| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 8.31 MB | Adobe PDF |
Authors
Abstract(s)
A indústria metalomecânica portuguesa apresenta um papel crucial no cenário industrial e no crescimento económico nacional. Contudo, enfrenta um desafio significativo: a elevada incidência de acidentes laborais. Inserida nesta indústria, encontra-se o grupo Mecwide, uma multinacional especializada na prestação de serviços em projetos com contratos a nível integrado. Esta organização destaca-se pela sua preocupação com a segurança dos seus colaboradores e pela ambição de otimizar os seus processos, visando aumentar a competitividade. Com o intuito de identificar soluções para otimizar estas áreas, foram conduzidos três estudos
distintos, todos focados no âmbito da Indústria 4.0: o RAG Chatbot, a Computer Vision e os Digital Twins.
O objetivo da solução proposta para o RAG Chatbot é proporcionar respostas rápidas e precisas a perguntas sobre os Equipamentos de Proteção Individual necessários para a realização de funções específicas em diversos projetos. Com este propósito, o estudo visou identificar a abordagem mais adequada para a implementação desta tecnologia na empresa, considerando duas alternativas: desenvolver um modelo próprio ou adotar uma solução integral na cloud. Além disso, foram analisados os custos associados a cada opção. O estudo concluiu que a solução mais viável seria a adoção de um modelo na cloud, com custos que variam entre 28,18€ e 1392,38€, consoante a escolha entre um modelo mais económico ou um mais dispendioso
O segundo estudo, focado em Computer Vision, teve como propósito determinar o sistema mais apropriado para implementação na Mecwide. A empresa pretende empregar este sistema para verificar se os colaboradores estão efetivamente a usar os Equipamentos de Proteção Individual, bem como para realizar o tracking de produtos. Foi escolhido um sistema de Computer Vision da NVIDIA, equipado com um processador Jetson, cujo investimento totaliza 8343,60€. Este valor abrange não apenas as câmaras necessárias e as lentes correspondentes, mas também o processador Jetson. O último estudo, focalizado nos Digital Twins, teve como objetivo identificar o modelo mais adequado para orientar a criação do modelo de Digital Twins da Mecwide. A empresa planeia integrar esta tecnologia numa das suas fábricas para otimizar os processos produtivos. A seleção do modelo baseou-se numa revisão sistemática conduzida por Eduardo Guzman, Beatriz Andres e Raul Poler, intitulada "Models and algorithms for production planning, scheduling and sequencing
problems: A holistic framework and a systematic review". O artigo de Hooker sobressaiu como a fonte ideal para guiar a conceção do modelo de Digital Twin da Mecwide. Os resultados dos estudos validaram a viabilidade das propostas, destacando a pertinência das soluções tecnológicas analisadas, as quais se recomendam para implementação na Mecwide.
The Portuguese metalworking industry plays a pivotal role in the industrial landscape and national economic growth. However, it faces a significant challenge: a high incidence of occupational accidents. Within this industry operates Mecwide Group, a multinational specialized in integrated contract project services. This organization distinguishes itself by its commitment to employee safety and its ambition to optimize processes to enhance competitiveness. To identify solutions for optimizing these areas, three distinct studies were conducted, all focused on Industry 4.0: RAG Chatbot, Computer Vision and Digital Twins. The proposed solution for the RAG Chatbot aims to provide quick and accurate responses to questions regarding the Personal Protective Equipment required for specific tasks across various projects. To achieve this goal, the study aimed to determine the most suitable approach for implementing this technology within the company, considering two alternatives: developing an in house model or adopting a comprehensive cloud-based solution. Additionally, the associated costs of each option were analyzed. The study concluded that the most viable solution would be adopting a cloud-based model, with costs ranging from 28,18€ to 1392,38€, depending on the choice between a more economical or more expensive model. The second study, focused on Computer Vision, aimed to determine the most suitable system for implementation at Mecwide. The company intends to use this system to verify employees compliance with wearing Personal Protective Equipment and to track products. A Computer Vision system from NVIDIA equipped with a Jetson processor was selected, with a total investment of 8343,60€. This amount covers not only the necessary cameras and corresponding lenses but also the Jetson processor. The final study, focused on Digital Twins, aimed to identify the most suitable model to guide the creation of Mecwide's Digital Twin model. The company plans to integrate this technology into one of its factories to optimize production processes. The model selection was based on a systematic review conducted by Eduardo Guzman, Beatriz Andres, and Raul Poler, titled "Models and algorithms for production planning, scheduling and sequencing problems: A holistic framework and a systematic review". Hooker's article emerged as the ideal source to guide the conception of Mecwide's Digital Twin model. The study results validated the feasibility of the proposals, emphasizing the relevance of the analyzed technological solutions, which are recommended for implementation at Mecwide.
The Portuguese metalworking industry plays a pivotal role in the industrial landscape and national economic growth. However, it faces a significant challenge: a high incidence of occupational accidents. Within this industry operates Mecwide Group, a multinational specialized in integrated contract project services. This organization distinguishes itself by its commitment to employee safety and its ambition to optimize processes to enhance competitiveness. To identify solutions for optimizing these areas, three distinct studies were conducted, all focused on Industry 4.0: RAG Chatbot, Computer Vision and Digital Twins. The proposed solution for the RAG Chatbot aims to provide quick and accurate responses to questions regarding the Personal Protective Equipment required for specific tasks across various projects. To achieve this goal, the study aimed to determine the most suitable approach for implementing this technology within the company, considering two alternatives: developing an in house model or adopting a comprehensive cloud-based solution. Additionally, the associated costs of each option were analyzed. The study concluded that the most viable solution would be adopting a cloud-based model, with costs ranging from 28,18€ to 1392,38€, depending on the choice between a more economical or more expensive model. The second study, focused on Computer Vision, aimed to determine the most suitable system for implementation at Mecwide. The company intends to use this system to verify employees compliance with wearing Personal Protective Equipment and to track products. A Computer Vision system from NVIDIA equipped with a Jetson processor was selected, with a total investment of 8343,60€. This amount covers not only the necessary cameras and corresponding lenses but also the Jetson processor. The final study, focused on Digital Twins, aimed to identify the most suitable model to guide the creation of Mecwide's Digital Twin model. The company plans to integrate this technology into one of its factories to optimize production processes. The model selection was based on a systematic review conducted by Eduardo Guzman, Beatriz Andres, and Raul Poler, titled "Models and algorithms for production planning, scheduling and sequencing problems: A holistic framework and a systematic review". Hooker's article emerged as the ideal source to guide the conception of Mecwide's Digital Twin model. The study results validated the feasibility of the proposals, emphasizing the relevance of the analyzed technological solutions, which are recommended for implementation at Mecwide.
Description
Keywords
Indústria 4.0 Digital Twins Computer vision RAG chatbot Indústria metalomecânica Segurança Industry 4.0 Metalworking industry Safety