Homework 4: Robotics Instructions

Contributors: Ferdinando Dionisio, Vittorio Lisi, Giovanni Gabriele Imbimbo, Emanuele Cifelli

Overview

This guide provides step-by-step instructions for working with the robotics package. It covers Gazebo simulation, navigation tasks, mapping and localization, and vision-based navigation. Follow these instructions to ensure a successful setup and operation.

---

Setup Instructions

1. Clone the Repository

Navigate to the src directory of your ROS2 workspace and execute:

cd src
git clone https://github.com/ferd-bot/RL_24_Homework_4_Robotics.git .

Important:
The above command (git clone with a dot .) works only if the target directory is empty. If it's not, you can:

1. Remove all files in the directory:

   rm -rf *

2. Alternatively, clone the repository without the dot and manually move the contents of the RL_24_Homework_4_Robotics folder into the src directory.

2. Configure and Build the Workspace

Navigate to your ROS2 workspace, clean previous builds, and rebuild:

cd ~/ros2_ws
rm -rf build/ install/ log/
colcon build
source install/setup.bash

---

Simulation and Navigation

1. Start the Simulation in Gazebo

Launch the Gazebo simulation environment:

ros2 launch rl_fra2mo_description gazebo_fra2mo.launch.py

2. Autonomous Exploration and Mapping

Mode 1: Autonomous Exploration

Start autonomous exploration by launching:

ros2 launch rl_fra2mo_description fra2mo_explore.launch.py

This will activate the autonomous exploration node and, after a brief delay, Rviz will open with the "explore.rviz" configuration. Once the robot completes the exploration, run the waypoint-following script:

ros2 run rl_fra2mo_description follow_waypoints.py

• Waypoints:
  The waypoints are specified in the new_goals.yaml file located in the config directory.
  To use the waypoints from point 2 of Homework, replace new_goals.yaml with goals.yaml in the script.

• Environment:
  The navigation operates in the leonardo_race_field_new.sdf world with the mappa_mondo_1.pgm map.

---

3. Vision-based Navigation

For vision-based navigation using AMCL, follow these steps:

1. Start the Gazebo simulation:

   ros2 launch rl_fra2mo_description gazebo_fra2mo.launch.py

2. Launch the vision-based navigation node, which includes AMCL, RViz, and SLAM-based localization:

   ros2 launch rl_fra2mo_description fra2mo_navigation_vision.launch.py

3. Once all nodes are active, start the task script for approaching an obstacle, detecting a marker, and returning to the initial position (for Aruco TF publish, check "Aruco TF Frame":

   ros2 run rl_fra2mo_description task.py

4. To visualize Aruco marker detection, open rqt in another terminal and subscribe to the /aruco_detect/result topic:

   rqt

---

4. Aruco TF Frame

To visualize transformation frames during the simulation:

1. While running the task.py script, publish and print the Aruco pose using:

   ros2 run rl_fra2mo_description aruco_tf

   This command publishes the Aruco pose in the global frame.

2. To read the tag pose relative to the camera:

   ros2 topic echo /aruco_detect/pose

Note: These topics can be monitored while running the task.py script during the Aruco detection task. This allows you to observe transformations and validate marker tracking dynamically within the simulation.

---

Additional Notes

1. Multiple Terminals:
   Run each command or node in a separate terminal after sourcing the workspace:

   source install/setup.bash

2. Video Demonstrations:

   • Autonomous Navigation with New Goals
   • Vision-based Navigation

3. Troubleshooting:

   • Ensure all required nodes are running.
   • Check terminal outputs for error messages.

---

By following these instructions, you can perform advanced robotics tasks such as simulation, autonomous navigation, exploration, and vision-based control for Homework 4.