Artificial intelligence and convolutional neural networks optimized for industrial processes

The aim of this scientific paper is an experimental of artificial intelligence techniques using deep learning and in particular convolutional neural networks (CNN) to optimize industrial processes. An application is presented that is able to recognize components within an electrical equipment and verify their state. At the same time, the application attempts to identify the coding of industrial components in order to be able to construct an enrichment of the component information. Using an optical character recognition system for detecting and reading the component coding, a search is conducted for the technical specifications of the components. On this aspect, an innovative category prediction system is presented that can recommend the best solution for possible modifications or changes in the event of component malfunctions or failures.