Behavior Tree-Based Task Planning for UR5 robot

Behavior Trees are frequently used to construct robot actions, allowing complex commands and actions to be gradually built from simpler behaviors. This thesis aims to develop a framework for planning and executing tasks based on Behavior Trees for the UR5 robot, implementing the Gazebo and ROS simulation environment. The idea is to design and implement some elementary actions (for example: reach a target pose with the robot), which will provide the building blocks for other control methods. These actions are initially developed and tested independently, in order to guarantee their optimal function. This provides certainty and ease in integrating these functions into the Behavior Tree for UR5. The evaluation is carried out through simulations to see the suitability and effectiveness for later implementation in the final structure.