Artificial Intelligence prototype for damage prediction in electric induction motors

Abstract

Electric induction motors are the most used today to carry out industrial processes, so keeping them in optimal conditions guarantees their efficiency. Maintaining a predictive maintenance plan guarantees us to prevent major damages and avoid stopping production lines or normal operations in common faults such as in bearings, stator and rotor. The purpose of this project is to create a prototype using a convolutional neural network (CNN) with supervised learning that processes its layers imitating the visual cortex of the human eye to identify different punctual characteristics of an image, this methodology is based on the "pre-processing" of data. With this, the idea arises of creating an APP from the cell phone to detect engine failures through the audio spectrogram collected with the microphone of a mobile device so that a prediction can be made in a practical way at any moment.