Sayed-Mouchaweh - Diagnosability, Security and Safety of Hybrid Dynamic and Cyber-Physical Systems

This Springer imprint is published by the registered company Springer International Publishing AG part of Springer Nature.

The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Cyber-physical systems (CPS) are characterized as a combination of physical (physical plant, process, network) and cyber (software, algorithm, computation) components whose operations are monitored, controlled, coordinated, and integrated by a computing and communicating core. The interaction between physical and computational components in CPS is intensive. They cover an increasing number of real life applications such as autonomous vehicles, aircrafts, smart manufacturing processes, surgical robots and human robot collaboration, smart electric grids, home appliances, air traffic control, automated farming, and implanted medical devices. The interaction between both physical and cyber components requires tools allowing analyzing and modeling both the discrete (discrete event control, communication protocols, discrete sensors/actuators, scheduling algorithms, etc.) and continuous (continuous dynamics, physics, continuous sensors/actuators, etc.) dynamics. Therefore, many CPS can be modeled as hybrid dynamic systems in order to take into account both discrete and continuous behaviors as well as the interactions between them.

Many critical infrastructures, such as power generation and distribution networks, water networks and mass transportation systems, autonomous vehicles and traffic monitoring, are CPS. Such systems, including critical infrastructures, are becoming widely used and covering many aspects of our daily life. Therefore, the security, safety, and reliability of CPS is essential for the success of their implementation and operation. However, these systems are prone to major incidents resulting from cyberattacks and system failures. These incidents affect significantly their security and safety. Attacks can be represented as an interference in the communication channel between the supervisor and the system intentionally generated by intruders in order to damage the system or as the enablement, respectively disablement, of actuators events that are disabled, respectively enabled, by the supervisor. In general, intruders hide, create, or even change intentionally events that transit from one device (actuator, sensor) to another in a control communication channel. These attacks in a supervisory control system can lead the plant to execute event sequences entailing the system to reach unsafe or dangerous states that can damage the system.

Therefore, reliable, scalable, and timely fault diagnosis is crucial in order to improve the robustness of CPS to failures. In addition, it is primordial to detect intrusions that exploit the vulnerabilities of industrial control systems in order to alter intentionally the integrity, confidentiality, and availability of CPS. These cyberattacks affect the control commands of the controller [Programmable Logic Controller (PLC)], the reports (sensors readings) coming from the plant as well as the communication between them. Moreover, a thorough understanding of the vulnerability of CPS components against such incidents can be incorporated in future design processes in order to better design such systems. Finally, the timely fault diagnosis can help operators to have better situation awareness and give them ample time to implement correction (maintenance) actions.

However, guaranteeing the security and safety of CPS requires verifying their behavioral or safety properties either at design stage such as state reachability, diagnosability, and predictability or online such as fault detection and isolation. This is a challenging task because of the inherent interconnected and heterogeneous combination of behaviors (cyber/physical, discrete/continuous) in these systems. Indeed, fault propagation in CPS is governed not only by the behaviors of components in the physical and cyber subsystems but also by their interactions. This makes the identification of the root cause of observed anomalies and predicting the failure events a hard problem. Moreover, the increasing complexity of CPS and the security and safety requirements of their operation as well as their decentralized resource management entail a significant increase in the likelihood of failures in these systems. Finally, it is worth mentioning that computing the reachable set of states of HDS is an undecidable matter due to the infinite state space of continuous systems.

This edited Springer book presents recent and advanced approaches and techniques that address the complex problem of analyzing the diagnosability property of CPS and ensuring their security and safety against faults and attacks. The CPS are modeled as hybrid dynamic systems using different model-based and data-driven approaches in different application domains (electric transmission networks, wireless communication networks, intrusions in industrial control systems, intrusions in production systems, wind farms, etc.). These approaches handle the problem of ensuring the security of CPS in presence of attacks and verifying their diagnosability in presence of different kinds of uncertainty (uncertainty related to the event occurrences, to their order of occurrence, to their value etc.).

Finally, the editor is very grateful to all authors and reviewers for their very valuable contribution allowing setting another cornerstone in the research and publication history of studding the diagnosability, security, and safety of CPS modeled as hybrid dynamic systems. I would like also to acknowledge Mrs. MaryE. James for establishing the contract with Springer and supporting the editor in any organizational aspects. I hope that this volume will be a useful basis for further fruitful investigations and fresh ideas for researcher and engineers as well as a motivation and inspiration for newcomers to address the problems related to this very important and promising field of research.