Human-centricity is the core value behind the evolution of manufacturing towards Industry 5.0. Nevertheless, there is a lack of architecture that considers safety, trustworthiness, and human-centricity at its core. Therefore, we propose an architecture that integrates Artificial Intelligence (Active Learning, Forecasting, Explainable Artificial Intelligence), simulated reality, decision-making, and users' feedback, focussing on synergies between humans and machines. Furthermore, we align the proposed architecture with the Big Data Value Association Reference Architecture Model. Finally, we validate it on three use cases from real-world case studies.

Intrusion detection systems (IDSs) are essential elements when it comes to the protection of an ICT infrastructure. Misuse IDSs a stable method that can achieve high attack detection rates (ADR), while keeping false alarm rates under acceptable levels. However, misuse IDSs suffer from the lack of agility, as they are unqualified to adapt to new and “unknown” environments. That is, such an IDS puts a security administrator into an intensive engineering task for keeping the IDS up-to-date every time it faces efficiency drops. Considering the extended size of modern networks and the complexity of big network traffic data, the problem exceeds by far the limits of human managing capabilities. In this regard, we propose a novel methodology which combines the benefits of self-taught learning and MAPE-K frameworks to deliver a scalable, self-adaptive and autonomous misuse IDS. Our methodology enables a misuse IDS to sustain a high ADR even if it is imposed to consecutive and drastic environmental changes. Through the utilization of deep-learning based methods, the IDS is able to grasp an attack’s nature based on generalized features reconstructions stemming directly from the unknown environment and its unlabeled data. The experimental results reveal that our methodology can breathe new life into the IDS without the constant need of manually refreshing its training set. We evaluate our proposal under several classification metrics, and we show that it is able to increase the ADR of the IDS up to 73.37% in critical situations where a statically trained IDS is rendered totally ineffective.

Addressing cyber and privacy risks has never been more critical for organisations. While a number of risk assessment methodologies and software tools are available, it is most often the case that one must, at least, integrate them into a holistic approach that combines several appropriate risk sources as input to risk mitigation tools. In addition, cyber risk assessment primarily investigates cyber risks as the consequence of vulnerabilities and threats that threaten assets of the investigated infrastructure. In fact, cyber risk assessment is decoupled from privacy impact assessment, which aims to detect privacy-specific threats and assess the degree of compliance with data protection legislation. Furthermore, a Privacy Impact Assessment (PIA) is conducted in a proactive manner during the design phase of a system, combining processing activities and their inter-dependencies with assets, vulnerabilities, real-time threats and Personally Identifiable Information (PII) that may occur during the dynamic life-cycle of systems. In this paper, we propose a cyber and privacy risk management toolkit, called AMBIENT (Automated Cyber and Privacy Risk Management Toolkit) that addresses the above challenges by implementing and integrating three distinct software tools. AMBIENT not only assesses cyber and privacy risks in a thorough and automated manner but it also offers decision-support capabilities, to recommend optimal safeguards using the well-known repository of the Center for Internet Security (CIS) Controls. To the best of our knowledge, AMBIENT is the first toolkit in the academic literature that brings together the aforementioned capabilities. To demonstrate its use, we have created a case scenario based on information about cyber attacks we have received from a healthcare organisation, as a reference sector that faces critical cyber and privacy threats.

The technological leap of smart technologies and the Internet of Things has advanced the conventional model of the electrical power and energy systems into a new digital era, widely known as the Smart Grid. The advent of Smart Grids provides multiple benefits, such as self-monitoring, self-healing and pervasive control. However, it also raises crucial cybersecurity and privacy concerns that can lead to devastating consequences, including cascading effects with other critical infrastructures or even fatal accidents. This paper introduces a novel architecture, which will increase the Smart Grid resiliency, taking full advantage of the Software-Defined Networking (SDN) technology. The proposed architecture called SDN-microSENSE architecture consists of three main tiers: (a) Risk assessment, (b) intrusion detection and correlation and (c) self-healing. The first tier is responsible for evaluating dynamically the risk level of each Smart Grid asset. The second tier undertakes to detect and correlate security events and, finally, the last tier mitigates the potential threats, ensuring in parallel the normal operation of the Smart Grid. It is noteworthy that all tiers of the SDN-microSENSE architecture interact with the SDN controller either for detecting or mitigating intrusions.

Adversarial machine learning (AML) is a class of data manipulation techniques that cause alterations in the behavior of artificial intelligence (AI) systems while going unnoticed by humans. These alterations can cause serious vulnerabilities to mission-critical AI-enabled applications. This work introduces an AI architecture augmented with adversarial examples and defense algorithms to safeguard, secure, and make more reliable AI systems. This can be conducted by robustifying deep neural network (DNN) classifiers and explicitly focusing on the specific case of convolutional neural networks (CNNs) used in non-trivial manufacturing environments prone to noise, vibrations, and errors when capturing and transferring data. The proposed architecture enables the imitation of the interplay between the attacker and a defender based on the deployment and cross-evaluation of adversarial and defense strategies. The AI architecture enables (i) the creation and usage of adversarial examples in the training process, which robustify the accuracy of CNNs, (ii) the evaluation of defense algorithms to recover the classifiers’ accuracy, and (iii) the provision of a multiclass discriminator to distinguish and report on non-attacked and attacked data. The experimental results show promising results in a hybrid solution combining the defense algorithms and the multiclass discriminator in an effort to revitalize the attacked base models and robustify the DNN classifiers. The proposed architecture is ratified in the context of a real manufacturing environment utilizing datasets stemming from the actual production lines.

As the upsurge of information and communication technologies has become the foundation of all modern application domains, fueled by the unprecedented amount of data being processed and exchanged, besides security concerns, there are also pressing privacy considerations that come into play. Compounding this issue, there is currently a documented gap between the cybersecurity and privacy risk assessment (RA) avenues, which are treated as distinct management processes and capitalise on rather rigid and make-like approaches. In this paper, we aim to combine the best of both worlds by proposing the APSIA (Automated Privacy and Security Impact Assessment) methodology, which stands for Automated Privacy and Security Impact Assessment. APSIA is powered by the use of interdependency graph models and data processing flows used to create a digital reflection of the cyber-physical environment of an organisation. Along with this model, we present a novel and extensible privacy risk scoring system for quantifying the privacy impact triggered by the identified vulnerabilities of the ICT infrastructure of an organisation. We provide a prototype implementation and demonstrate its applicability and efficacy through a specific case study in the context of a heavily regulated sector (i.e., assistive healthcare domain) where strict security and privacy considerations are not only expected but mandated so as to better showcase the beneficial characteristics of APSIA. Our approach can complement any existing security-based RA tool and provide the means to conduct an enhanced, dynamic and generic assessment as an integral part of an iterative and unified risk assessment process on-the-fly. Based on our findings, we posit open issues and challenges, and discuss possible ways to address them, so that such holistic security and privacy mechanisms can reach their full potential towards solving this conundrum.