Frazão, Luís Alexandre LopesCosta, Nuno Alexandre Ribeiro daRamos, João Pedro FerreiraPereira, António Manuel de JesusBernarda, Luis de Castro Soveral da2024-03-152023-11-07http://hdl.handle.net/10400.8/9534Cloud-based solutions are among the most widely proposed solutions to address the challenges inherent in modern robotics. Cloud robotics is a field of robotics that leverages cloud-based technologies and concepts such as cloud computing or storage to enhance robotic solutions while also driving down the cost of each individual robot. By employing containerization technology, robots can offload computationally intensive tasks to the cloud while ensuring low-latency communication through edge devices. This approach aims to enhance scalability, resource efficiency, and real-time responsiveness in cloud-edge robotics applications. The use of such technologies facilitates dynamic assignment of tasks to decentralized heterogeneous robotic fleets. To enable this, an architecture is proposed where a container orchestration system is configured across nodes in the cloud and edge, including robotic end nodes. This architecture is also capable of addressing many of the challenges faced by resourceconstrained robots and can facilitate communication avenues between robotic end nodes that cannot be integrated with the cluster. Several tests are carried out and metrics are collected and analysed that enable insights into master node resource provisioning and the advantages of a cloud-edge based architecture. Other functional tests are also conducted and elaborated upon, including a prototype integrating multiple projects. The test results demonstrate that the architecture is valid and can serve as a foundation for further work in the field.engRobóticaRobótica em nuvemArmazenamento em nuvenTecnologia de virtualizaçãoComputação móvelmaster thesis203553683