Modern Cyber-physical Systems Research Articles

From Near-Sensor to In-Sensor: A State-of-the-Art Review of Embedded AI Vision Systems.

In modern cyber-physical systems, the integration of AI into vision pipelines is now a standard practice for applications ranging from autonomous vehicles to mobile devices. Traditional AI integration often relies on cloud-based processing, which faces challenges such as data access bottlenecks, increased latency, and high power consumption. This article reviews embedded AI vision systems, examining the diverse landscape of near-sensor and in-sensor processing architectures that incorporate convolutional neural networks. We begin with a comprehensive analysis of the critical characteristics and metrics that define the performance of AI-integrated vision systems. These include sensor resolution, frame rate, data bandwidth, computational throughput, latency, power efficiency, and overall system scalability. Understanding these metrics provides a foundation for evaluating how different embedded processing architectures impact the entire vision pipeline, from image capture to AI inference. Our analysis delves into near-sensor systems that leverage dedicated hardware accelerators and commercially available components to efficiently process data close to their source, minimizing data transfer overhead and latency. These systems offer a balance between flexibility and performance, allowing for real-time processing in constrained environments. In addition, we explore in-sensor processing solutions that integrate computational capabilities directly into the sensor. This approach addresses the rigorous demand constraints of embedded applications by significantly reducing data movement and power consumption while also enabling in-sensor feature extraction, pre-processing, and CNN inference. By comparing these approaches, we identify trade-offs related to flexibility, power consumption, and computational performance. Ultimately, this article provides insights into the evolving landscape of embedded AI vision systems and suggests new research directions for the development of next-generation machine vision systems.

Codesign of reactor-oriented hardware and software for cyber-physical systems

Modern cyber-physical systems often make use of heterogeneous systems-on-chip with reconfigurable logic to provide adequate computing power and flexible I/O. However, modeling, verifying, and implementing the computations spanning CPUs and reconfigurable logic is still challenging. The hardware and software components are often designed by different teams and at different levels of abstraction, making it hard to reason about the resulting computation. We propose to lift both hardware and software design to the same level of abstraction by using the Lingua Franca coordination language. Lingua Franca is based on a sparse synchronous model that allows modeling concurrency and timing while keeping a sequential model for the actual computation. We define hardware reactors as a subset of the reactor model of computation underlying Lingua Franca. We also present and evaluate reactor-chisel, a hardware runtime implementing the semantics of hardware reactors, and an extension to the Lingua Franca compiler enabling reactor-oriented hardware-software codesign.

Elements of Timed Pattern Matching

The rise of machine learning and cloud technologies has led to a remarkable influx of data within modern cyber-physical systems. However, extracting meaningful information from this data has become a significant challenge due to its volume and complexity. Timed pattern matching has emerged as a powerful specification-based runtime verification and temporal data analysis technique to address this challenge. In this paper, we provide a comprehensive tutorial on timed pattern matching that ranges from the underlying algebra and pattern specification languages to performance analyses and practical case studies. Analogous to textual pattern matching, timed pattern matching is the task of finding all time periods within temporal behaviors of cyber-physical systems that match a predefined pattern. Originally we introduced and solved several variants of the problem using the name of match sets, which has evolved into the concept of timed relations over the past decade. Here we first formalize and present the algebra of timed relations as a standalone mathematical tool to solve the pattern matching problem of timed pattern specifications. In particular, we show how to use the algebra of timed relations to solve the pattern matching problem for timed regular expressions and metric compass logic in a unified manner. We experimentally demonstrate that our timed pattern matching approach performs and scales well in practice. We further provide in-depth insights into the similarities and fundamental differences between monitoring and matching problems as well as regular expressions and temporal logic formulas. Finally, we illustrate the practical application of timed pattern matching through two case studies, which show how to extract structured information from temporal datasets obtained via simulations or real-world observations. These results and examples show that timed pattern matching is a rigorous and efficient technique in developing and analyzing cyber-physical systems.

A Machine Learning Approach Towards Cyber-Physical Security of Battery Systems

Abstract Modern battery systems exhibit a cyber-physical nature due to the extensive use of communication technologies in battery management. This makes modern cyber-physical battery systems vulnerable to cyber threats where an adversary can manipulate sensing and actuation signals to satisfy certain malicious objectives. In this work, we present a machine learning-based approach to enable resilience to adversarial attacks by detecting and estimating the attack and subsequently taking corrective action to mitigate the attack. In particular, we focus on false data injection type attacks on battery systems. The overall diagnostic algorithm consists of an adaptive boosting-based attack detector, a Long Short Term Memory (LSTM) neural network-based attack estimator, and a corrective controller. The proposed algorithm is trained and tested by utilizing data from a complex battery electrochemical battery simulator. Simulation results are presented to verify the effectiveness of the approach.

An Approach to Improving Availability of Microservices for Cyber-Physical Systems

The design of modern Cyber-Physical Systems (CPS) connects physical and digital realms from cloud systems to edge devices. Microservice architecture has been widely used for IT solutions and emerges as a promising approach for supporting CPS that are more efficient, adaptable, and interconnected. However, there is an increasing need to improve the availability, reliability, and resilience of microservice systems according to the needs. This paper summarizes the challenges and drawbacks of microservice architecture used for CPS. Then, the simplified microservice model has been created, initial properties have been defined, and an improvement plan has been presented. The microservice model’s availability has been improved using a novel approach with endpoint containerization. Then, the discussion and conclusions have been offered to explore the full potential of integrating the physical and digital realms.

AD-NEv: A Scalable Multilevel Neuroevolution Framework for Multivariate Anomaly Detection.

Anomaly detection tools and methods present a key capability in modern cyberphysical and failure prediction systems. Despite the fast-paced development in deep learning architectures for anomaly detection, model optimization for a given dataset is a cumbersome and time-consuming process. Neuroevolution could be an effective and efficient solution to this problem, as a fully automated search method for learning optimal neural networks, supporting both gradient and nongradient fine-tuning. However, existing methods mostly focus on optimizing model architectures without taking into account feature subspaces and model weights. In this work, we propose anomaly detection neuroevolution (AD-NEv)-a scalable multilevel optimized neuroevolution framework for multivariate time-series anomaly detection. The method represents a novel approach to synergically: 1) optimize feature subspaces for an ensemble model based on the bagging technique; 2) optimize the model architecture of single anomaly detection models; and 3) perform nongradient fine-tuning of network weights. An extensive experimental evaluation on widely adopted multivariate anomaly detection benchmark datasets shows that the models extracted by AD-NEv outperform well-known deep learning architectures for anomaly detection. Moreover, results show that AD-NEv can perform the whole process efficiently, presenting high scalability when multiple graphics processing units (GPUs) are available.

Learning System Descriptions for Cyber-Physical Systems

Fault diagnosis algorithms compute faulty components by comparing actual observations against some model of known behaviour. A major challenge for fault diagnosis lies in creating such a suitable model. In the past, models were usually assumed to be given by experts. But in modern cyber-physical systems this assumption cannot be held, as experts are expensive and system architectures may be subject to change. This article presents a novel algorithm to obtain those models automatically and apply them for fault diagnosis. The evaluation was done on the Tennessee Eastman process and on two benchmarks of multiple-tank systems.

HYPERTEXT CASE MODEL FOR TEACHING SEMANTIC READING TO SECONDARY SCHOOL STUDENTS IN MATH LESSONSPROBLEMS OF MANAGING A CYBER-PHYSICAL SYSTEM IN THE PRESENCE OF MALFUNCTIONS IN IT

The study of various properties of information-technical systems from the point of view of interaction of their physical and digital components is a new and actual direction of modern science of cyber-physical systems. When considering modern cyber-physical systems, various authors recommend considering them as complex systems within the framework of system analysis. Due to the complexity of systems, adaptive algorithms should be relied upon in human-machine interaction. In this case, the specialist is seen as a decision maker. In the course of his work, he will strive to ensure the required mode of functioning of the cyber-physical system, including in the presence of faults in it. This paper proposes to consider the possible actions of a specialist in a cyber-physical system from the perspective of a simulation experiment. The paper provides an illustration of an algorithm scheme related to adaptive resource selection in a cyber-physical system. The scheme of rational decision making when resources are selected is shown. A methodology for generating logic models is proposed. The paper shows how game modeling can be applied to cyber-physical systems. The results obtained in this paper can be applied to a wide range of cyberphysical systems.

Digital Twins for Anomaly Detection in the Industrial Internet of Things: Conceptual Architecture and Proof-of-Concept

Modern cyber-physical systems based on the Industrial Internet of Things (IIoT) can be highly distributed and heterogeneous, and that increases the risk of failures due to misbehavior of interconnected components, or other interaction anomalies. In this paper, we introduce a conceptual architecture for IIoT anomaly detection based on the paradigms of Digital Twins (DT) and Autonomic Computing (AC), and we test it through a proof-of-concept of industrial relevance. The architecture is derived from the current state-of-the-art in DT research and leverages on the MAPE-K feedback loop of AC in order to monitor, analyze, plan, and execute appropriate reconfiguration or mitigation strategies based on the detected deviation from prescriptive behavior stored as shared knowledge. We demonstrate the approach and discuss results by using a reference operational scenario of adequate complexity and criticality within the European Railway Traffic Management System.

Continuous agile cyber–physical systems architectures based on digital twins

Modern cyber-physical systems, for the most part, are large-scale multilevel heterogeneous distributed systems that integrate subsystems of different kinds and are built on the Internet of Things platforms, where system structure and behavior are not constant. Managing such systems and keeping them in working condition throughout their lifetime is a difficult task. The proposed article discusses one of the possible approaches to solving this problem, based on the use of well-known continuous and agile architecture paradigms. However, there are currently no effective mechanisms for implementing these paradigms. The proposed article suggests a new approach to implementing continuous agile architectures by utilizing digital twins and proposes a reference architecture for a run-time dynamic digital twin. This method is unique because it builds a series of dynamic digital twins that model the system in real time, utilizing data about system events. Build the first models using the models used in earlier stages of the system lifecycle. This gives the following opportunities: i) a way to use dynamic digital twins to implement the continuous agile architecture paradigm; ii) a generalized three-level model of the life cycle of the continuous agile architecture; iii) a reference architecture for dynamic digital twins; and iv) a set of models that are all about using dynamic digital twins. The suggested approach enables the management of heterogeneous multilevel cyber-physical systems with variable structure and behavior variability.

High-Speed Network DDoS Attack Detection: A Survey.

Having a large number of device connections provides attackers with multiple ways to attack a network. This situation can lead to distributed denial-of-service (DDoS) attacks, which can cause fiscal harm and corrupt data. Thus, irregularity detection in traffic data is crucial in detecting malicious behavior in a network, which is essential for network security and the integrity of modern Cyber-Physical Systems (CPS). Nevertheless, studies have shown that current techniques are ineffective at detecting DDoS attacks on networks, especially in the case of high-speed networks (HSN), as detecting attacks on the latter is very complex due to their fast packet processing. This review aims to study and compare different approaches to detecting DDoS attacks, using machine learning (ML) techniques such as k-means, K-Nearest Neighbors (KNN), and Naive Bayes (NB) used in intrusion detection systems (IDSs) and flow-based IDSs, and expresses data paths for packet filtering for HSN performance. This review highlights the high-speed network accuracy evaluation factors, provides a detailed DDoS attack taxonomy, and classifies detection techniques. Moreover, the existing literature is inspected through a qualitative analysis, with respect to the factors extracted from the presented taxonomy of irregular traffic pattern detection. Different research directions are suggested to support researchers in identifying and designing the optimal solution by highlighting the issues and challenges of DDoS attacks on high-speed networks.

Designing distributed decision-making authorities for smart factories – understanding the role of manufacturing network architecture

The availability of cyber-physical systems (CPS) in modern manufacturing networks provides a multitude of interesting opportunities from a manufacturing control perspective. Providing sensors, data gathering, local computation and communication capabilities modern CPS fulfil the technical requirements to act completely autonomously in a manufacturing network. While the distribution of decision-making authority to autonomous entities is feasible given such requirements, practice often sees the monopolisation of decision-making authority for centralised control. However, distributed production control approaches might be better suited given current manufacturing challenges, ranging from unreliable supply chains over highly volatile markets, to the demand for increasingly efficient and highly customisable production. In this article, we extend an existing scheduling complexity framework which enables practitioners and researchers alike to assess the aptitude of given manufacturing networks for both centralised and distributed control. In particular, we study the influence of a manufacturing network's topology ranging from assembly line to job shops on the aforementioned aptitude, with total production costs as objective. We utilise a multi-agent-based discrete-event simulation comparing an MILP-based centralised control approach and an autonomy based distributed control approach with weighted costs as decision function to evaluate this framework. Our results provide novel insights regarding the influence of manufacturing network topologies on the scheduling complexity of manufacturing networks.

SAFLA: Scheduling Multiple Real-Time Periodic Task Graphs on Heterogeneous Systems

Many modern Cyber Physical Systems (CPSs) are composed of multiple independent periodically executing real-time control tasks having inter-dependent component sub-tasks. Each such control task is therefore usually represented as <i>Directed-acyclic Task Graphs</i> (DTGs). These CPSs are often distributed in nature and are quickly shifting from homogeneous to heterogeneous processing platforms in order to meet ever increasing demands for performance and energy savings, within limited resource budgets. In spite of the practical relevance of the problem in today's CPS design scenario, very few research works in literature have tried to address this due to its inherent computational as well as design complexity. This work endeavors to solve the problem of co-scheduling a set of periodic real-time applications each modelled as an independent DTG, to be executed on a distributed platform consisting of heterogeneous processors communicating using shared buses. Assuming the processing platform to be DVFS (Dynamic Voltage Frequency Scaling) enabled, we attempt to minimize dynamic energy dissipation associated with the execution of all DTGs over an hyperperiod <inline-formula><tex-math notation="LaTeX">$\mathcal {H}$</tex-math></inline-formula> while ensuring that no DTG instance within <inline-formula><tex-math notation="LaTeX">$\mathcal {H}$</tex-math></inline-formula> misses its deadline. The problem has first been formally represented as a constraint optimization problem. However, an optimal solution using standard solvers become prohibitively compute as well as memory intensive and doesn't scale even for moderate problem sizes. Hence, in this work, we attempt to develop a three-phase list-based hierarchical scheduling algorithm called <i>Slack Aware Frequency Level Allocator</i> ( <i>SAFLA</i> ). The efficacy of <i>SAFLA</i> has been critically evaluated through simulation using benchmark DTGs.

FalsifAI: Falsification of AI-Enabled Hybrid Control Systems Guided by Time-Aware Coverage Criteria

Modern Cyber-Physical Systems (CPSs) that need to perform complex control tasks (e.g., autonomous driving) are increasingly using AI-enabled controllers, mainly based on deep neural networks (DNNs). The quality assurance of such types of systems is of vital importance. However, their verification can be extremely challenging, due to their complexity and uninterpretable decision logic. Falsification is an established approach for CPS quality assurance, which, instead of attempting to prove the system correctness, aims at finding a time-variant input signal violating a formal specification describing the desired behavior; it often employs a search-based testing approach that tries to minimize the <i>robustness</i> of the specification, given by its quantitative semantics. However, guidance provided by robustness is mostly black-box and only related to the system output, but does not allow to understand whether the temporal internal behavior determined by multiple consecutive executions of the neural network controller has been explored sufficiently. To bridge this gap, in this paper, we make an early attempt at exploring the temporal behavior determined by the repeated executions of the neural network controllers in hybrid control systems and first propose eight time-aware coverage criteria specifically designed for neural network controllers in the context of CPS, which consider different features by design: the simple temporal activation of a neuron, the continuous activation of a neuron for a given duration, and the differential neuron activation behavior over time. Secondly, we introduce a falsification framework, named <inline-formula><tex-math notation="LaTeX">$\mathtt {FalsifAI}$</tex-math></inline-formula>, that exploits the coverage information for better falsification guidance. Namely, inputs of the controller that increase the coverage (so improving the <i>exploration</i> of the DNN behaviors), are prioritized in the <i>exploitation</i> phase of robustness minimization. Our large-scale evaluation over a total of 3 typical CPS tasks, 6 system specifications, 18 DNN models and more than 12,000 experiment runs, demonstrates 1) the advantage of our proposed technique in outperforming two state-of-the-art falsification approaches, and 2) the usefulness of our proposed time-aware coverage criteria for effective falsification guidance.

An Automatic Assembly SMT Production Line Operation Technological Process Simulation Model Development

This article describes an automatic assembly SMT production line operation technological process simulation model development within the framework of modern cyber-physical production systems. In the course of the research, the authors conduct an analysis of modern publications on the topic of research, and the need to find new approaches for modeling the work of a production line at the initial stages of design. This makes it possible to assess the load on each element of the system at the design stage, and identify disadvantages at the early stages of production planning. The authors give an example of the task solution for modeling a fragment of an automated production line of SMT assembly, based on the use of typical mathematical schemes of mass service systems (MSS). A description of the functioning of the technological link of the process is carried out, and a Q-diagram of the technological process of SMT assembly is developed. Using the simulation modeling language GPSS and the technological parameters of the selected equipment, a block diagram was developed, on the basis of which the program was written and the work of the SMT assembly production line was simulated. The obtained results further allow us to estimate the production capacity, the loading of each equipment, the functioning processes of both the selected fragment, and the developed model and they can be applied to the simulation of large production lines within the framework of PCB production.

A Blockchained AutoML Network Traffic Analyzer to Industrial Cyber Defense and Protection

Network traffic analysis can raise privacy concerns due to its ability to reveal sensitive information about individuals and organizations. This paper proposes a privacy-preserving Block-chained AutoML Network Traffic Analyzer (BANTA). The system securely stores network traffic logs in a decentralized manner, providing transparency and security. Differential privacy algorithms protect sensitive information in the network flow logs while allowing administrators to analyze network traffic without the risk of leakages. The BANTA uses blockchain technology, where smart contracts automate the process of network traffic analysis, and a multi-signature system ensures the system’s security, safety, and reliability. The proposed approach was evaluated using a real-world network traffic dataset. The results demonstrate the system’s high accuracy and real-time anomaly detection capabilities, which makes it well-suited for scalable cybersecurity operations. The system’s privacy protection, decentralized storage, automation, multi-signature system, and real-world effectiveness ensure that the organization’s data is private, secure, and effectively protected from cyber threats, which are the most vexing issue of modern cyber-physical systems.

Digital twins in cyber effects modelling of IoT/CPS points of low resilience

The exponential increase of data volume and velocity have necessitated a tighter linkage of physical and cyber components in modern Cyber–physical systems (CPS) to achieve faster response times and autonomous component reconfiguration. To attain this degree of efficiency, the integration of virtual and physical components reinforced by artificial intelligence also promises to improve the resilience of these systems against organised and often skillful adversaries. The ability to visualise, validate, and illustrate the benefits of this integration, while taking into account improvements in cyber modelling and simulation tools and procedures, is critical to that adoption. Using Cyber Modelling and Simulation (M&S) this study evaluates the scale and complexity required to achieve an acceptable level of cyber resilience testing in an IoT-enabled critical national infrastructure (CNI). This research focuses on the benefits and challenges of integrating cyber modelling and simulation (M&S) with digital twins and threat source characterisation methodologies towards a cost-effective security and resilience assessment. Using our dedicated DT environment, we show how adversaries can utilise cyber–physical systems as a point of entry to a broader network in a scenario where they are trying to attack a port.

Optimization strategies for GPUs: an overview of architectural approaches

Modern Cyber Physical Systems (CPS) applications require hardware capable of optimized performance-per-watt efficency. This is usually obtained through massively parallel accelerators such as the GPU. Recent research is therefore investigating novel designs to optimize GPU energy consumption and performance for various applications in the Internet-of-things, autonomous navigation, and industrial robotics domains. This paper presents a survey of the current state-of-the-art approaches for optimizing GPU performance metrics; we present a complete and up-to-date summary of ideas, mechanisms, and potential improvements for next-generation GPU devices.

Quantifying the econometric loss of a cyber-physical attack on a seaport

Modern Cyber-Physical Systems (CPS) show increasing levels of automation, intelligence, and optimization capabilities. Both creators and adopters of these technologies emphasize the benefits: safety, cost-savings, and efficiencies. Existing literature under-represents the potential risks associated with the adoption of individually complex systems and interconnected Systems-of-Systems (SoS). Research primarily focuses on the potential benefits, and many recent deployments of new technology have received little rigorous scrutiny. The studies that do explore issues, such as cyber-attacks on CPS, often lack an analysis of the potential impact based on the way cyber-physical risks are communicated. Due to high levels of interaction and inter dependencies in CPS, disruptions in one sector can propagate to other sectors. The initial economic impacts may start locally but will often lead to global effects if response mechanisms prove insufficient. This case study aims to address a research gap in modeling, and quantifies how a cyber-attack, with physical consequences, can affect local and global trade if a single maritime port is affected. Additionally, to further understanding of the public's perception of risk about cyber-attacks and its emergent threat to global supply chains, members of the public were asked to react to the potential econometric losses. This study produced the five-part CyPEM (Cyber Physical Econometric Model) that has the capability to translate a cyber-attack to an econometric loss. Each framework part may also be swapped out to analyze different CPS. With this, we discovered a gap in the perceptions of cyber-physical risk regarding maritime transport.

Advantages of the usage of the Infinity Computer for reducing the Zeno behavior in hybrid system models

To capture the dynamics of modern Cyber-Physical Systems, hybrid system models are introduced to combine their continuous dynamics with the discrete ones. Unfortunately, one important negative issue can affect hybrid system models: the so-called Zeno phenomenon, which results in an infinite number of discrete transitions in a finite amount of time occurring during the model’s simulation that leads to inconsistent results. In this context, the paper investigates the use of a recently proposed numerical algorithm, based on the Infinity Computer methodology, to handle the Zeno phenomenon and evaluate it with respect to standard numerical methods by considering the hybrid system models of two exemplary Cyber-Physical Systems: the Water tanks and the Thermostat.