Chapter 1: Physics Simulation with Gazebo

Learning Objectives

• Understand why simulation is essential for robotics development
• Install and configure Gazebo Fortress
• Create robot models using URDF and SDF
• Spawn and control robots in simulated environments
• Add physics properties and constraints

1.1 Why Simulation?

The Reality of Robot Development

Problem: Physical robots are:

• Expensive: $50,000 - $1,000,000+ for humanoids
• Dangerous: A malfunctioning humanoid can cause serious injury
• Slow: Each test requires setup, deployment, and reset
• Limited: Can't test extreme scenarios (falling, collisions

)

Solution: Validate algorithms in simulation first, then deploy to hardware.

Simulation Benefits

Aspect	Simulation	Physical Robot
Cost	Free	$$$$
Safety	Perfectly safe	Can damage robot/environment
Iteration Speed	Seconds	Minutes to hours
Parallelization	Run 1000 instances	Limited by hardware
Scenarios	Any imaginable	Limited by lab
Debugging	Full state access	Limited sensors

The Sim-to-Real Gap

Simulation is not perfect:

• Physics: Simplified contact models
• Sensors: Noise models are approximations
• Actuators: Ideal vs. real motor dynamics

Solution: Domain randomization and careful calibration minimize the gap.

1.2 Gazebo Architecture

Gazebo Classic vs. Gazebo (Ignition/Fortress)

We use Gazebo Fortress (also called Ignition Gazebo):

• Modern architecture
• Better performance
• Improved rendering (Ogre 2)
• Plugins for ROS 2 integration

Installation

# Install Gazebo Fortress
sudo apt-get update
sudo apt-get install ros-humble-ros-gz

# Verify installation
gz sim --version

Architecture

┌──────────────┐
│  Gazebo GUI  │ ← Visualization
└──────┬───────┘
       │
┌──────▼───────┐
│ Gazebo Server│ ← Physics engine
│   (gz sim)   │
└──────┬───────┘
       │
   ┌───▼───┐
   │ PhysX │ ← Contact dynamics
   └───────┘

1.3 Creating World Files

SDF (Simulation Description Format)

Gazebo uses SDF to describe worlds and models.

worlds/empty.sdf

<?xml version="1.0" ?>
<sdf version="1.8">
  <world name="empty_world">
    
    <!-- Physics engine settings -->
    <physics name="1ms" type="dart">
      <max_step_size>0.001</max_step_size>
      <real_time_factor>1.0</real_time_factor>
    </physics>
    
    <!-- Lighting -->
    <light type="directional" name="sun">
      <cast_shadows>true</cast_shadows>
      <pose>0 0 10 0 0 0</pose>
      <diffuse>0.8 0.8 0.8 1</diffuse>
      <specular>0.2 0.2 0.2 1</specular>
      <attenuation>
        <range>1000</range>
      </attenuation>
      <direction>-0.5 0.1 -0.9</direction>
    </light>
    
    <!-- Ground plane -->
    <model name="ground_plane">
      <static>true</static>
      <link name="link">
        <collision name="collision">
          <geometry>
            <plane>
              <normal>0 0 1</normal>
              <size>100 100</size>
            </plane>
          </geometry>
        </collision>
        <visual name="visual">
          <geometry>
            <plane>
              <normal>0 0 1</normal>
              <size>100 100</size>
            </plane>
          </geometry>
          <material>
            <ambient>0.8 0.8 0.8 1</ambient>
            <diffuse>0.8 0.8 0.8 1</diffuse>
          </material>
        </visual>
      </link>
    </model>
    
  </world>
</sdf>

Launching a World

gz sim worlds/empty.sdf

1.4 URDF to SDF Conversion

Why Two Formats?

• URDF: ROS-specific, simpler, good for robot description
• SDF: Gazebo-specific, more features, good for world description

Gazebo can import URDF but converts it to SDF internally.

Enhanced URDF with Gazebo Tags

urdf/humanoid_gazebo.urdf

<?xml version="1.0"?>
<robot name="humanoid_gazebo">
  
  <link name="base_link">
    <visual>
      <geometry>
        <box size="0.3 0.2 0.5"/>
      </geometry>
      <material name="blue">
        <color rgba="0.2 0.4 0.8 1"/>
      </material>
    </visual>
    
    <collision>
      <geometry>
        <box size="0.3 0.2 0.5"/>
      </geometry>
    </collision>
    
    <inertial>
      <mass value="10.0"/>
      <inertia ixx="0.2" ixy="0" ixz="0"
               iyy="0.3" iyz="0" izz="0.2"/>
    </inertial>
  </link>
  
  <!-- Gazebo-specific material -->
  <gazebo reference="base_link">
    <material>Gazebo/Blue</material>
    <mu1>0.8</mu1> <!-- Friction coefficient -->
    <mu2>0.8</mu2>
  </gazebo>
  
  <link name="right_foot">
    <visual>
      <geometry>
        <box size="0.2 0.1 0.05"/>
      </geometry>
    </visual>
    
    <collision>
      <geometry>
        <box size="0.2 0.1 0.05"/>
      </geometry>
    </collision>
    
    <inertial>
      <mass value="1.0"/>
      <inertia ixx="0.01" ixy="0" ixz="0"
               iyy="0.01" iyz="0" izz="0.01"/>
    </inertial>
  </link>
  
  <joint name="right_ankle" type="revolute">
    <parent link="base_link"/>
    <child link="right_foot"/>
    <origin xyz="0.1 -0.15 -0.4" rpy="0 0 0"/>
    <axis xyz="1 0 0"/>
    <limit lower="-0.5" upper="0.5" effort="100" velocity="2"/>
  </joint>
  
  <!-- Gazebo plugin for joint control -->
  <gazebo>
    <plugin filename="libgazebo_ros2_control.so" name="gazebo_ros2_control">
      <robot_param>robot_description</robot_param>
      <parameters>config/controllers.yaml</parameters>
    </plugin>
  </gazebo>
  
</robot>

1.5 Spawning Robots

Using ros2 launch

launch/spawn_humanoid.launch.py

from launch import LaunchDescription
from launch_ros.actions import Node
from ament_index_python.packages import get_package_share_directory
import os

def generate_launch_description():
    pkg_path = get_package_share_directory('humanoid_gazebo')
    urdf_file = os.path.join(pkg_path, 'urdf', 'humanoid_gazebo.urdf')
    
    with open(urdf_file, 'r') as f:
        robot_description = f.read()
    
    spawn_entity = Node(
        package='ros_gz_sim',
        executable='create',
        arguments=[
            '-name', 'humanoid',
            '-topic', 'robot_description',
            '-x', '0',
            '-y', '0',
            '-z', '1.0'
        ],
        output='screen'
    )
    
    robot_state_publisher = Node(
        package='robot_state_publisher',
        executable='robot_state_publisher',
        parameters=[{'robot_description': robot_description}]
    )
    
    return LaunchDescription([
        robot_state_publisher,
        spawn_entity
    ])

Testing Basic Motion

# Start Gazebo
gz sim worlds/empty.sdf

# Spawn robot
ros2 launch humanoid_gazebo spawn_humanoid.launch.py

# Command a joint
gz topic -t /model/humanoid/joint/right_ankle/cmd_pos -m gz.msgs.Double -p "data: 0.3"

1.6 Physics Properties

Contact Parameters

Critical for walking:

<gazebo reference="right_foot">
  <mu1>1.0</mu1>  <!-- Friction in primary direction -->
  <mu2>1.0</mu2>  <!-- Friction in secondary direction -->
  <kp>1000000.0</kp>  <!-- Contact stiffness -->
  <kd>1.0</kd>  <!-- Contact damping -->
  <minDepth>0.001</minDepth>  <!-- Minimum penetration for contact -->
</gazebo>

Self-Collision

<gazebo>
  <self_collide>true</self_collide>
</gazebo>

Summary

✅ Understand simulation advantages and limitations
✅ Install and use Gazebo Fortress
✅ Create world and robot descriptions
✅ Spawn robots and apply physics

Next: Chapter 2: Sensor Simulation