Skip to main content

Control-Theory-Informed Feature Selection for Detecting Malicious Tampering in Additive Layer Manufacturing Processes...

Publication Type
Conference Paper
Book Title
Proceedings of the 16th International Conference on Cyber Warfare and Security
Publication Date
Page Numbers
55 to 64
Publisher Location
Reading, United Kingdom
Conference Name
16th International Conference on Cyber Warfare and Security (ICCWS 2021)
Conference Location
Cookeville, Tennessee, United States of America
Conference Sponsor
Tennessee Tech University CEROC, Tennessee Tech Computer Science Department, Oak Ridge National Laboratory
Conference Date

Additive layer manufacturing (ALM) is rapidly becoming an appealing solution to the low-volume manufacturing of metal, polymer, or composite parts. However, ALM’s reliance on digital part specifications, microcontrollers, and modern networking makes these devices vulnerable to malicious tampering by cyber attackers, which can negatively affect part performance and even result in catastrophic failure. We present a hybrid analytic approach to feature discovery using control theoretic techniques and linear modelling on input-output data collected from a representative controller system. Employing this approach, we design, train, and test an anomaly detection system. The preliminary results show that the proposed approach effectively discovers useful input-output relationships for anomaly detection in a simulated ALM process. Application to larger and more complex systems are discussed.