Skip to main content

Chapter 2: Gazebo — Physics, Gravity, Collisions, and Sensors

Introduction to Gazebo Physics​

Gazebo provides a highly configurable physics engine to accurately simulate robot dynamics and interactions with their environment. Understanding these physics properties is fundamental to creating realistic digital twins.

Gravity​

Gravity is a crucial environmental parameter in any physics simulation. In Gazebo, you can define the direction and magnitude of the gravitational force acting on all objects in the simulation world.

Configuring Gravity​

Gazebo worlds are defined in SDF (Simulation Description Format) files. Gravity is typically set within the <physics> tag of the world file.

<physics name="default_physics" default="0" type="ode">
  <ode>
    <solver>
      <type>quick</type>
      <iters>50</iters>
      <world_cfm>0</world_cfm>
      <world_erp>0.2</world_erp>
    </solver>
    <constraints>
      <cfm>0</cfm>
      <erp>0.2</erp>
    </constraints>
  </ode>
  <gravity>0 0 -9.8</gravity> <!-- Standard Earth gravity in Z-direction -->
  <magnetic_field>6e-06 2.3e-05 -4.2e-05</magnetic_field>
  <max_step_size>0.001</max_step_size>
</physics>

Code Example 2.1: Configuring gravity in a Gazebo SDF world file.

Collisions​

Accurate collision detection and response are vital for robots to interact realistically with their surroundings. Gazebo allows you to define collision shapes for each link of a robot and configure their physical properties.

Collision Shapes and Properties​

Each link in a robot model (URDF or SDF) can have one or more <collision> elements. Inside a collision element, you define a geometric <geometry> (box, sphere, cylinder, mesh) and <surface> properties that control how it behaves during contact.

<link name="base_link">
  <collision name="base_collision">
    <geometry>
      <box>
        <size>0.2 0.2 0.5</size>
      </box>
    </geometry>
    <surface>
      <friction>
        <ode>
          <mu>1.0</mu>    <!-- Coefficient of friction -->
          <mu2>1.0</mu2>   <!-- Secondary coefficient of friction -->
          <slip1>0.0</slip1>
          <slip2>0.0</slip2>
        </ode>
      </friction>
      <bounce>
        <restitution_coefficient>0.1</restitution_coefficient> <!-- Bounciness -->
        <threshold>0.0</threshold>
      </bounce>
      <contact>
        <ode>
          <kd>1000000.0</kd> <!-- Spring stiffness -->
          <kp>1000000.0</kp> <!-- Damping constant -->
        </ode>
      </contact>
    </surface>
  </collision>
  ...
</link>

Code Example 2.2: Defining collision properties in a Gazebo SDF link.

  • mu / mu2: Coefficients of friction (primary and secondary).
  • restitution_coefficient: Determines how "bouncy" a collision is (0.0 for no bounce, 1.0 for perfect bounce).
  • kd / kp: Contact stiffness and damping parameters.

Sensor Simulation​

Gazebo offers a rich set of simulated sensors that mimic their real-world counterparts, providing realistic data for robot perception.

Conceptual Diagram of Gazebo Sensor Simulation​

Conceptual Diagram of Gazebo Sensor Simulation Figure 2.1: A conceptual diagram illustrating how various sensors (LiDAR, Camera, IMU) are simulated in Gazebo and generate data streams. (Note: An actual image would be embedded here, showing sensors attached to a robot model in Gazebo.)

LiDAR (Laser Range Finder) Simulation​

LiDAR sensors measure distances to objects by emitting laser pulses. Gazebo can simulate 2D and 3D LiDAR, generating realistic point cloud data.

<sensor name="laser_sensor" type="ray">
  <pose>0.1 0 0.2 0 0 0</pose>
  <ray>
    <scan>
      <horizontal>
        <samples>640</samples>
        <resolution>1</resolution>
        <min_angle>-1.57</min_angle>
        <max_angle>1.57</max_angle>
      </horizontal>
    </scan>
    <range>
      <min>0.1</min>
      <max>10.0</max>
      <resolution>0.01</resolution>
    </range>
  </ray>
  <always_on>1</always_on>
  <update_rate>30</update_rate>
  <visualize>true</visualize>
</sensor>

Code Example 2.3: Configuring a 2D LiDAR sensor in Gazebo.

  • ray type: Indicates a LiDAR (laser ray) sensor.
  • horizontal / vertical: Defines the scan properties (samples, angle range).
  • range: Minimum, maximum, and resolution of distance measurements.

Depth Camera Simulation​

Depth cameras provide an image where each pixel's value represents the distance from the camera to the corresponding point in the scene.

<sensor name="depth_camera" type="depth">
  <pose>0.05 0 0.3 0 0 0</pose>
  <camera>
    <horizontal_fov>1.047</horizontal_fov>
    <image>
      <width>640</width>
      <height>480</height>
      <format>R8G8B8</format>
    </image>
    <clip>
      <near>0.1</near>
      <far>10</far>
    </clip>
  </camera>
  <always_on>1</always_on>
  <update_rate>30</update_rate>
  <visualize>false</visualize>
</sensor>

Code Example 2.4: Configuring a depth camera sensor in Gazebo.

IMU (Inertial Measurement Unit) Simulation​

IMUs measure linear acceleration and angular velocity, providing crucial data for robot localization and state estimation.

<sensor name="imu_sensor" type="imu">
  <pose>0 0 0.1 0 0 0</pose>
  <imu>
    <orientation>
      <x>0</x> <y>0</y> <z>0</z>
    </orientation>
    <angular_velocity>
      <x>0</x> <y>0</y> <z>0</z>
    </angular_velocity>
    <linear_acceleration>
      <x>0</x> <y>0</y> <z>0</z>
    </linear_acceleration>
  </imu>
  <always_on>1</always_on>
  <update_rate>100</update_rate>
</sensor>

Code Example 2.5: Configuring an IMU sensor in Gazebo.

Summary​

This chapter explored the fundamental aspects of physics simulation within Gazebo, covering how to configure gravity and define realistic collision properties for robot links. We also delved into simulating various crucial sensors—LiDAR, depth cameras, and IMUs—demonstrating their configuration and the types of data they generate, which are essential for developing robust robot perception systems in digital twins.

Self-Assessment Questions​

  1. How is gravity typically configured in a Gazebo world, and what file format is used for this configuration?
  2. Explain the purpose of mu and restitution_coefficient within Gazebo's collision surface properties. How do they affect simulated interactions?
  3. Describe the key parameters you would configure for a 2D LiDAR sensor in Gazebo, including its scan and range properties.
  4. What type of data does a depth camera provide in a Gazebo simulation, and how is this different from a standard RGB camera?
  5. Why is IMU data critical for robot localization and state estimation, and what physical quantities does a simulated IMU measure?

References​

[1] Gazebo. (n.d.). SDF Format. Retrieved from https://gazebosim.org/docs/garden/sdf [2] Gazebo. (n.d.). Sensors. Retrieved from https://gazebosim.org/docs/garden/sensors