Institutional Repository
Technical University of Crete
EN  |  EL



My Space

Vision-Based autonomous robotic arm for Pick-and-Stack applications in the Gazebo-ROS environment

Theodorakis Charalampos

Full record

Year 2022
Type of Item Diploma Work
Bibliographic Citation Charalampos Theodorakis, "Vision-Based autonomous robotic arm for Pick-and-Stack applications in the Gazebo-ROS environment", Diploma Work, School of Electrical and Computer Engineering, Technical University of Crete, Chania, Greece, 2022
Appears in Collections


In recent years, the use of robots as a part of the automation process is a rapidly expanding field. The most prevalent type of robot in industrial environments is robotic arms. This type of robots is capable of performing a variety of operations, including transportation, assembly, packing and welding. The complexity of the tasks assumed for these growing disciplines is also rising. Thus, the ability of the robot to comprehend the surrounding environment is a required feature to carry out new tasks. Interacting with the environment and being aware of the changing conditions enables the robot to make autonomous decisions in more complex situations. This thesis describes the implementation of an autonomous 6-DOF (Degrees Of Freedom) robotic manipulator with visual guidance, where the objective for the robot arm is to stack cylinder blocks, recreating a gradually-presented structured tower pattern. The available blocks, initially placed randomly in the robot workspace, have unique ArUco markers displayed on them, so they can be identified. The manipulator is equipped with vacuum grippers and a 2D camera attached to its end effector. The robot acquires information about the desired tower pattern from a remote stationary camera, responsible to track the spawning of building blocks. Since the robotic arm is equipped with its own camera, it employs computer vision techniques to locate the desired cylinder and, consequently, constructs a trajectory to pick and stack that cylinder to its correct position using motion planning algorithms. The entire project has been implemented within the Robot Operating System (ROS) and Gazebo open-source 3D robotics simulator. The proposed robotic system has been tested extensively in simulations to ensure its reliability and investigate its efficiency.

Available Files