Automatic quality inspection of car bodies

• Grant number: 17/25873-8
• Support Opportunities: Research Grants - Innovative Research in Small Business - PIPE
• Duration: September 01, 2018 - October 31, 2020
• Field of knowledge: Engineering - Mechanical Engineering - Manufacturing Processes
• Convênio/Acordo: FINEP - PIPE/PAPPE Grant
• Principal Investigator: Jorge Augusto de Bonfim Gripp
• Grantee: Jorge Augusto de Bonfim Gripp
• Host Company: Autaza Tecnologia S/A
• CNAE: Fabricação de máquinas e equipamentos para uso industrial específico não especificados anteriormente ; Fabricação de automóveis, camionetas e utilitários
• City: São José dos Campos
• Associated researchers: Enivaldo Amaral de Souza
• Associated grant(s): 18/50097-4 - Automatic quality inspection of car bodies, AP.PIPE
• Associated scholarship(s): 18/15704-7 - Automatic quality inspection of car bodies, BP.TT

Abstract

Among the quality dimensions perceived by a customer, appearance is a key factor that influences the purchase of a car. Since appearance is a differentiator of the brand to win customers, the automotive industry performs strict quality control to ensure standardization in manufacturing and the absence of unwanted marks, beats or undulations. Currently, the method for surface quality evaluation is done visually by a trained inspector to identify and classify the defects. Besides being an intense and repetitive work, this method is subject to the subjectivity of the inspector's evaluation. As the inspection is not perfect, the industry loses millions of dollars every year due to wrong classifications, unnecessary rework or defects that are identified only in later stages of the process, where the correction is much more expensive. This project consists of an automatic quality inspection solution that will improve the objectivity and reproducibility of the inspection process. The product of this project is an Inspection Cell, containing equipment and software to assist the automotive industry in ensuring the quality of its products during manufacture. Cameras are used to capture images of the car body, which is illuminated with a special wall of light that projects a zebra pattern on the car. The software analyzes the reflected fringes of light on the surface of the car and creates an automated report that shows the defects identified and their classification using color, facilitating communication for the correction teams. The company AUTAZA is a startup resulting from an three-year research and development project between the General Motors (GM) automotive industry in São Caetano do Sul and the Center of Competence in Manufacturing (CCM-ITA), a laboratory of the Aeronautics Institute of Technology (ITA) in São José dos Campos-SP. Recognizing the innovative potential of the prototypes developed in this project, its three executing members formed AUTAZA on March 17, 2016, with GM as their first customer. In 2017 Autaza had a PIPE Phase 1 project approved by FAPESP for the improvement of the algorithm for classifying defects in car bodies. Its main technological differential is not only the identification of surface defects but also the classification using artificial intelligence in a compatible and adaptable way with the standards currently adopted by the industry. (AU)