The Use of Artificial Intelligence for Safeguarding Cyber-Physical Business Systems

Several security issues, such as the forgery of information-sharing channels, hardware components, and operating systems, hampered the steady expansion of cyber-physical systems (CPS), especially those with safety-critical implementations. As a result of these security concerns, different CPS stacks with accurate static and flexible control and prevention mechanisms were developed, including cross-layer and intra-layer coupling. This study offers a framework for applying machine intelligence to cyber-physical business systems (MI-CPBS). An overview of the many CPS-level security difficulties, including threat modeling and the underlying scientific challenges that must be surmounted to develop suitable security solutions, is presented first in this article. The limitations of static and dynamic detection and avoidance methods, such as their inability to spot runtime safety attacks owing to dormancy or ambiguity, are also summarized in this study. This article uses machine learning-based techniques to investigate new, cutting-edge security solutions for the many vulnerabilities presented in the context of different CPS stacks. It also includes open-ended research topics and challenges to creating innovative security measures for CPS. In conclusion, it summarizes the security research for intelligent CPS and provides significant reviews.