, check that WebGL is enabled in your browser
• For slow performance, try reducing quality to 1x or 2x mode
• On mobile devices, start with 1x quality for smoother experience

Hardware Requirements

• Raspberry Pi 4 (or 3B+)
• CamJam EduKit 3 (or compatible motor driver)
• 2x DC Motors with wheels
• USB Microphone
• Freenove 8x16 LED Matrix (HT16K33) - Optional but recommended
• Power supply for motors

Wiring

Motor Connections (CamJam EduKit 3)

Motor	GPIO Pin 1	GPIO Pin 2
Left	9	10
Right	7	8

LED Matrix (I2C)

Matrix Pin	Pi Pin
VCC	3.3V
GND	GND
SDA	SDA
SCL	SCL

Status LED

LED	GPIO Pin
Status	25

Installation

1. Enable I2C (for LED eyes)

sudo raspi-config
# Navigate to: Interface Options > I2C > Enable

2. Install System Dependencies

sudo apt-get update
sudo apt-get install -y portaudio19-dev i2c-tools

3. Install Python Dependencies

pip3 install flask numpy sounddevice scipy --break-system-packages
pip3 install adafruit-circuitpython-ht16k33 --break-system-packages

4. Verify I2C Connection (optional)

i2cdetect -y 1
# Should show devices at 0x70 and 0x71 for the LED matrix

Usage

Start the Robot

cd /path/to/robot-rave
python3 robot_backend.py

Access the Web Interface

Open a browser and navigate to:

• From the Pi: http://localhost:5000
• From another device: http://<pi-ip-address>:5000

Controls

• START AUTONOMOUS MODE: Robot listens for music and dances automatically
• Manual Controls: D-pad for direct motor control
• Sensitivity Slider: Adjust music detection sensitivity
• Gain Slider: Adjust microphone input gain
• EMERGENCY STOP: Immediately stops all movement

Project Structure

robot-rave/
├── robot_backend.py      # Main backend server with all logic
├── robot_frontend.html   # Web interface (served by Flask)
├── ravitto_studio.html   # Interactive 360° photo viewer
├── ravitto_splat.html    # 3D Gaussian splat WebGL viewer
├── ravitto_360.gif       # Robot showcase animation
├── images/               # Robot photos
├── README.md             # This file
├── CLAUDE.md             # Claude Code guidelines
└── AGENT.md              # AI assistant guidelines

API Endpoints

Endpoint	Method	Description
/	GET	Serves the web frontend
/splat	GET	3D Gaussian splat viewer
/studio	GET	Interactive 360° photo studio
/api/status	GET	Returns current robot state as JSON
/api/control/<cmd>	POST	Send control commands (forward, backward, left, right, stop, toggle_auto)
/api/sens/<val>	POST	Set sensitivity (0-100)
/api/gain/<val>	POST	Set microphone gain (0-100)
/api/eyes/<expression>	POST	Set eye expression
/api/eyes/special/<type>	POST	Trigger special eye animation

Eye Expressions

The LED matrix eyes support multiple expressions:

• normal - Default relaxed eyes
• happy - Curved happy eyes
• excited - Wide open eyes
• sleepy - Droopy tired eyes
• angry - Furrowed brow
• look_left, look_right, look_up - Directional looking

Special animations: heart, star, dizzy

How It Works

1. Audio Capture: USB microphone captures ambient audio at 44.1kHz
2. Feature Extraction: Analyzes RMS energy, spectral features, and frequency bands
3. Beat Detection: Identifies beats using onset detection and tempo estimation
4. Music Classification: Distinguishes music from silence, noise, and speech
5. Dance Engine: Selects appropriate dance patterns based on BPM, energy, and dominant frequency band
6. Motor Control: Executes dance moves through PWM-controlled DC motors
7. Eye Animation: Updates LED matrix based on energy, beats, and movement direction

Contributing

Contributions are welcome! Feel free to submit issues and pull requests.

License

MIT License - Feel free to use, modify, and distribute.

Acknowledgments

• SoTa Event & Cocoa - For hosting the hackathon that sparked this project
• CamJam EduKit - For the excellent robotics kit
• Adafruit - For the HT16K33 LED matrix library
• All the hackathon participants who cheered on the dancing robot!

---

Made with music and motors at The Robot Rave hackathon