| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 5.91 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
A indústria 4.0, também denominada como IIoT, através da implementação massiva de sensores e dispositivos IoT, leva à produção de grandes quantidades de dados (big data). O processamento destes dados é um grande desafio à indústria atual pois obriga a grandes capacidades de armazenamento e de recursos computacionais. Assim sendo, as técnicas de big data analytics vêm tentar fazer face a estes desafios. É neste contexto que surge o projeto europeu IS4SE, na medida em que o seu principal objetivo é desenvolver uma infraestrutura de IIoT numa empresa de siderurgia com as ferramentas capazes de analisar dados. A equipa de I&D da Controlar encontra-se envolvida na realização deste projeto. Assim, este trabalho pretende executar algumas das principais tarefas atribuídas à Controlar no âmbito deste projeto, incluindo a criação de uma rede IoT e uma aplicação capaz de analisar os dados provenientes desta rede. Numa primeira fase foi desenvolvida a rede IoT, tendo por base dois protocolos principais: o LoRaWAN e o Wi-Fi via MQTT. Por um lado, utilizou-se o LoRaWAN devido ao seu baixo consumo e longo alcance. Por outro lado, recorreu-se ao Wi-Fi fazendo uso do protocolo MQTT devido à flexibilidade que este oferece. De seguida, criou-se uma gateway multiprotocolar capaz de gerir estes dois protocolos e desenvolveu-se um servidor para armazenamento, no mesmo local, de toda a informação proveniente dos diferentes protocolos da rede IoT. A fase final do trabalhou incidiu sobre o desenvolvimento de uma aplicação web capaz de tratar os dados provenientes dos diferentes nós conectados à rede IoT, de analisá-los e, consequentemente, criar estatísticas sobre os mesmos. Esta análise, engloba, igualmente, a aplicação de algoritmos de machine learning na informação adquirida. De forma a testar e validar a robustez da rede IoT e o bom funcionamento da aplicação web, foram integrados nós LoRaWAN e Wi-Fi na rede IoT criada ao longo de alguns meses, o que gerou um grande conjunto de dados a serem analisados pela aplicação web.
Industry 4.0, also known as IIoT, works through the massive implementation of IoT sensors and devices, which lead to the production of massive quantities of data (big data). Big data processing presents a great challenge to the current industry because it requires large storage and computational resources. big data analytics is related to the approaches developed to face these challenges. It is in this context that the European project IS4SE emerges, having as main goal to develop an IIoT infrastructure capable of analyzing data in a steel company. Controlar's I&D team is involved in carrying out this project. This thesis intends to carry out some of the main tasks assigned to Controlar in the scope of this project, in particular, the creation of an IoT network and a web application capable of analyzing the data coming from this network. In a first phase, an IoT network, based on two main protocols, was developed: LoRaWAN and Wi-Fi via MQTT. On the one hand, LoraWAN was used due to its low consumption and long reach. On the other hand, Wi-Fi was used making use of the MQTT protocol because of the flexibility it offers. Then, a multiprotocol gateway capable of managing these two protocols was created. It was also necessary to develop a server that would store, in the same location, all the information from the different protocols of the IoT network. In a later phase, a web app, capable of handling data from different nodes connected to the IoT network, to analyze them and, consequently, to create statistics on them, was developed. This analysis includes the use of machine learning algorithms in the acquired information. In order to test and validate the strength of the IoT network and the good functioning of the web app, LoRaWAN and Wi-Fi nodes were integrated into the IoT network enabling the collection of a large set of data, during a few month, to be analyzed by the web app.
Industry 4.0, also known as IIoT, works through the massive implementation of IoT sensors and devices, which lead to the production of massive quantities of data (big data). Big data processing presents a great challenge to the current industry because it requires large storage and computational resources. big data analytics is related to the approaches developed to face these challenges. It is in this context that the European project IS4SE emerges, having as main goal to develop an IIoT infrastructure capable of analyzing data in a steel company. Controlar's I&D team is involved in carrying out this project. This thesis intends to carry out some of the main tasks assigned to Controlar in the scope of this project, in particular, the creation of an IoT network and a web application capable of analyzing the data coming from this network. In a first phase, an IoT network, based on two main protocols, was developed: LoRaWAN and Wi-Fi via MQTT. On the one hand, LoraWAN was used due to its low consumption and long reach. On the other hand, Wi-Fi was used making use of the MQTT protocol because of the flexibility it offers. Then, a multiprotocol gateway capable of managing these two protocols was created. It was also necessary to develop a server that would store, in the same location, all the information from the different protocols of the IoT network. In a later phase, a web app, capable of handling data from different nodes connected to the IoT network, to analyze them and, consequently, to create statistics on them, was developed. This analysis includes the use of machine learning algorithms in the acquired information. In order to test and validate the strength of the IoT network and the good functioning of the web app, LoRaWAN and Wi-Fi nodes were integrated into the IoT network enabling the collection of a large set of data, during a few month, to be analyzed by the web app.
Description
Keywords
IIoT IoT LoRaWAN MQTT Aplicação web Análise de dados Machine learning Web Application Data analysis