Defesa de Mestrado – Afonso da Fonseca Braga – 21/12/2021
09/12/2021 09:37
| Defesa de Dissertação de Mestrado | |
| Aluno | Afonso da Fonseca Braga |
| Orientador
Coorientador |
Prof. Edson Roberto De Pieri, Dr. – DAS/UFSC
Profa. Patrícia Della Méa Plentz, Dra. – INE/UFSC |
| Data | 21/12/2021 (terça-feira) – 9h
Videoconferência (https://meet.google.com/khw-qyuf-twq) |
| Banca | Prof. Edson Roberto De Pieri, Dr. – DAS/UFSC (presidente);
Prof. Mário Antônio Ribeiro Dantas, Dr. – UFJF; Prof. Leandro Buss Becker, Dr. – DAS/UFSC; Profa. Eliane Pozzebon, Dra. – UFSC/Araranguá. |
| Título | HeMuRo Framework: Development of a Generic Framework for Multi-Robot Systems |
| Abstract: Mobile robotics presents as an alternative due to the lack of mobility of robotic manipulators. Biological behaviors inspired many ways of locomotion increasing the usability of robots in multiple areas, adding robots in academic environments and also in environments that are harmful to humans. Over time, new tools emerged to simplify the use of robotic systems. ROS has a collection of open-source tools and algorithms to help engineers build robotic systems, featuring robot control, mapping, localization, and navigation for mobile robots. However, adding more functionalities to a mobile robot impacts higher costs and higher energy consumption. Therefore, a balance between functionality and feasibility is needed. Multi-Robot System (MRS) are systems where it is possible to combine different types of robots with multiple abilities to perform missions, increasing the efficiency and robustness of the system. This way it is possible to use multiple robots with specific abilities, instead of using a single robot with all the sensors and functionalities. New issues are introduced such as, for example, task allocation and task decomposition taking into consideration each type of robot, robot coordination to execute missions, and also the ability to reallocate missions in case of failure. This work presents a framework responsible for task allocation and decomposition of missions for heterogeneous robots. With the main goal to be open-source and flexible, HeMuRo Framework was built with a modular and distributed architecture allowing modification and improvements. With simple missions as input, the framework performs mission decomposition and task allocation using an auction algorithm taking into consideration battery level, time to execute the mission, and if the robot has the capability of executing the mission. In case of failure or not being able to finish the mission, there is also the possibility to reallocate to another robot. This framework works independently but there is also the possibility of interaction with ROS to communicate with real robots or simulated environments. Simulations were also conducted, involving different scenarios such as hospital and logistics warehouse. HeMuRo Framework applied versatility to decompose different types of results, obtaining graphical information for analysis. | |
