Risk mitigation measures during adoption of ML techniques for additive manufacturing quality control and data security

Abstract

Additive manufacturing (AM) has arisen as a promising advanced manufacturing innovation. Notwithstanding, its expansive selection in industry is as yet impeded by high passage boundaries of design for additive manufacturing (DfAM), restricted materials library, different preparing deserts, and conflicting product quality. Lately, machine learning (ML) has acquired expanding consideration in AM because of its unprecedented performance in information undertakings like order, relapse and grouping. This article gives a comprehensive audit on the cutting edge of ML applications in an assortment of AM spaces. In the DfAM, ML can be utilized to yield new elite Meta materials and advanced topological designs. In AM preparing, contemporary ML calculations can assist with upgrading measure parameters, and lead examination of powder spreading and in-measure deformity observing. On the production of AM, ML can help professionals in pre-manufacturing planning, and product quality assessment and control. In addition, there has been an expanding worry about information security in AM as information penetrates could happen with the guide of ML procedures. This paper puts forth the challenges arising when machine learning techniques are used during quality control and data security in the field of additive manufacturing. Then we propose few risk mitigation strategies to counter those challenges. This paper can be a readymade guide for practitioners who are involved in AM process considering ML solutions in the process.