Diagnosis of high-speed dental pieces from their sound analysis

Abstract

The computational tools are developed to help professionals to determine anomalies in different equipment.  These tools seek to determine any damage without disassembly for the purpose of optimize processes, in this case the operation of the diagnose high speed dental piece.  This article presents the results of the implementation of a computational algorithm for obtaining, from the sounds generated by turbines high speed parts, in what state is this.  This is accomplished by capturing the sound of high-speed components in good and bad state, in order to build a database from these sounds, each of these signals are extracted features in different domains to train a neural network, which diagnose the state of the workpiece.  With the implementation of this system has been possible to achieve an 81% success rate for the classification of defective pieces.