Bayesian additive modeling for quality control of 3D printed products

Three-dimensional (3D) printing is a disruptive technology with the potential to revolutionize manufacturing. However, control of product boundary deformation is a major issue that can limit its impact in practice. The fundamental requirement for quality control is a generic methodology that can predict deformations for a wide range of designs based on the available data of a few previously manufactured products, potentially of different designs. We develop a Bayesian methodology to effectively update prior conceptions of deformation for a new design based on printed products of different shapes. Our approach is applied to infer deformation models for regular polygons based on deformation models and data for circles. Ultimately, our methodology fills a gap in comprehensive quality control for 3D printing, and can advance it as a high-impact manufacturing technology.