Risk assessment of human-automation conflict under cyberattacks in process systems

Human-automation conflict is a frontier subject to be vigilant against, and it may become even more elusive and confusing under cyberattacks. Conflicts due to cyberattacks are often beyond the expertise of the human oper-ator, making the resolution much more difficult. Therefore, this study transforms common attacks into under-standable representations with process variables, explores human-automation conflict under five generalized attacks, e.g., false data injection, denial of service, etc., and explains the conflict with a proposed cooperative Pareto paradigm based on game theory, then applies on a Continuous Stirred Tank Reactor. The results show that cyberattacks could cause conflicts significantly. It also highlights that attack strength determines the strength of action conflict. The control actions could buffer some impact of attacks within a limited range. The conflict risk could be applied to distinguish a fault and an attack, and measures could be taken accordingly.