Defesa de Mestrado – Marcelo Elias Simon – 11/2/2022
17/01/2022 14:44
| Defesa de Dissertação de Mestrado | |
| Aluno | Marcelo Elias Simon |
| Orientador
Coorientador Coorientador |
Prof. Max Hering de Queiroz, Dr. – DAS/UFSC
Prof. Felipe Gomes de Oliveira Cabral, Dr. – DAS/UFSC Prof. Fábio Baldissera, Dr. |
| Data | 11/2/2022 (sexta-feira) – 14h30
Videoconferência (https://meet.google.com/yod-ybfa-crn) |
| Banca | Prof. Max Hering de Queiroz, Dr. – DAS/UFSC (presidente);
Prof. Sandro Battistella, Dr. – CECE/UINIOESTE; Prof. Ubirajara Franco Moreno, Dr. – DAS/UFSC; Prof. Rodrigo Castelan Carlson, Dr. – DAS/UFSC. |
| Título | Multi-Robots Coordination System for Urgan Search and Rescue Post-Disaster Assistance based on Supervisory Control Theory |
| Abstract: The application of robots in Urban Search and Rescue (USAR) is a domain receiving increasing attention from the scientific community. Although it does not yet grant the robustness, reliability, and safety required for field implementation. In this work, we propose a Multi-robots Coordination System (MCS) that implements the interaction between reactive and deliberative layers with the objective of increasing the safety and reliability assigned to USAR robots. Formal models were used to model robot plants and constraints, allowing us to synthesize supervisors using the Supervisory Control Theory (SCT). By implementing these supervisors on the control system of each robot, we formulate the reactive layer in such a way that the robot will attain the designed specifications. The deliberative layer, represented by the Task Dispatcher (TD) component, is in charge of defining the best allocation of tasks to the robots based on their current status and the required sequences of tasks (missions) assigned by a human. The full integration of the MCS components was implemented in the Robotic Operating System (ROS) with the purpose of validating the proposed architecture in a simulated environment. A scenario composed of two ground robots and two aerial robots applied to a post-disaster scene was designed and sequences of tasks were assigned according to international USAR agency common procedures. As a result of the proposed architecture, we have accomplished a reliable and safe USAR robots system that has the behaviors controlled by means of mathematically proved formalisms. Besides, the reactive layer showed the potential of being scalable since it did not require a lot of effort from the robot. Therefore, the proposed system may increase the USAR community’s acceptance on the use of robots, enabling the actual implementation of already developed technologies. | |
