Skip to main content

Module 2 Assessment: The Digital Twin (Gazebo & Unity)

This assessment evaluates your understanding of simulation environments, including Gazebo, Unity, physics simulation, and sensor modeling for robotics applications.

Assessment Objectives​

By completing this assessment, you will demonstrate:

  • Proficiency in setting up and configuring Gazebo simulation environments
  • Understanding of physics simulation principles including gravity, collisions, and constraints
  • Ability to work with URDF and SDF for robot modeling in simulation
  • Knowledge of Unity integration for high-fidelity rendering
  • Skills in modeling and simulating various sensors (LiDAR, cameras, IMUs)
  • Competency in applying simulation techniques for robotics development

Part 1: Gazebo Environment Setup (20 points)​

Question 1.1 (5 points)​

List and briefly explain the five main components of Gazebo's architecture.

Question 1.2 (5 points)​

What is the difference between gzserver and gzclient? When would you use each?

Question 1.3 (5 points)​

Explain how to install Gazebo with ROS 2 integration on Ubuntu 22.04. Include the specific package names required.

Question 1.4 (5 points)​

Describe the purpose of the Gazebo-ROS bridge and explain how it enables communication between Gazebo and ROS 2.

Part 2: Physics Simulation (25 points)​

Question 2.1 (8 points)​

Calculate the moment of inertia for a cylindrical robot wheel with radius 0.1m, length 0.05m, and mass 0.5kg about its central axis. Show your work and explain why proper inertia values are important for simulation.

Question 2.2 (8 points)​

Explain the difference between static and dynamic friction in physics simulation. How would you configure these values in a Gazebo model for a rubber wheel on a concrete surface?

Question 2.3 (9 points)​

Compare and contrast the three physics engines supported by Gazebo (ODE, Bullet, DART). For each engine, describe:

  • Strengths
  • Weaknesses
  • Best use cases

Part 3: URDF and SDF (20 points)​

Question 3.1 (10 points)​

Write a complete URDF definition for a simple differential drive robot with:

  • Rectangular chassis (0.5m x 0.3m x 0.15m)
  • Two cylindrical wheels (radius 0.1m, length 0.05m)
  • Proper mass, inertial, visual, and collision properties
  • Continuous joints connecting wheels to chassis

Question 3.2 (10 points)​

Convert the following URDF snippet to equivalent SDF format:

<joint name="wheel_joint" type="continuous">
  <parent link="chassis"/>
  <child link="wheel"/>
  <origin xyz="0 0.2 0" rpy="0 0 0"/>
  <axis xyz="0 1 0"/>
</joint>

Part 4: Unity Rendering (15 points)​

Question 4.1 (8 points)​

Explain the advantages of using Unity for robotics simulation compared to traditional simulators like Gazebo. What types of applications benefit most from Unity's capabilities?

Question 4.2 (7 points)​

Describe the Unity-ROS bridge and explain how to establish communication between Unity and ROS 2. What are the key components required for this integration?

Part 5: Sensor Simulation (20 points)​

Question 5.1 (10 points)​

Configure a 3D LiDAR sensor in SDF with the following specifications:

  • 1024 horizontal samples
  • 64 vertical samples
  • Horizontal FOV: 360 degrees
  • Vertical FOV: 30 degrees (from -15° to +15°)
  • Range: 0.1m to 100m
  • Gaussian noise with 1cm standard deviation

Question 5.2 (10 points)​

Explain the process of sensor fusion in robotics. Describe how you would combine LiDAR and IMU data to improve robot localization. What are the advantages of using multiple sensors?

Practical Exercise (30 points)​

Exercise 6.1 (30 points)​

Create a complete simulation setup with the following components:

  1. Robot Model (10 points):

    • Create a URDF model of a simple wheeled robot with 2 driven wheels and 2 casters
    • Include proper mass, inertial, visual, and collision properties
    • Add a camera sensor and an IMU

  2. Gazebo World (10 points):

    • Create an SDF world file with a room containing obstacles
    • Include lighting and basic physics configuration
    • Place your robot in the world

  3. ROS Integration (10 points):

    • Create a launch file that starts Gazebo with your world
    • Launch the robot state publisher
    • Create a simple ROS node that subscribes to the camera and IMU data
    • Publish velocity commands to move the robot

Provide the complete code for each component and explain how they work together.

Answer Key and Scoring​

Part 1: Gazebo Environment Setup (20 points)​

  • Question 1.1: 5 points for explaining all 5 components (Gazebo Server, Client, Physics Engine, Rendering Engine, Sensor System)
  • Question 1.2: 5 points for distinguishing between gzserver (headless) and gzclient (GUI)
  • Question 1.3: 5 points for correct installation commands and package names
  • Question 1.4: 5 points for explaining the bridge functionality

Part 2: Physics Simulation (25 points)​

  • Question 2.1: 8 points (4 for calculation, 4 for explanation)
  • Question 2.2: 8 points (4 for definitions, 4 for configuration example)
  • Question 2.3: 9 points (3 points per engine: strengths, weaknesses, use cases)

Part 3: URDF and SDF (20 points)​

  • Question 3.1: 10 points for complete, valid URDF
  • Question 3.2: 10 points for correct SDF conversion

Part 4: Unity Rendering (15 points)​

  • Question 4.1: 8 points for advantages and use cases
  • Question 4.2: 7 points for bridge explanation

Part 5: Sensor Simulation (20 points)​

  • Question 5.1: 10 points for complete SDF sensor configuration
  • Question 5.2: 10 points for fusion explanation

Practical Exercise (30 points)​

  • Exercise 6.1: 30 points distributed across components (10 each)

Passing Criteria​

To pass this assessment, you must achieve:

  • Minimum 70% (70/100) overall score
  • Minimum 60% in each section
  • Complete implementation of the practical exercise

Learning Objectives Review​

This assessment covers:

  • Setting up and configuring Gazebo simulation environments ✓
  • Understanding physics simulation principles including gravity, collisions, and constraints ✓
  • Working with URDF and SDF for robot modeling in simulation ✓
  • Integrating Unity for high-fidelity rendering and visualization ✓
  • Modeling and simulating various sensors (LiDAR, cameras, IMUs) ✓
  • Applying simulation techniques for robotics development ✓

Answers to Part 1, Question 1.1 (Example Solution)​

The five main components of Gazebo's architecture are:

  1. Gazebo Server (gzserver): The headless simulation engine that handles physics, collision detection, and sensor simulation
  2. Gazebo Client (gzclient): The graphical user interface that provides visualization and user interaction
  3. Physics Engine: Handles rigid body dynamics, collisions, and contacts (ODE, Bullet, or DART)
  4. Rendering Engine: Manages 3D visualization using OGRE
  5. Sensor System: Simulates various sensor types (cameras, LiDAR, IMUs, etc.)

Answers to Part 2, Question 2.1 (Example Solution)​

For a solid cylinder rotating about its central axis: I = (1/2) * m * r² I = (1/2) * 0.5kg * (0.1m)² I = 0.25 * 0.5 * 0.01 I = 0.00125 kg⋅m²

Proper inertia values are crucial for realistic physics simulation. Incorrect values can lead to unstable simulation, unrealistic motion, or numerical errors. The moment of inertia affects how the object responds to applied torques and forces, determining its angular acceleration.

Answers to Part 3, Question 3.1 (Example Solution)​

<?xml version="1.0"?>
<robot name="diff_drive_robot">
  <!-- Chassis link -->
  <link name="chassis">
    <inertial>
      <mass value="1.0"/>
      <origin xyz="0 0 0"/>
      <inertia ixx="0.05" ixy="0" ixz="0" iyy="0.08" iyz="0" izz="0.1"/>
    </inertial>
    <visual>
      <origin xyz="0 0 0"/>
      <geometry>
        <box size="0.5 0.3 0.15"/>
      </geometry>
      <material name="grey">
        <color rgba="0.5 0.5 0.5 1"/>
      </material>
    </visual>
    <collision>
      <origin xyz="0 0 0"/>
      <geometry>
        <box size="0.5 0.3 0.15"/>
      </geometry>
    </collision>
  </link>

  <!-- Left wheel -->
  <link name="left_wheel">
    <inertial>
      <mass value="0.2"/>
      <origin xyz="0 0 0"/>
      <inertia ixx="0.001" ixy="0" ixz="0" iyy="0.001" iyz="0" izz="0.002"/>
    </inertial>
    <visual>
      <geometry>
        <cylinder radius="0.1" length="0.05"/>
      </geometry>
      <material name="black">
        <color rgba="0 0 0 1"/>
      </material>
    </visual>
    <collision>
      <geometry>
        <cylinder radius="0.1" length="0.05"/>
      </geometry>
    </collision>
  </link>

  <!-- Right wheel -->
  <link name="right_wheel">
    <inertial>
      <mass value="0.2"/>
      <origin xyz="0 0 0"/>
      <inertia ixx="0.001" ixy="0" ixz="0" iyy="0.001" iyz="0" izz="0.002"/>
    </inertial>
    <visual>
      <geometry>
        <cylinder radius="0.1" length="0.05"/>
      </geometry>
      <material name="black">
        <color rgba="0 0 0 1"/>
      </material>
    </visual>
    <collision>
      <geometry>
        <cylinder radius="0.1" length="0.05"/>
      </geometry>
    </collision>
  </link>

  <!-- Caster wheels -->
  <link name="front_caster">
    <inertial>
      <mass value="0.05"/>
      <origin xyz="0 0 0"/>
      <inertia ixx="0.0001" ixy="0" ixz="0" iyy="0.0001" iyz="0" izz="0.0001"/>
    </inertial>
    <visual>
      <geometry>
        <sphere radius="0.05"/>
      </geometry>
      <material name="silver">
        <color rgba="0.7 0.7 0.7 1"/>
      </material>
    </visual>
    <collision>
      <geometry>
        <sphere radius="0.05"/>
      </geometry>
    </collision>
  </link>

  <!-- Joints -->
  <joint name="left_wheel_joint" type="continuous">
    <parent link="chassis"/>
    <child link="left_wheel"/>
    <origin xyz="0 0.2 0" rpy="-1.5708 0 0"/>
    <axis xyz="0 0 1"/>
  </joint>

  <joint name="right_wheel_joint" type="continuous">
    <parent link="chassis"/>
    <child link="right_wheel"/>
    <origin xyz="0 -0.2 0" rpy="-1.5708 0 0"/>
    <axis xyz="0 0 1"/>
  </joint>

  <joint name="front_caster_joint" type="ball">
    <parent link="chassis"/>
    <child link="front_caster"/>
    <origin xyz="0.2 0 -0.05"/>
  </joint>

  <!-- Camera sensor -->
  <joint name="camera_joint" type="fixed">
    <parent link="chassis"/>
    <child link="camera_link"/>
    <origin xyz="0.25 0 0.1"/>
  </joint>

  <link name="camera_link">
    <inertial>
      <mass value="0.01"/>
      <inertia ixx="0.0001" ixy="0" ixz="0" iyy="0.0001" iyz="0" izz="0.0001"/>
    </inertial>
  </link>

  <!-- IMU sensor -->
  <joint name="imu_joint" type="fixed">
    <parent link="chassis"/>
    <child link="imu_link"/>
    <origin xyz="0 0 0.05"/>
  </joint>

  <link name="imu_link">
    <inertial>
      <mass value="0.01"/>
      <inertia ixx="0.0001" ixy="0" ixz="0" iyy="0.0001" iyz="0" izz="0.0001"/>
    </inertial>
  </link>
</robot>

Answers to Part 3, Question 3.2 (Example Solution)​

<joint name="wheel_joint" type="revolute">
  <parent>chassis</parent>
  <child>wheel</child>
  <pose>0 0.2 0 0 0 0</pose>
  <axis>
    <xyz>0 1 0</xyz>
    <use_parent_model_frame>false</use_parent_model_frame>
  </axis>
</joint>

Answers to Part 5, Question 5.1 (Example Solution)​

<sensor name="3d_lidar" type="ray">
  <pose>0.3 0 0.5 0 0 0</pose>
  <ray>
    <scan>
      <horizontal>
        <samples>1024</samples>
        <resolution>1</resolution>
        <min_angle>-3.14159</min_angle>
        <max_angle>3.14159</max_angle>
      </horizontal>
      <vertical>
        <samples>64</samples>
        <resolution>1</resolution>
        <min_angle>-0.2618</min_angle> <!-- -15 degrees -->
        <max_angle>0.2618</max_angle>   <!-- 15 degrees -->
      </vertical>
    </scan>
    <range>
      <min>0.1</min>
      <max>100.0</max>
      <resolution>0.01</resolution>
    </range>
  </ray>
  <noise type="gaussian">
    <mean>0.0</mean>
    <stddev>0.01</stddev>
  </noise>
</sensor>

Answers to Part 5, Question 5.2 (Example Solution)​

Sensor fusion combines data from multiple sensors to improve the accuracy and reliability of the system. For LiDAR and IMU fusion:

  • LiDAR provides accurate position and mapping data but at lower frequency and can be affected by environmental conditions
  • IMU provides high-frequency motion data but suffers from drift over time

Fusion typically uses filtering techniques like Kalman filters or particle filters to combine the strengths of both sensors. The IMU provides high-frequency motion updates while LiDAR corrects for drift and provides absolute position references.

Advantages of multi-sensor fusion:

  • Improved accuracy through redundancy
  • Increased robustness against sensor failures
  • Better temporal and spatial resolution
  • Compensation for individual sensor limitations

Practical Exercise Solution Outline​

For the practical exercise, you would need to create:

  1. Robot URDF File (diff_drive_robot.urdf):
<!-- Provided in Question 3.1 solution -->
  1. Gazebo World File (simple_room.world):
<?xml version="1.0" ?>
<sdf version="1.7">
  <world name="simple_room">
    <physics name="default_physics" type="ode">
      <max_step_size>0.001</max_step_size>
      <real_time_factor>1</real_time_factor>
      <gravity>0 0 -9.8</gravity>
    </physics>

    <include>
      <uri>model://ground_plane</uri>
    </include>

    <include>
      <uri>model://sun</uri>
    </include>

    <!-- Room walls -->
    <model name="wall_1">
      <pose>-3 0 1 0 0 1.57</pose>
      <link name="wall_link">
        <visual name="visual">
          <geometry>
            <box><size>6 0.2 2</size></box>
          </geometry>
          <material>
            <ambient>0.8 0.8 0.8 1</ambient>
            <diffuse>0.8 0.8 0.8 1</diffuse>
          </material>
        </visual>
        <collision name="collision">
          <geometry>
            <box><size>6 0.2 2</size></box>
          </geometry>
        </collision>
      </link>
    </model>

    <!-- Similar for other walls -->

    <!-- Include robot -->
    <include>
      <uri>model://diff_drive_robot</uri>
      <pose>0 0 0.1 0 0 0</pose>
    </include>
  </world>
</sdf>
  1. Launch File (simulation.launch.py):
from launch import LaunchDescription
from launch.actions import DeclareLaunchArgument
from launch.substitutions import LaunchConfiguration
from launch_ros.actions import Node
from launch.actions import IncludeLaunchDescription
from launch.launch_description_sources import PythonLaunchDescriptionSource
from ament_index_python.packages import get_package_share_directory
import os

def generate_launch_description():
    # Gazebo launch
    gazebo = IncludeLaunchDescription(
        PythonLaunchDescriptionSource(
            os.path.join(get_package_share_directory('gazebo_ros'), 'launch', 'gazebo.launch.py')
        )
    )

    # Robot spawn node
    spawn_entity = Node(
        package='gazebo_ros',
        executable='spawn_entity.py',
        arguments=['-topic', 'robot_description', '-entity', 'diff_drive_robot'],
        output='screen'
    )

    # Robot state publisher
    robot_state_publisher = Node(
        package='robot_state_publisher',
        executable='robot_state_publisher',
        name='robot_state_publisher',
        output='screen'
    )

    # Sensor processing node
    sensor_processor = Node(
        package='my_robot_package',
        executable='sensor_processor',
        name='sensor_processor',
        output='screen'
    )

    return LaunchDescription([
        gazebo,
        spawn_entity,
        robot_state_publisher,
        sensor_processor
    ])
  1. ROS Node for Sensor Processing (sensor_processor.py):
import rclpy
from rclpy.node import Node
from sensor_msgs.msg import Image, Imu
from geometry_msgs.msg import Twist

class SensorProcessor(Node):
    def __init__(self):
        super().__init__('sensor_processor')

        # Subscribe to sensors
        self.camera_sub = self.create_subscription(Image, '/camera/image_raw', self.camera_callback, 10)
        self.imu_sub = self.create_subscription(Imu, '/imu/data', self.imu_callback, 10)

        # Publisher for robot control
        self.cmd_vel_pub = self.create_publisher(Twist, '/cmd_vel', 10)

        self.get_logger().info('Sensor processor initialized')

    def camera_callback(self, msg):
        self.get_logger().info(f'Received camera image: {msg.width}x{msg.height}')

    def imu_callback(self, msg):
        self.get_logger().info(f'Received IMU data: {msg.linear_acceleration.x:.2f} m/s²')

def main(args=None):
    rclpy.init(args=args)
    processor = SensorProcessor()
    rclpy.spin(processor)
    processor.destroy_node()
    rclpy.shutdown()

if __name__ == '__main__':
    main()

Next Steps​

After completing this assessment, proceed to Module 3: The AI-Robot Brain to learn about NVIDIA Isaac platform and AI integration in robotics.

Additional Resources​

Assessment Completion​

When you have completed this assessment:

  1. Review all questions and ensure you've addressed each requirement
  2. Test your practical exercise code in a simulation environment
  3. Verify that all components work together as expected
  4. Document any challenges you encountered and how you resolved them

This completes Module 2: The Digital Twin (Gazebo & Unity). You should now have a solid understanding of simulation environments and be ready to explore AI integration in robotics with Module 3.