Machine Learning Applications for Additive Manufacturing

Abstract

Additive manufacturing (AM) is the advanced fabrication technology through which an object is created one layer at a time. AM becomes vital to fulfil various Industry 4.0 requirements. It can drastically minimize material waste, production time and reduce the number of components required for assembly. Meanwhile, Machine Learning (ML) is gradually making its way into the manufacturing industry to ease industrial automation. ML increases production efficiency and also brings machine interaction closer to human interaction. Because of its exemplary service in data-related operations such as clustering, categorization, and regression, ML has received growing attention in recent years in AM. This chapter provides an in-depth examination of ML applications in AM, including design for AM, AM production, inspection, testing, validation, and repair and restoration of AM components. ML can be used to produce novel high-performance metamaterials and optimal topological designs for AM. Contemporary ML methods can aid in the optimization of input variables, as well as the assessment of in-process defect detection in ML. ML can assist practitioners in pre-manufacturing planning and product quality assessment and control when it comes to the production of AM. The study found that, despite growing interest in the use of AM for repair and restoration, research on the application of ML for repair and restoration design optimization remains scarce. Additionally, no standard guidelines exist for designing for repair and restoration via AM. Additionally, there has been growing worry regarding AM data security, as data breaches could arise due to ML algorithms. The chapter concludes with a section that summarizes the key results from the literature and gives future scope and recommendations for the advancement of AM using ML.