Optimal Sabotage Attack on Composite Material Parts (1810.03010v1)

Abstract: Industry 4.0 envisions a fully automated manufacturing environment, in which computerized manufacturing equipment--Cyber-Physical Systems (CPS)--performs all tasks. These machines are open to a variety of cyber and cyber-physical attacks, including sabotage. In the manufacturing context, sabotage attacks aim to damage equipment or degrade a manufactured part's mechanical properties. In this paper, we focus on the latter, specifically for composite materials. Composite material parts are predominantly used in safety-critical systems, e.g., as load-bearing parts of aircraft. Further, we distinguish between the methods to compromise various manufacturing equipment, and the malicious manipulations that will sabotage a part. As the research literature has numerous examples of the former, in this paper we assume that the equipment is already compromised, our discussion is solely on manipulations. We develop a simulation approach to designing sabotage attacks against composite material parts. The attack can be optimized by two criteria, minimizing the "footprint" of manipulations. We simulate two optimal attacks against the design of a spar, a load bearing component of an airplane wing. Our simulation identifies the minimal manipulations needed to degrade its strength to three desired levels, as well as the resulting failure characteristics. Last but not least, we outline an approach to identifying sabotaged parts.