Industrial Cyber-physical Systems Research Articles

Contradictions Identification of Safety and Security Requirements for Industrial Cyber–Physical Systems

Contradictions Identification of Safety and Security Requirements for Industrial Cyber–Physical Systems

Endpoint cybersecurity: When smart devices turn stupid

With the proliferation of ever-cheaper IoTs (or smart devices) from smart doorbells to WiFi security cameras and baby monitors and hybrid work arrangements involving relaxed BYOD (Bring Your Own Device) policies, cybersecurity education must build awareness on the threats that endpoints pose. While securing corporate networks from cyberattack gains much attention, there is a dearth of educational material that exemplifies the processes by which endpoints are used in hacks, and how to build defenses against endpoint vulnerabilities. This teaching case provides a Defense in Depth (DiD) learning framework on how to defend industrial cyber-physical systems, digital infrastructure and assets. The case illustrates the applicability of a DiD framework independent of attack vector yet particular to endpoint vulnerabilities and shows how lessons from the past can help defend and mitigate future cyberattacks.

A Systematic Analysis of Security Metrics for Industrial Cyber–Physical Systems

Nowadays, as the cyber-threat landscape is evolving and digital assets are proliferating and becoming more and more interconnected with the internet and heterogeneous devices, it is fundamental to be able to obtain a sensible measure of the security of devices, networks, and systems. Industrial cyber–physical systems (ICPSs), in particular, can be exposed to high operational risks that entail damage to revenues, assets, and even people. A way to overcome the open question of measuring security is with the use of security metrics. With metrics it is possible to rely on proven indicators that benchmark systems, identify vulnerabilities, and show practical data to assess the risk. However, security metrics are often proposed with specific contexts in mind, and a set of them specifically crafted for ICPSs is not explicitly available in the literature. For this reason, in this work, we analyze the current state of the art in the selection of security metrics and we propose a systematic methodology to gather, filter, and validate security metrics. Then, we apply the procedure to the ICPS domain, gathering almost 300 metrics from the literature, analyzing the domain to identify the properties useful to filter the metrics, and applying a validation framework to assess the validity of the filtered metrics, obtaining a final set capable of measuring the security of ICPSs from different perspectives.

CyberDefender: an integrated intelligent defense framework for digital-twin-based industrial cyber-physical systems

CyberDefender: an integrated intelligent defense framework for digital-twin-based industrial cyber-physical systems

A Digital Twin Development Framework for an Electrical Submersible Pump (ESP)

Premature failure of a sub-system can be critical for an industrial Cyber-Physical System (CPS). A digital twin (DT) assisted predictive maintenance procedure can reduce the risk of costly unplanned maintenance. This study presents a generalized DT development framework for an Electrical Submersible Pump (ESP) that can assist in predictive maintenance. The framework is applied on a single-phase ESP as a proof of concept. The maximum winding temperature of the selected ESP is simulated using a multiphysics simulation tool with transient electromagnetic and transient heat transfer solvers. The simulation parameters were refined using data captured through an ESP free-run experiment. Simulating the total energy loss in the ESP stator and rotor and the transfer of heat from the outer fluid domain facilitates a relationship between the measurable external temperature and the maximum temperature in the stator winding. Following a design of experiments (DOE) approach, a series of simulations were run to establish a statistical model for the winding temperature in terms of the fluid temperature, the time duration a particular temperature was persistent, and the initial maximum stator winding temperature. As the instantaneous maximum stator winding temperature is related to the remaining useful lifetime, it was shown using a case study that the proposed framework can prognosticate the ESP failure, assisting effective decision-making for predictive maintenance of a CPS.

Lightweight privacy-preserving authentication mechanism in 5G-enabled industrial cyber physical systems

With the deep integration of informatization and industrialization, cyber-physical system (CPS), which integrates computing, communication and control technologies, comes into being and has been widely used in industrial applications. A large number of information physical system devices and control systems are based on open internet connections, and security and privacy protection issues are gradually emerging, especially in the process of dynamic and difficult to control information physical system stream data transmission and storage, which will face new security threats. To alleviate these issues, user authentication mechanisms can prevent unauthorized access by adversaries in the CPS environment. For this purpose, we put forth a lightweight privacy-preserving authentication scheme for 5G-enabled ecosystems. This method utilizes edge devices to save online transaction data in network and securely transmits IoT devices data from the sensors to the edge gateway. Then, the edge gateway delivers and stores these data in the cloud. A detailed comparative study based on experimental results indicates that our proposal achieves a better balance between security and functionality features, communication and capitulation costs compared to other existing competitive solutions.

The Impact of Network Design Interventions on the Security of Interdependent Systems

We study the problem of defending a Cyber-Physical System (CPS) consisting of interdependent components with heterogeneous sensitivity to investments. In addition to the optimal allocation of limited security resources, we analyze the impact of an orthogonal set of defense strategies in the form of network design interventions in the CPS to protect it against the attacker. We first propose an algorithm to simplify the CPS attack graph to an equivalent form which reduces the computational requirements for characterizing the defender's optimal security investments. We then evaluate four types of design interventions in the network in the form of adding nodes in the attack graph, interpreted as introducing additional safeguards, introducing structural redundancies, introducing functional redundancies, and introducing new functionalities. We identify scenarios in which interventions that strengthen internal components of the CPS may be more beneficial than traditional approaches such as perimeter defense. We showcase our proposed approach in two practical use cases: a remote attack on an industrial CPS and a remote attack on an automotive system. We highlight how our results closely match recommendations made by security organizations and discuss the implications of our findings for CPS design.

AGRIC: Artificial-Intelligence-Based Green Routing for Industrial Cyber–Physical System Pertaining to Extreme Environment

AGRIC: Artificial-Intelligence-Based Green Routing for Industrial Cyber–Physical System Pertaining to Extreme Environment

Bayesian and stochastic game joint approach for Cross-Layer optimal defensive Decision-Making in industrial Cyber-Physical systems

The propagation of cyber-attacks targeting modern industrial cyber-physical systems (ICPSs) is considered a sophisticated and persistent cross-layer penetration process, posing significant cyber-to-physical (C2P) risks to critical industrial infrastructures. However, existing defensive solutions for ICPSs lack the global defensive decision-making capacity against such persistent and stealthy cross-layer threats. This paper proposes a Bayesian-Stochastic hybrid game-theoretic approach that can generate optimal defensive strategies throughout the cross-layer penetration lifecycle, even without full information about attackers. Specifically, we first present a unified quantification framework to guarantee consistent utility function configurations in a cyber-physical layer integrated system. Then, we propose a state transition-based stochastic game model to characterize the dynamic evolution process of cross-layer penetration. The attack-defense interaction in each state is modeled as an incomplete-information Bayesian game to capture the inherent uncertainties of attacker information. Furthermore, a multi-agent Bayesian Q-learning (MABQL) algorithm is developed to learn the optimal defensive strategies despite incomplete information. The proposed approach is implemented and assessed based on a real-world ICPS testbed and the numerical results validate its feasibility and effectiveness for making optimal defensive decisions.

Detect Insider Attacks in Industrial Cyber-physical Systems Using Multi-physical Features-based Fingerprinting

ICPS software and hardware suffer from low update frequency, making it easier for insiders to bypass external defenses and launch concealed destructive attacks. To address these concerns, we design a device fingerprinting method based on multi-physical features, augmenting current intrusion detection techniques in the ICPS environment. In this article, we use the sorting system as an example, demonstrating that the proposed device fingerprinting technology has generality in the intrusion detection of ICPS control flow. Specifically, we first formalize the physical model of the sorting system to analyze the critical device features. Then, we extract these physical features from the sensor data collected in a physical testbed. Utilizing featurized data, we train a classifier that generates fingerprints in real-time in the production environment. Moreover, we develop a differential detection model based on device fingerprints to discover stealthy insider attacks efficiently. We evaluate the proposed method in a real-world testbed. Experiment results show that the detecting performance of classifiers approaches 100% when the the number of component types is small.

A framework for throughput bottleneck analysis using cloud-based cyber-physical systems in Industry 4.0 and smart manufacturing

The performance of a production system is primarily evaluated by its throughput, which is constrained by throughput bottlenecks. Thus, bottleneck analysis (BA), encompassing bottleneck identification, diagnosis, prediction, and prescription, is a crucial analytical process contributing to the success of manufacturing industries. Nevertheless, BA requires a substantial quantity of information from the manufacturing system, making it a data-intensive task. Based on the dynamic nature of bottlenecks, the optimal strategy for BA entails making well-informed decisions in real-time and executing necessary modifications accordingly. The efficient implementation of BA requires gathering, storing, analyzing, and illustrating data from the shop floor. Utilizing Industry 4.0 technologies, such as cyber-physical systems and cloud technology, facilitates the execution of data-intensive operations for the successful management of BA in real-world settings. The main objective of this study is to establish a framework for BA through the utilization of Cloud-Based Cyber-Physical Systems (CB-CPSs). First, a literature review was conducted to identify relevant research and current applications of CB-CPSs in BA. Using the results of the review, a CB-CPSs framework was subsequently introduced for BA. The application of the framework was assessed via simulation in a real-world manufacturer of marine engines. The findings indicate that the implementation of CB-CPSs can contribute significantly to throughput improvement.

Anomaly Detection Based on Data Super-Resolution in Industrial Cyber-Physical Systems with Multi-Rate Sampling

Anomaly Detection Based on Data Super-Resolution in Industrial Cyber-Physical Systems with Multi-Rate Sampling

Unsupervised Correlation- and Interaction-Aware Anomaly Detection for Cyber-Physical Production Systems based on Graph Neural Networks

Anomaly detection is essential for realising state-of-the-art and safe cyber-physical production systems. With the early detection of possible anomalies, the rise and the carryover of a failure throughout the entire manufacture can be prevented. Current anomaly detection approaches use sensor data from the live system only. Therefore, many important characteristics, interactions, and correlations within the system are neglected. Graphs open a particular way of modelling cyber-physical production systems, while at the same time, they allow the inclusion of a wider variation and variety of data about the system. In this study, we introduce an unsupervised correlation- and interaction-aware anomaly detection concept for cyber-physical production systems based on graph neural networks. We define a concept that allows modelling the entire system as a graph, including external knowledge. We introduce a reconstruction-based graph neural network to detect and analyse the anomalies in these graph structures. The concept is evaluated in a real industrial cyber-physical production system. The test results confirm that the presented concept can be applied to detect anomalies while concurrently considering the correlations and interactions of the system.

Dual-Reinforcement-Learning-Based Attack Path Prediction for 5G Industrial Cyber–Physical Systems

5G industrial cyber-physical systems have attracted substantial research interests due to their capability in the interconnection of everything. However, integrating the 5G network may expose systems to more potential risks. To reveal attack propagation, an attack path prediction approach based on dual reinforcement learning is proposed. First, a dual-network model is established, incorporating the security constraints for attacks against the 5G network into the attack graph. Second, employing reinforcement learning, Q-value updating functions and reward mechanisms based on topology and vulnerability are designed. Finally, an optimal attack path prediction algorithm is developed. Unlike traditional methods, the proposed approach does not rely on the monotonicity assumption that a system component has only one vulnerability, enabling it to accurately predict the optimal attack paths. Our simulation results demonstrate that the proposed approach can identify possible attack sources and paths from a 5G industrial cyber-physical system.

Fog Computing for Control of Cyber-Physical Systems in Industry Using BCI.

Brain-computer interfaces use signals from the brain, such as EEG, to determine brain states, which in turn can be used to issue commands, for example, to control industrial machinery. While Cloud computing can aid in the creation and operation of industrial multi-user BCI systems, the vast amount of data generated from EEG signals can lead to slow response time and bandwidth problems. Fog computing reduces latency in high-demand computation networks. Hence, this paper introduces a fog computing solution for BCI processing. The solution consists in using fog nodes that incorporate machine learning algorithms to convert EEG signals into commands to control a cyber-physical system. The machine learning module uses a deep learning encoder to generate feature images from EEG signals that are subsequently classified into commands by a random forest. The classification scheme is compared using various classifiers, being the random forest the one that obtained the best performance. Additionally, a comparison was made between the fog computing approach and using only cloud computing through the use of a fog computing simulator. The results indicate that the fog computing method resulted in less latency compared to the solely cloud computing approach.

Statistical knowledge and game-theoretic integrated model for cross-layer impact assessment in industrial cyber-physical systems

Cyberspace intrusions targeting modern industrial cyber-physical systems (ICPSs) are considered highly persistent and stealthy penetration processes that can result in catastrophic consequences for industrial infrastructures. Existing studies predominantly concentrated on detection and defense strategies for specific stages of intrusions without much knowledge of underlying system operational mechanisms, characteristics and evolutionary patterns. In this paper, we present a novel approach that integrates statistical knowledge and game theory to establish a comprehensive security model covering various aspects, including anomaly detection, behavioral analysis, strategy generation and impact assessment. Specifically, a statistical model, i.e. Poisson intrusion model (PIM), is developed to characterize the probabilistic properties of intrusions by leveraging knowledge of their occurrence patterns and frequencies. A Bayesian inference-based model is proposed to analyze the intrusion behaviors for anomaly detection. Then, by integrating statistical knowledge of intrusions and detections, a Markov game model is formulated to characterize the interactive actions and strategies between attackers and defenders throughout the intrusion process. Further, the cross-layer impact is assessed by quantifying the potential consequences under corresponding cyber security conditions in terms of production performance degradation and unintended incident losses. Finally, the proposed approach is validated through extensive experiments for power plant operational scenarios.

Ensemble Active Learning by Contextual Bandits for AI Incubation in Manufacturing

An Industrial Cyber-physical System (ICPS) provides a digital foundation for data-driven decision-making by artificial intelligence (AI) models. However, the poor data quality (e.g., inconsistent distribution, imbalanced classes) of high-speed, large-volume data streams poses significant challenges to the online deployment of offline-trained AI models. As an alternative, updating AI models online based on streaming data enables continuous improvement and resilient modeling performance. However, for a supervised learning model (i.e., a base learner), it is labor-intensive to annotate all streaming samples to update the model. Hence, a data acquisition method is needed to select the data for annotation to ensure data quality while saving annotation efforts. In the literature, active learning methods have been proposed to acquire informative samples. Different acquisition criteria were developed for exploration of under-represented regions in the input variable space or exploitation of the well-represented regions for optimal estimation of base learners. However, it remains a challenge to balance the exploration-exploitation trade-off under different online annotation scenarios. On the other hand, an acquisition criterion learned by AI adapts itself to a scenario dynamically, but the ambiguous consideration of the trade-off limits its performance in frequently changing manufacturing contexts. To overcome these limitations, we propose an ensemble active learning method by contextual bandits ( CbeAL ). CbeAL incorporates a set of active learning agents (i.e., acquisition criteria) explicitly designed for exploration or exploitation by a weighted combination of their acquisition decisions. The weight of each agent will be dynamically adjusted based on the usefulness of its decisions to improve the performance of the base learner. With adaptive and explicit consideration of both objectives, CbeAL efficiently guides the data acquisition process by selecting informative samples to reduce the human annotation efforts. Furthermore, we characterize the exploration and exploitation capability of the proposed agents theoretically. The evaluation results in a numerical simulation study and a real case study demonstrates the effectiveness and efficiency of CbeAL in manufacturing process modeling of the ICPS.

Asynchronous event‐triggered adaptive robust switching filtering for nonlinear industrial cyber physical systems under data injection attacks

AbstractBy fully taking the effects of data injection attacks and nonlinear disturbances into consideration, based on dynamic asynchronous event‐triggered approach, an adaptive robust H‐Infinity switching filtering (ARSF) method is proposed for the stability of industrial cyber physical systems. An adaptive feedback controller utilizing anticipated states and error‐dependent parameters can proficiently counteract the deleterious impacts of data injection attacks originating from the measurement channel. Furthermore, a filter, which includes the value of the measurement signal, is designed to monitor the system's state changes in real‐time and ensure the stable operation of the system under attacks. Based on the design of an adaptive feedback controller and a filter eliminating the effects of data injection attacks originating from the measurement channel, a switching scheme is further proposed for compensating the effect of data injection attacks in the control channel. Finally, simulation tests are conducted by using MATLAB to control the robustness of a practical quadrotor UAV system's angles by adopting ARSF, whose results verify that the correctness and effectiveness of the proposed ARSF.

Securing healthcare data in industrial cyber-physical systems using combining deep learning and blockchain technology

Industrial cyber–physical systems (ICPS) are emerging platforms for various industrial applications. For instance, remote healthcare monitoring, real-time healthcare data generation, and many other applications have been integrated into the ICPS platform. These healthcare applications encompass workflow tasks, such as processing within hospitals, laboratory tests, and insurance companies for patient payments, which necessitate a sequential flow. The external wireless, fog, and cloud services within ICPS face security issues that impact end-users’ healthcare applications. Blockchain technology offers an optimal solution for ICPS-enabled applications. However, blockchain technology for the ICPS platform is still vulnerable to cyberattacks, while microservices are essential for executing applications. This paper introduces the novel “Pattern-Proof Malware Validation” (PoPMV) algorithm designed for blockchain in ICPS. It exploits a deep learning model (LSTM) with reinforcement learning techniques to receive feedback and rewards in real-time. The primary objective is to mitigate security vulnerabilities, enhance processing speed, identify both familiar and unfamiliar attacks, and optimize the functionality of ICPS. Simulations demonstrate the superiority of the proposed approach compared to current blockchain frameworks, showcasing dynamic allocation of microservices and improved security with comprehensive attack detection by 30%.

Efficient Contrast Adjustment and Fusion Method for Underexposed Images in Industrial Cyber-Physical Systems

Owing to imaging equipment's environment and limitations, the images obtained in industrial cyber-physical systems (ICPSs) are degraded and available in various visual appearances. The process of highlighting the hidden contents in night-time contrast-distorted images is complex. Earlier approaches have solved this problem from a different perspective and achieved remarkable results that are generally unsatisfactory for images with diverse illumination distortions in ICPSs. Hence, an effective visibility enhancement model is proposed to eliminate inconsistent color casts while highlighting more hidden content for improved inspection, safety in large spaces, and monitoring of large systems. Our proposed model has four steps: 1) removal of the unnatural color cast via a white balance technique, 2) use of probability density and softplus functions to process the images with the actual color cast, 3) using an optimization algorithm to estimate the illumination and adjusting it using a nonlinear function, and 4) blending by multiscale fusion to obtain the most effective visual result. Evaluation of ten benchmark datasets using 14 quality metrics for 22 conventional and modern algorithms shows that our approach is robust, flexible, and applicable to numerous vision-based applications, such as ICPSs, autonomous vehicles, smart cameras, smart mobility, and transportation, especially in low-light environments.