Artificial intelligence for advanced manufacturing quality

Abstract

This Thesis addresses the challenge of AI-based image quality control systems applied to manufacturing industry, aiming to improve this field through the use of advanced techniques for data acquisition and processing, in order to obtain robust, reliable and optimal systems. This Thesis presents contributions onthe use of complex data acquisition techniques, the application and design of specialised neural networks for the defect detection, and the integration and validation of these systems in production processes. It has been developed in the context of several applied research projects that provided a practical feedback of the usefulness of the proposed computational advances as well as real life data for experimental validation.