Thiago Domingos Lemos
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Thiago Domingos Lemos has a total of 2 co-authors. Here are the top co-authors by number of shared publications:
- Luiz Augusto Zelaquett De Souza
Escola Politécnica de Pernambuco, Universidade de Pernambuco
1 shared publication - Roberta Andrade de Araújo Fagundes
Universidade de Pernambuco
1 shared publication
Latest Publications
- 1 December 2025
- Published by Associacao Brasileira de Inteligencia Computacional - ABRICOM
- p. 1-8
Quality Prediction of Resistance Spot Welding with Preheating in Imbalanced Dataset
Resistance Spot Welding (RSW) holds significant economic importance across various industries, particularly in the automotive sector, where it plays a crucial role in ensuring the safety and structural integrity of manufactured automobiles. However, when working with galvanized metal sheets, RSW faces unique challenges due to the presence of zinc oxide layer on the surface, which can reduce weld quality. Therefore, a commonly adopted solution is the preheating technique, which helps disrupt this oxide layer and improve weld formation. Despite its widespread use, preheating is often overlooked in predictive quality prediction models. In response, this research addresses that gap by investigating the ability of machine learning techniques to classify welding process data between ‘good’ or ‘defective’ welds with information from preheating phase. To further improve classification performance—especially in the presence of imbalanced datasets—this study proposes an ensemble-based approach combined with characteristics extraction from physical features of the dynamic resistance curve. An ensemble approach was explored by integrating various base classifiers within a Balanced Bagging Classifier, combined through a Voting Classifier to leverage their complementary strengths. The performance of the ensemble model was rigorously evaluated using cross-validation technique to ensure robustness and generalizability. The results revealed that the use of balanced bagging classifiers improved the results of classifying defective and good spot welding with respect to default bagging classifiers in the context of imbalaced data.
- 1 December 2025
- Published by Associacao Brasileira de Inteligencia Computacional - ABRICOM
- p. 1-8
Quality Prediction of Resistance Spot Welding with Preheating in Imbalanced Dataset
Resistance Spot Welding (RSW) holds significant economic importance across various industries, particularly in the automotive sector, where it plays a crucial role in ensuring the safety and structural integrity of manufactured automobiles. However, when working with galvanized metal sheets, RSW faces unique challenges due to the presence of zinc oxide layer on the surface, which can reduce weld quality. Therefore, a commonly adopted solution is the preheating technique, which helps disrupt this oxide layer and improve weld formation. Despite its widespread use, preheating is often overlooked in predictive quality prediction models. In response, this research addresses that gap by investigating the ability of machine learning techniques to classify welding process data between ‘good’ or ‘defective’ welds with information from preheating phase. To further improve classification performance—especially in the presence of imbalanced datasets—this study proposes an ensemble-based approach combined with characteristics extraction from physical features of the dynamic resistance curve. An ensemble approach was explored by integrating various base classifiers within a Balanced Bagging Classifier, combined through a Voting Classifier to leverage their complementary strengths. The performance of the ensemble model was rigorously evaluated using cross-validation technique to ensure robustness and generalizability. The results revealed that the use of balanced bagging classifiers improved the results of classifying defective and good spot welding with respect to default bagging classifiers in the context of imbalaced data.
- Open Access
- 30 November 2022
- Published by Revista de Engenharia e Pesquisa Aplicada in Revista de Engenharia e Pesquisa Aplicada
- Vol. 7 (3) , 54-63
Abordagem Preditiva de Quebras Baseada em Logs de Eventos na Indústria Automotiva
Nas últimas décadas, a indústria automobilística vem experimentando diversas e significativas mudanças. A Indústria 4.0 surge neste contexto aspirando a um alto nível de conectividade ao longo de todo o ciclo de vida do produto, exigindo cada vez mais tecnologias de controles de dados, permitindo o desenvolvimento da manutenção preditiva. O presente trabalho focaliza na abordagem da manutenção preditiva estatística, através da análise de logs de falhas de um grupo específico e semelhante de máquinas, tendo como objetivo o desenvolvimento de um modelo preditivo de quebras. A abordagem de aprendizado de máquina foi escolhida devido à sua capacidade de previsão de variáveis-saída a partir de dados históricos como entrada. Nossos dados foram coletados ao longo de vários anos de 400 robôs de um grande fabricante desses tipos de braços robotizados. Por fim foi obtida uma configuração final do modelo através da aplicação de uma rede neural artificial.
- 30 November 2022
- Published by Revista de Engenharia e Pesquisa Aplicada in Revista de Engenharia e Pesquisa Aplicada
- Vol. 7 (3) , 54-63
Open AccessAbordagem Preditiva de Quebras Baseada em Logs de Eventos na Indústria Automotiva
Nas últimas décadas, a indústria automobilística vem experimentando diversas e significativas mudanças. A Indústria 4.0 surge neste contexto aspirando a um alto nível de conectividade ao longo de todo o ciclo de vida do produto, exigindo cada vez mais tecnologias de controles de dados, permitindo o desenvolvimento da manutenção preditiva. O presente trabalho focaliza na abordagem da manutenção preditiva estatística, através da análise de logs de falhas de um grupo específico e semelhante de máquinas, tendo como objetivo o desenvolvimento de um modelo preditivo de quebras. A abordagem de aprendizado de máquina foi escolhida devido à sua capacidade de previsão de variáveis-saída a partir de dados históricos como entrada. Nossos dados foram coletados ao longo de vários anos de 400 robôs de um grande fabricante desses tipos de braços robotizados. Por fim foi obtida uma configuração final do modelo através da aplicação de uma rede neural artificial.