Real time 3D reconstruction for enhanced cybersecurity of additive manufacturing processes

Abstract

Industry 4.0 has enhanced automation and connectivity in manufacturing but also increased the risk of cyber-intrusions, particularly in additive manufacturing (AM) used in critical sectors such as defense, aerospace, and healthcare. This study presents a real-time process monitoring framework that detects cyber intrusions and halts the printing in a 3D printer. We retrofit existing printers with low-cost depth sensors that continuously capture 3D spatial data. Our framework reconstructs the printed object in real-time and compares it against a virtual ground truth model to detect geometric discrepancies indicative of cyber-intrusions. The entire detection process operates within the time taken to print a single layer for our case study of a standard 3D Benchy. We validate our approach by simulating cyber-intrusions that alter the G-code sent to the printer, successfully detecting and stopping compromised prints. This framework enhances AM cybersecurity by providing real-time threat detection and intervention, ensuring secure and resilient automated manufacturing in Industry 4.0.