Information Embedding in Additive Manufacturing through Printing Speed Control

Abstract

Additive manufacturing (AM) is rapidly developing, and new applications are continuously emerging. While AM is increasingly becoming integral to many industries, including aerospace, automotive, and biomedical, it has opened a host of unique security concerns, from theft of technical data to process sabotage and counterfeiting. In this work, we present a method to address the counterfeiting problem by embedding information in additively manufactured parts through controlling printing process parameters. Variations in printing speed, the encoding parameter in this work, introduce subtle localized height differences on parts' surfaces, which are readable using an optical profilometer. The profilometry data is captured after printing, and this data is processed to predict the intended bit response for each embedding region on the surface of the part. We experimentally demonstrate the feasibility of the proposed scheme for embedding and reading the information in 3D printed parts and show that it achieves 80% accuracy for a 53 mm/s difference in printing speed between the encoded bits. Finally, we characterize the performance of the proposed scheme, measured as the accuracy in decoded messages, as a function of the difference in printing speed used to perform the embedding.