Sprocket Ai Vision

AI quality system for the metalworking industry

Keywords: Predictive Maintenance, Ai, Artificial Vision, Industry 4.0

Authors: Pedro Apolinário, António H. J. Moreira

Currently, high standards of quality in the manufacture of components for the most diverse industries, from electrical to automotive, oblige suppliers to guarantee the tolerances defined in the components and to prove the absence of defects that could harm the final product. This project intends to implement an artificial vision solution complemented by artificial intelligence algorithms for the detection of anomalies or defects in sprockets used in the steering units of vehicles. Being a critical element, the presence of scobs or wear on the teeth must be detected and the component rejected. These defects often arise from the wear of the punch cutting components that give rise to the component. Thus, the system, in addition to rejecting, should also alert the person in charge of production for tool wear, in this way minimizing losses and associated costs.