Controle orientado a eventos de uma fábrica inteligente baseado em árvores de comportamento e numa estrutura de programação distribuída.

Abstract

This thesis presents the development of an event-driven control system for a smart fac- tory using behavior trees integrated with the ROS framework. The implementation was

carried out in an experimental environment that includes the UR10 robotic manipula- tor, the TurtleBot2i mobile robot, and vision sensors such as the Kinect camera. The

project focuses on the autonomous coordination of these devices for object manipulation and transportation tasks, utilizing ArUco marker detection and real-time adaptation to

environmental conditions. The methodology involves designing a behavior tree-based su- pervisor that manages the UR10 operations based on object location data provided by

the Kinect and transportation conducted by the TurtleBot2i. Results demonstrate the systemŠs ability to perform tasks in a synchronized and efficient manner, with Ćexibility to integrate new devices and adapt to various production scenarios. Future work suggests implementing autonomous navigation for the TurtleBot2i, dynamic object sectorization based on visual identiĄcation, and the development of a digital twin for simulation and enhancement of UR10Šs kinematic control.