Automated industrial inspection workbench for human machine interface (HMI) consoles

• Autor(a) principal: Nakashima, Alison Shigueo
• Data de Publicação: 2018
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10198/19451
• topic: Adaptive industrial inspection; TFT and LCD display inspection; Adaptive workbench; Force sensor; Machine learning; Domínio/Área Científica::Engenharia e Tecnologia::Outras Engenharias e Tecnologias

The actual moment of the industrial production is changing the way of production. Now the systems are adaptable to produce different items in the same production line with a very reduced time to setup the systems. In the same way, the quality control systems must be more adaptable and intelligent possible. The present work propose the creation of intelligent and adaptable inspection cell to inspect Human Machine Interface (HMI) consoles of different types. This cell is composed by an image acquisition system with controlled illumination, a force sensor installed on the robot tool to verify the buttons’ functionality. The force tests are processed and classified using decision three, k-Nearest Neighbors (k-NN) and Support Vector Machine (SVM) classification method. Besides, the Thin-Film Transistor (TFT) display uses Normalized Cross-Correlation (NCC) and Correlation Coefficients (CC) to check the display’s regions. To Liquid Cristal Display (LCD) is used the same method and also be used a Neural Network Classification (NNC). In the experimental tests, four different types of consoles prototypes are tested, one of them has a TFT display and buttons, others two have only buttons and one has only a LCD display. In the inspection workbench is created, all the hardware necessary to execute the inspection was installed successfully. Moreover, the inspection methods obtained a precision higher than 90% to the buttons and display inspection.