RL24_HW_3

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

Overview

This is Homework 3, where the objective is to implement a vision-based controller for a 7-degrees-of-freedom robotic manipulator arm within the Gazebo simulation environment.

---

Instructions

1. Clone the repository from GitHub:

   cd src
   git clone -b REV_3 https://github.com/ferd-bot/RL_24_Homework_3_Robotics.git .

   Important:
   The git clone command with the dot (.) works only if the target directory is empty.

   • If not, you can remove extra files using:

     rm -rf *

   • Alternatively, clone the repository normally (without the dot) and manually move the files from the RL_24_Homework_3_Robotics folder to the src directory.

2. Configure and build all packages in the ROS2 workspace:

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

---

Launching the Simulation in Gazebo

Start the IIWA Robot with Position Controller and Camera Enabled

To launch the Gazebo simulation with the Position controller and enable the camera:

ros2 launch iiwa_bringup iiwa.launch.py use_sim:="true" use_vision:="true"

• Note: If the camera is not required, set use_vision:=false:

  ros2 launch iiwa_bringup iiwa.launch.py use_sim:="true" use_vision:="false"

---

Start the IIWA Robot with Position Controller, Camera, and Spherical Object Enabled

To launch the Gazebo simulation with the Position controller, camera, and spherical object:

ros2 launch iiwa_bringup iiwa.launch.py use_sim:="true" use_vision:="true" spherical_object:="true"

To move the manipulator toward the sphere (position control), open a new terminal and run:

ros2 topic pub /iiwa_arm_controller/commands std_msgs/msg/Float64MultiArray "data: [1.7, 2.0, 0.0, 2.0, 0.0, -1.5, 0.0]"

---

OpenCV Integration

Launch the OpenCV Node

To use OpenCV functionalities, run the following command in a new terminal:

ros2 run ros2_opencv ros2_opencv_node

---

To visualize the image processed by OpenCV, open rqt in another terminal and subscribe to the topic /processed_image:

rqt

---

ArUco Marker Integration

To launch Gazebo in velocity mode with ArUco tags enabled:

ros2 launch iiwa_bringup iiwa.launch.py command_interface:="velocity" robot_controller:="velocity_controller" use_sim:="true" use_vision:="true" aruco_tag:="true"

This command also opens RViz with a pre-configured setup for the aruco_single node to detect the marker.

• To perform the Positioning Task in velocity mode, open another terminal and run:

  ros2 run ros2_kdl_package ros2_kdl_node --ros-args -p cmd_interface:=velocity -p task:=positioning

• To perform the Look-at-Point Task in velocity mode, open another terminal and run:

  ros2 run ros2_kdl_package ros2_kdl_node --ros-args -p cmd_interface:=velocity -p task:=look-at-point

To visualize the image of Aruco, open rqt in another terminal and subscribe to the topic /aruco_detect/result:

rqt

---

Executing Trajectories in Effort Mode

You can execute the four trajectories defined in the previous homework using effort mode.
To start Gazebo in effort mode:

ros2 launch iiwa_bringup iiwa.launch.py command_interface:="effort" robot_controller:="effort_controller" use_sim:="true" use_vision:="true" aruco_tag:="true"

To execute the trajectories, open a new terminal and run one of the following commands depending on the desired control mode:

• Effort Mode in Joint Space:

  ros2 run ros2_kdl_package ros2_kdl_node 0 --ros-args -p cmd_interface:=effort -p control_space:=joint_space

• Effort Mode in Operational Space:

  ros2 run ros2_kdl_package ros2_kdl_node 0 --ros-args -p cmd_interface:=effort -p control_space:=operational_space

To visualize the image of Aruco, open rqt in another terminal and subscribe to the topic /aruco_detect/result:

rqt

---

Available Trajectories

The following trajectories are supported, all controlled in Effort Mode with Joint Space as the default.
Trajectories are numbered from 0 to 3:

1. Linear with Trapezoidal Velocity Profile:

   ros2 run ros2_kdl_package ros2_kdl_node 0 --ros-args -p cmd_interface:=effort -p control_space:=joint_space

2. Linear with Cubic Velocity Profile:

   ros2 run ros2_kdl_package ros2_kdl_node 1 --ros-args -p cmd_interface:=effort -p control_space:=joint_space

3. Circular with Trapezoidal Velocity Profile:

   ros2 run ros2_kdl_package ros2_kdl_node 2 --ros-args -p cmd_interface:=effort -p control_space:=joint_space

4. Circular with Cubic Velocity Profile:

   ros2 run ros2_kdl_package ros2_kdl_node 3 --ros-args -p cmd_interface:=effort -p control_space:=joint_space

To execute trajectories in Operational Space, simply set control_space:=operational_space.

---

Additional Notes

A new topic /torque_plot has been added.
This topic allows you to visualize the torques applied to the manipulator's joints.

To monitor the topic in the terminal:

ros2 topic echo /torque_plot

Alternatively, use rqt_plot for graphical visualization:

rqt

In rqt, set up the plotting plugin and add the following data points for each joint:

/torque_plot/data[0]
/torque_plot/data[1]
/torque_plot/data[2]
/torque_plot/data[3]
/torque_plot/data[4]
/torque_plot/data[5]
/torque_plot/data[6]

Additional Topics:

• Visualize the detected ArUco pose:

  ros2 topic echo /aruco_detect/pose

---

Videos

For simplicity, only videos demonstrating Effort-Based Control in Operational Space with Gazebo are included:

• Effort mode - Joint Space
• Effort mode - Operational Space
• Velocity mode - Positioning
• Velocity mode - Look at Point