Framework for Intelligent Data Exchange in Cyber Physical Systems

• Autor(a) principal: António Miguel Silva Pereira
• Data de Publicação: 2022
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: https://hdl.handle.net/10216/140756
• topic: Engenharia electrotécnica, electrónica e informática; Electrical engineering, Electronic engineering, Information engineering

In a shop-floor there exists a production line of several controllers in which their aim is to perform welding. These controllers must all communicate with each other to keep the production active. The problem lies when there is a need of retrieving data that these controllers have generated during their work where they save their welding value and timestamp. They are usually in high number and each with plenty of information due to the fact that data is saved each millisecond. Because the controllers must be working all the time to ensure maximum production efficiency, any downtime while uploading information is detrimental to the system. Through a developed platform called DINASORE (Dynamic INtelligent Architecture for Software and MOdular REconfiguration), we will build a simulation model where we will try different techniques to minimize the downtime of the system while the uploading happens. This simulated model will be emulated with the help of the usage of Raspberry Pi emulating the controllers performance replicating the shop-floor production line. With this model we will attempt several methods between distributed database and self-organizing to reach a solution where we can request the upload without crashing the system. The metrics we will use to compare the results of the different approaches will be 1) the acknowledgement of all the information requested arriving to the end-point, 2) the delta-time between the request and the information arriving and 3) the average communication time between the controllers while uploading and not uploading data. The expected result of this simulated experiment will be a framework where it will be possible to test several different approaches towards managing data intensive applications with clear cut metrics to be able to give a comparison between them.