A predictive analysis data-based for Additive Manufacturing

Additive Manufacturing (AM) is considered as an essential technology for the new industrial revolution, the use of such technology helps reducing considerably the prototyping cost and complexity. Adding to its recent appearing, the powerful advances of this technology allowed introducing several new tracks, like the completely AM However, the AM suffers from several issues (printing quality, robustness, efficiency, etc.), thus delaying its large-scale adoption in industry. One of the most considered issues in AM is the production reliability, in this paper we present a data-based predictive analysis approach for detecting printing failure in AM (3D printer). The proposed approach is based on sound analysis to predict the printing conditions. The goal is to predict the last state of object in press (conforming objects or not) based on previously extracted features from collected sounds.