(375d) Practical Issues in Cybersecurity: From Encryption to Images

Cybersecurity has gained interest for chemical processes in recent years. For example, [1] discusses techniques for considering cybersecurity during traditional chemical engineering procedures such as hazard and risk analysis. One of the main drivers for the cybersecurity discussion is an increased interest in recent years in concepts such as Industry 4.0 [2] and the Industrial Internet of Things [3], and a general increase in data analytics and artificial intelligence approaches. Future work in this domain would benefit from research that enables a more integrated approach to control system cybersecurity in the chemical engineering research field, in which the cyber defenses at a plant are more explicitly considered, as well as advanced control frameworks and concepts. These concepts include resilient frameworks, to analyze how a system functions as a whole and the types of attacks it remains vulnerable to after both, the theoretical requirements for certain resilient control frameworks as well as, practical communication and networking protocols that need to be accounted for. In addition, there are a wide variety of issues becoming more important to consider from a cybersecurity perspective over time. For example, image analysis is important for certain types of decision-making systems, though image data represents another type of information that could be falsified to manipulate such decision-making [4], and computations on the Cloud, where it may be desired to encrypt not only the data but the computations themselves [5, 6], are additional concerns.

Motivated by the above, this talk highlights multiple concepts in the practical aspects of control system cybersecurity as it interfaces with advanced control. For example, using a case study of a benzene hydealkylation process under distributed Lyapunov-based economic model predictive control [7], we explain the role of encryption protocols in enabling safe transfer of information between distributed units, and the extent to which the encryption might enable attacked controllers to be identified on the network. We utilize a process example of a semi-batch polymerization reactor under model predictive control with uncertainty scenarios [8] and offset-free model predictive control [9] to demonstrate the manner in which disturbance-handling is not equivalent to cyberattack resilience, and we also discuss profit targeting as a key metric in distinguishing attacks and faults. We follow up with a discussion of considerations in image analysis, which form part of the increasingly automated world that is more open to attacks, and verification in that domain, and how such concepts relate to cybersecurity and cyberattack resilience. For example, focusing on the zinc flotation process, in which image-based control is considered today [10], we explore the potential use of software such as Blender for creating a preliminary closed-loop simulation methodology for simulating the zinc flotation process, including both system state evolution under control actions, as well as image generation and processing, along with analysis on the impacts of cyberattacks on image data that determine process characteristics. Finally, we will discuss the role of communication techniques, as well as encryption techniques which facilitate communication with the Cloud, in control for nonlinear chemical process systems.

[1] Cormier, A., & Ng, C. (2020). Integrating cybersecurity in hazard and risk analyses. Journal of Loss Prevention in the Process Industries, 64, 104044.

[2] Reis, M. S., & Kenett, R. (2018). Assessing the value of information of data‐centric activities in the chemical processing industry 4.0. AIChE Journal, 64(11), 3868-3881.

[3] Sisinni, E., Saifullah, A., Han, S., Jennehag, U., & Gidlund, M. (2018). Industrial internet of things: Challenges, opportunities, and directions. IEEE Transactions on Industrial Informatics, 14(11), 4724-4734.

[4] Bucci, E. M. (2018). Automatic detection of image manipulations in the biomedical literature. Cell death & disease, 9(3), 1-9.

[5] Darup, M. S., Redder, A., & Quevedo, D. E. (2018). Encrypted cloud-based MPC for linear systems with input constraints. IFAC-PapersOnLine, 51(20), 535-542.

[6] Alexandru, A. B., Morari, M., & Pappas, G. J. (2018, December). Cloud-based MPC with encrypted data. In 2018 IEEE Conference on Decision and Control (CDC) (pp. 5014-5019). IEEE.

[7] Liu, J., Chen, X., Muñoz de la Peña, D., & Christofides, P. D. (2010). Sequential and iterative architectures for distributed model predictive control of nonlinear process systems. AIChE Journal, 56(8), 2137-2149.

[8] Lucia, S., Finkler, T., & Engell, S. (2013). Multi-stage nonlinear model predictive control applied to a semi-batch polymerization reactor under uncertainty. Journal of process control, 23(9), 1306-1319.

[9] Das, B., & Mhaskar, P. (2015). Lyapunov‐based offset‐free model predictive control of nonlinear process systems. The Canadian Journal of Chemical Engineering, 93(3), 471-478.

[10] Kaartinen, J., Hätönen, J., Hyötyniemi, H., & Miettunen, J. (2006). Machine-vision-based control of zinc flotation—a case study. Control Engineering Practice, 14(12), 1455-1466.