Robot Kit Setup

Complete hardware and software setup for ZeroClaw robots

Robot Kit Setup

Complete guide to building a ZeroClaw-powered robot on Raspberry Pi 5.

Prerequisites

• Raspberry Pi 5 (8GB recommended) or Pi 4 (4GB+)
• All hardware components (see Overview)
• Basic soldering and wiring skills
• ~2-3 hours for full setup

Phase 1: Base Operating System

1. Flash Raspberry Pi OS

Use Raspberry Pi Imager:

# Download from https://www.raspberrypi.com/software/

# Configure:
# - OS: Raspberry Pi OS (64-bit, Bookworm)
# - Storage: Your SD card or NVMe
# - Settings (gear icon):
#   - Enable SSH
#   - Set hostname: robot
#   - Set username/password
#   - Configure WiFi

2. Update System

ssh [e-mail]

sudo apt update && sudo apt upgrade -y

# Install build essentials
sudo apt install -y \
    build-essential \
    git \
    curl \
    cmake \
    pkg-config \
    libssl-dev \
    libasound2-dev \
    libclang-dev \
    ffmpeg \
    fswebcam

3. Install Rust

curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh
source ~/.cargo/env
rustc --version

Phase 2: AI Software Stack

1. Install Ollama (Local LLM)

# Install Ollama
curl -fsSL https://ollama.ai/install.sh | sh

# Start service
sudo systemctl enable ollama
sudo systemctl start ollama

# Pull models (choose based on RAM)
# For 8GB Pi:
ollama pull llama3.2:3b      # 3B params, fast
ollama pull moondream        # Vision model, small

# For 4GB Pi:
ollama pull phi3:mini        # 3.8B, very fast
ollama pull moondream        # Vision

# Test
curl http://localhost:11434/api/tags

2. Install Whisper.cpp (Speech-to-Text)

# Clone and build
git clone https://github.com/ggerganov/whisper.cpp
cd whisper.cpp
make -j4

# Download model
bash ./models/download-ggml-model.sh base

# Install
sudo cp main /usr/local/bin/whisper-cpp
mkdir -p ~/.zeroclaw/models
cp models/ggml-base.bin ~/.zeroclaw/models/

# Test
echo "Testing microphone" | espeak -w /tmp/test.wav
whisper-cpp -m ~/.zeroclaw/models/ggml-base.bin -f /tmp/test.wav

3. Install Piper TTS (Text-to-Speech)

# Download Piper binary
wget https://github.com/rhasspy/piper/releases/download/v1.2.0/piper_arm64.tar.gz
tar -xzf piper_arm64.tar.gz
sudo cp piper/piper /usr/local/bin/

# Download voice model
mkdir -p ~/.zeroclaw/models/piper
cd ~/.zeroclaw/models/piper
wget https://huggingface.co/rhasspy/piper-voices/resolve/main/en/en_US/lessac/medium/en_US-lessac-medium.onnx
wget https://huggingface.co/rhasspy/piper-voices/resolve/main/en/en_US/lessac/medium/en_US-lessac-medium.onnx.json

# Test
echo "Hello, I am your robot!" | piper --model ~/.zeroclaw/models/piper/en_US-lessac-medium.onnx --output_file test.wav
aplay test.wav

4. Install RPLidar SDK

# Python driver
pip3 install rplidar-roboticia

# Add user to dialout group for serial access
sudo usermod -aG dialout $USER
# Logout and login for group change to take effect

Phase 3: Hardware Wiring

Wiring Diagram

                     ┌─────────────────────────────────────┐
                     │          Raspberry Pi 5             │
                     │                                     │
   ┌─────────────────┤ GPIO 4  ←── E-Stop Button (NC)      │
   │                 │ GPIO 5  ←── Bump Sensor Left        │
   │                 │ GPIO 6  ←── Bump Sensor Right       │
   │                 │ GPIO 12 ──► Motor PWM 1             │
   │                 │ GPIO 13 ──► Motor PWM 2             │
   │                 │ GPIO 17 ←── PIR Motion 1            │
   │                 │ GPIO 18 ──► LED Matrix (WS2812)     │
   │                 │ GPIO 23 ──► Ultrasonic Trigger      │
   │                 │ GPIO 24 ←── Ultrasonic Echo         │
   │                 │ GPIO 27 ←── PIR Motion 2            │
   │                 │                                     │
   │ ┌───────────────┤ USB-A   ←── RPLidar A1              │
   │ │               │ USB-A   ←── USB Microphone          │
   │ │               │ USB-A   ←── USB Webcam              │
   │ │               │ 3.5mm   ─► Speaker/Amp             │
   │ │               └─────────────────────────────────────┘
   │ │
   │ └────── RPLidar A1 (USB)
   │
   │    ┌──────────────────┐      ┌─────────────┐
   └────┤  Motor Controller├──────┤  4× Motors  │
        │  (L298N/TB6612)  │      │ Omni Wheels │
        └──────────────────┘      └─────────────┘

Pin Assignments (BCM Numbering)

Function	GPIO Pin	Device
E-Stop	4	Red mushroom button (normally closed)
Bump Left	5	Microswitch
Bump Right	6	Microswitch
Motor PWM 1	12	L298N ENA
Motor PWM 2	13	L298N ENB
PIR Motion 1	17	PIR sensor
LED Matrix	18	WS2812B data pin
Ultrasonic Trig	23	HC-SR04 trigger
Ultrasonic Echo	24	HC-SR04 echo
PIR Motion 2	27	PIR sensor

E-Stop Wiring (Critical)

E-Stop Button (NC)      Raspberry Pi
──────────────────      ────────────
NO (not used)
COM ─────────────────────────► GPIO 4
NC ──────────────────────────► 3.3V

Normal: Circuit closed, GPIO 4 reads HIGH
Pressed: Circuit open, GPIO 4 reads LOW (STOP!)

Phase 4: ZeroClaw Robot Kit Build

1. Clone and Build

git clone https://github.com/zeroclaw-labs/zeroclaw
cd zeroclaw

# Build robot kit
cargo build -p zeroclaw-robot-kit --release

# Build main zeroclaw (optional)
cargo build --release --features peripheral-rpi

2. Install Configuration

mkdir -p ~/.zeroclaw
cp crates/robot-kit/robot.toml ~/.zeroclaw/
cp crates/robot-kit/SOUL.md ~/.zeroclaw/workspace/

3. Configure robot.toml

Edit ~/.zeroclaw/robot.toml for your hardware:

# Drive system
[drive]
backend = "serial"  # "ros2", "serial", "gpio", or "mock"
serial_port = "/dev/ttyACM0"  # Arduino motor controller
max_speed = 0.3     # Start conservative!
max_rotation = 0.5

# Camera
[camera]
device = "/dev/video0"  # USB webcam or Pi Camera
width = 640
height = 480
vision_model = "moondream"
ollama_url = "http://localhost:11434"

# Audio
[audio]
mic_device = "plughw:1,0"      # Find with: arecord -l
speaker_device = "plughw:0,0"  # Find with: aplay -l
whisper_model = "base"
whisper_path = "/usr/local/bin/whisper-cpp"
piper_path = "/usr/local/bin/piper"
piper_voice = "en_US-lessac-medium"

# Sensors
[sensors]
lidar_port = "/dev/ttyUSB0"  # RPLidar
lidar_type = "rplidar"
motion_pins = [17, 27]
ultrasonic_pins = [23, 24]

# Safety (CRITICAL!)
[safety]
min_obstacle_distance = 0.3    # 30cm minimum
slow_zone_multiplier = 3.0     # Slow at 90cm
approach_speed_limit = 0.3     # 30% near obstacles
max_drive_duration = 30        # Auto-stop after 30s
estop_pin = 4
bump_sensor_pins = [5, 6]
bump_reverse_distance = 0.15
confirm_movement = false       # Set true for extra safety
predict_collisions = true
sensor_timeout_secs = 5
blind_mode_speed_limit = 0.2

Phase 5: Component Testing

Test LIDAR

python3 << 'EOF'
from rplidar import RPLidar
lidar = RPLidar('/dev/ttyUSB0')
for scan in lidar.iter_scans():
    print(f'Got {len(scan)} points')
    break
lidar.stop()
lidar.disconnect()
EOF

Test Camera

ffmpeg -f v4l2 -video_size 640x480 -i /dev/video0 -frames:v 1 test.jpg
xdg-open test.jpg  # View on desktop

Test Microphone

arecord -D plughw:1,0 -f S16_LE -r 16000 -c 1 -d 3 test.wav
aplay test.wav

Test Speaker

echo "Testing speaker" | piper --model ~/.zeroclaw/models/piper/en_US-lessac-medium.onnx --output_file - | aplay

Test Ollama

curl http://localhost:11434/api/generate -d '{"model":"llama3.2:3b","prompt":"Say hello"}'

Phase 6: Robot Startup

Manual Start

cd ~/zeroclaw

# Start Ollama (if not running)
ollama serve &

# Test in mock mode (no hardware)
export MOCK_MODE=true
./target/release/zeroclaw agent -m "Say hello and show a happy face"

# Test with real hardware (robot on blocks!)
unset MOCK_MODE
./target/release/zeroclaw agent -m "Move forward 0.5 meters"

Automatic Startup Script

Create ~/start_robot.sh:

#!/bin/bash
set -e

echo "Starting robot..."

# Start Ollama
if ! pgrep -x "ollama" > /dev/null; then
    ollama serve &
    sleep 5
fi

# Start sensor loop (optional background task)
if [ -f ~/sensor_loop.py ]; then
    python3 ~/sensor_loop.py &
    SENSOR_PID=$!
fi

# Start zeroclaw
cd ~/zeroclaw
./target/release/zeroclaw daemon

chmod +x ~/start_robot.sh

Systemd Service (Auto-start on Boot)

Create /etc/systemd/system/zeroclaw-robot.service:

[Unit]
Description=ZeroClaw Robot
After=network.target ollama.service

[Service]
Type=simple
User=pi
WorkingDirectory=/home/pi/zeroclaw
ExecStart=/home/pi/start_robot.sh
Restart=on-failure
RestartSec=10

[Install]
WantedBy=multi-user.target

sudo systemctl daemon-reload
sudo systemctl enable zeroclaw-robot
sudo systemctl start zeroclaw-robot

# Check status
sudo systemctl status zeroclaw-robot
journalctl -u zeroclaw-robot -f

Phase 7: Personality Customization

Edit ~/.zeroclaw/workspace/SOUL.md to customize the robot’s personality:

# Buddy the Robot

You are Buddy, a friendly robot companion!

## Personality
- Playful
- Patient
- Safe
- Curious

## Safety Rules (NEVER BREAK)
1. Never move faster than walking speed
2. Always stop if asked
3. Keep 1 meter distance
4. Never go near stairs or pools
5. Alert adults if child hurt

## Games
1. Hide and Seek
2. Follow the Leader
3. Simon Says
...

Safety Checklist

Before first run with real motors:

• E-stop button wired and tested
• Bump sensors wired and tested
• LIDAR spinning and returning data
• max_speed set to 0.3 or lower
• Robot on blocks (wheels not touching ground)
• First test with backend = "mock" in config
• Adult supervision ready
• Clear space around robot (3+ meters)
• Emergency stop response time < 1 second

Troubleshooting

See detailed troubleshooting in:

• Raspberry Pi Setup
• Robot Kit API

Common Issues

LIDAR not detected:

ls -la /dev/ttyUSB*
dmesg | grep -i usb

Ollama slow:

# Use smaller model
ollama rm llama3.2:3b
ollama pull phi3:mini

Motors not responding:

# Check serial
ls /dev/ttyACM*
screen /dev/ttyACM0 115200

Next Steps

API Reference

Tool specifications and usage

Overview

Architecture and features

Raspberry Pi

GPIO and Pi-specific setup

Examples

Example robot programs