Human-machine collaboration requires unambiguous communication to limit misunderstandings. Although semantic interoperability manages to remove ambiguity in machine-to-machine communication, it is insufficient when humans are involved. Humans process and understand information differently based on past experience and the current context, exceeding semantic interoperability’s scope. Cognitive interoperability aims to achieve an aligned understanding, shared intentions, and enable joint decision-making between agents. However, the cognitive state of the human is hard to detect and model, representing a major obstacle to cognitive interoperability. We propose a cognitive Human Digital Twin (cHDT) that emulates a human’s cognitive processes by exploiting cognitive architectures. In particular, we investigate ACT-R as a candidate model. It is a mature cognitive architecture that has been developed based on decades of experimental results from cognitive science and neuroscience. We discuss how the internal state of ACT-R models, and thus the cHDTs, may contribute to cognitive interoperability. With a simplified use case, we illustrate how a cHDT hosting a personalised ACT-R model could track and continuously share the human’s internal cognitive states. This enables external systems, such as robots, to adapt to human perspectives and avoid resource conflicts in human-robot collaboration. Finally, we discuss the applicability of ACT-R as an emulation model, the components of a cHDT, and outline a two-phase implementation plan to validate the proposed solution.

In this paper, we present a comprehensive ontology of human cognition and abilities, designed as a formal framework for digital twins. This model enables digital twins of humans to replicate aspects of human cognition, while cognitive digital twins enhance cyber-physical systems with human-like reasoning and intelligence. The cognition meta-model results from the integration of multiple perspectives on cognition in Neuropsychology, Education Sciences, Engineering Sciences, Cognitive Informatics, and Cognitive Architectures, including indirectly also perspectives of Cognitive Sciences and Artificial Intelligence. Our Human Ontology (HUMO) is an extension of the SOMA (Socio-physical Model of Activities) ontology, which serves as a basis to design cases for cognitive robot - human collaboration. It aims at being exploited by digital twins in their representation of the world and the entity they twin.

In the context of Industry 5.0, where human-centred collaboration is paramount, effective interaction between humans and Cyber–Physical Systems (CPS) requires us to reconsider the traditional concept of interoperability. Although technical, syntactic and semantic interoperability facilitated system integration in Industry 4.0, they are insufficient when humans become active collaborators. In particular, semantic interoperability fails to capture the cognitive and contextual nuances of human reasoning, often resulting in gaps in mutual understanding. This article discusses and defines cognitive interoperability, which goes beyond semantic alignment by integrating shared perception, contextual modelling, and reasoning mechanisms to support seamless collaboration between humans and CPS. First, we provide a comprehensive analysis of how cognitive interoperability has been defined in various fields, offering a unified and clarified perspective. Next, we present a structured diagram detailing the cognitive functions necessary to achieve cognitive interoperability between humans and CPSs. Finally, we examine a real-world use case in the field of Human–Robot Collaboration (HRC) to illustrate the practical implications of this concept. Through this example, we demonstrate how cognitive interoperability can address the limitations of semantic interoperability in dynamic and cooperative contexts by enabling mutual understanding and adaptive interaction. Our work contributes to advancing human-centred intelligent systems and provides a foundation for designing next-generation collaborative manufacturing environments.

Semantic web models such as ontologies and knowledge graphs were designed to represent knowledge explicitly and to infer implicit knowledge. This brings several benefits like flexibility and reasoning power. On the other hand, since the adoption of these models has been growing during the last years, the demand for consistency has risen in parallel. Therefore, the validation of these representations started to be an important issue for research. Several types of related works have been done; however, there is still a lack of an integrated review to understand the state of the art of formal semantic constraint validation. Addressing this issue is crucial for broader adoption of both the models discussed and the Semantic Web itself. For the community, it is important to review existing progress, outline future avenues, and examine related areas that can benefit from these developments. The contribution of this article includes a taxonomy of related research trends, a classification of selected works into these categories, and an overview of the main open challenges in validating semantic constraints for the Semantic Web. Each of these contributions was obtained through the conduction of a systematic literature review. In addition, an extension of the study is done to consider a very promising model which has a growing intersection with the field: property graphs. Finally, this paper concludes with an outlook that summarizes the contributions made and the challenges that the authors decided to continue researching in further steps.

This paper explores the combination of ontological reasoning and Deep Reinforcement Learning (DRL) as an approach for automated robot capability learning. While ontologies enable robots to autonomously choose and perform tasks through environmental reasoning, they cannot execute unprogrammed actions. Complementarily, DRL cannot be used to reason and make decisions, and it requires predefined objectives and constraints, but can support new behaviour learning to be reprogrammed into the robot. We propose an ontology-guided DRL framework that automatically initializes learning parameters when robots encounter unfamiliar tasks. Using an ontology for capability inference, the system maps ontological knowledge to DRL components including state space, action space, and reward functions. We illustrate with a collaborative assembly scenario, where a robot is not pre-programmed to screw. The ontology is used to infer the feasibility of the task and the required capabilities, while the DRL component is dynamically configured based on this information, allowing the robot to learn the skill autonomously. Copyright (c) 2025 The Authors. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0/)

Decision-making can be significantly enhanced by knowledge extraction techniques such as Formal Concept Analysis (FCA), which provides a structured way to analyze and organize data. However, in distributed environments, information is often fragmented, arriving in streams rather than as a complete dataset. Consulting the entire dataset at once is not always feasible due to time and resource constraints. While an existing batch algorithm allows concept lattice computation without requiring full attribute knowledge, batch processing is inherently unsuitable for dynamic, continuously evolving data. In this article, we propose an incremental algorithm for concept lattice computation in arbitrarily distributed formal contexts, which can efficiently update the lattice as new information becomes available. Our method ensures that knowledge extraction remains computationally feasible even when data is incomplete, distributed, or evolving over time. We further analyze the complexity of our approach in comparison to the existing batch-based distributed algorithm, highlighting its advantages in stream processing scenarios. By addressing the limitations of batch computation and enabling real-time lattice updates, our work contributes to enhancing FCA's applicability in distributed and dynamic knowledge systems. The proposed incremental approach paves the way for more adaptive, efficient, and scalable knowledge extraction methods, particularly in fields requiring real-time decision-making and pattern discovery.

Cognition, the set of mental processes that enable humans to perceive, reason, learn and decide, plays an essential role in effective collaboration between humans and Cyber-Physical Systems (CPSs). To achieve seamless cognitive interoperability between humans and CPSs, it is necessary to integrate a Cognitive Digital Twin (CDT) and a Human Digital Twin (HDT) to provide digital representations of both physical assets and human cognitive states. In this article, we first analyse the three essential functions of CDT and HDT: emulation, cognition and simulation, and review the state-of-the-art technologies for each of them, from supervised learning and knowledge graphs to deep reinforcement learning. Focusing on the cognitive layer, we review the state of the art in cognitive architectures, describing their symbolic, sub-symbolic and hybrid types and reporting on their real-world implementations in different domains. We then assess the relevance of these architectures for the integration of human-like reasoning in CDTs. Finally, we identify the main technological challenges and gaps that need to be addressed in order to implement fully operational CDTs.

In modern Cyber-Physical Enterprises, the place of cognition to build more flexible, autonomous, self-adaptive and ultimately more intelligent systems remains unclear. However it could be an essential element to ensure interoperability, as highlighted in recent works on cognitive interoperability. Assuming an environment where cyber-physical systems and humans are coupled with their individual digital twins, we discuss where cognition can exist in the couple and how Cognitive Digital Twin and Cyber-Physical Systems can emerge. This is supported by a proposition for a high-level architecture for Entity- Digital Twin coupling. The place of Cognitive Architectures as a special model of cognition is discussed. Exploring the integration with the ISO23247 Digital Twin framework for manufacturing, we also highlight its lacks to include cognition and humans.

Human Digital Twins (HDT) are a key technology to enable human-centric manufacturing systems. This paper presents a novel meta-model for HDTs, grounded in general systems theory, to provide a unified and robust conceptualisation for HDT development. The model describes four essential interacting systems categorised by the physical or digital nature of their components; (i) the Human Individual is a physical system composed of a physiological, cognitive and mechanical sub-system among others; (ii) the Human Digital Twin is a digital system composed of a human model, data storage and management capabilities, and functions to generate feedback for the human; (iii) Sensors are cyber-physical systems capturing observations about the human and digitising them for the HDT; (iv) Human-Machine Interaction Devices are cyber-physical systems enabling the human to receive feedback form the HDT and interact with it. The systemic grounding provides a general model of an HDT as an extension of a DT representing any kind of system. It clarifies the relations and interfaces between its four main sub-systems, as well as the data flows and feedback mechanism. The HDT's Communication Interface is identified as key component to integrate HDTs into larger digital ecosystems and enable human-centric system development. Copyright (c) 2025 The Authors.

Human Digital Twins (HDTs) are an emerging concept with the potential to create human-centric systems for Industry 5.0. The concept has rapidly spread to new application domains, most notably Healthcare, leading to diverging conceptual interpretations. This Systematic Literature Review analyses the conceptual understanding of HDTs across all application domains to clarify the conceptual foundation. Our review reveals a consensus that an HDT’s twinned entity is a human individual. However, there is little agreement on the data flows between the individual and their HDT. We address this shortcoming by proposing three categories based on the level of data integration: Human Digital Models, Human Digital Shadows, and Human Digital Twins. Finally, we synthesise our findings in a domain-agnostic general definition for HDT. We highlight an edge case where the twinned entity is a human individual alongside a strongly coupled technical system, and name it augmented Human Digital Twin (aHDT). The definition and categorisation scheme provide the needed conceptual clarity for inter-disciplinary collaboration to address open challenges. Notable challenges are sensing human data, reliable data transfers and modelling, especially behavioural modelling. Additional ethical issues concerning security, privacy and consent are central to successful HDT adoption. We call for cross-disciplinary efforts to establish a standardised framework and ethical guidelines to enable future developments.

The Internet of Things massive adoption in many industrial areas in addition to the requirement of modern services is posing huge challenges to the field of data mining. Moreover, the semantic interoperability of systems and enterprises requires to operate between many different formats such as ontologies, knowledge graphs, or relational databases, as well as different contexts such as static, dynamic, or real time. Consequently, supporting this semantic interoperability requires a wide range of knowledge discovery methods with different capabilities that answer to the context of distributed architectures (DA). However, to the best of our knowledge there is no general review in recent time about the state of the art of Concept Analysis (CA) and multi-relational data mining (MRDM) methods regarding knowledge discovery in DA considering semantic interoperability. In this work, a systematic literature review on CA and MRDM is conducted, providing a discussion on the characteristics they have according to the papers reviewed, supported by a clusterization technique based on association rules. Moreover, the review allowed the identification of three research gaps toward a more scalable set of methods in the context of DA and heterogeneous sources.

The transition from automated processes to mechanisms that manifest intelligence through cognitive abilities such as memorisation, adaptability and decision-making in uncertain contexts, has marked a turning point in the field of industrial systems, particularly in the development of cyber–physical systems and digital twins. This evolution, supported by advances in cognitive science and artificial intelligence, has opened the way to a new era in which systems are able to adapt and evolve autonomously, while offering more intuitive interaction with human users. This article proposes a systematic literature review to gather and analyse current research on Cognitive Cyber–Physical Systems (CCPS), Cognitive Digital Twins (CDT), and cognitive interoperability, which are pivotal in a contemporary Cyber–Physical Enterprise (CPE). From this review, we first seek to understand how cognitive capabilities that are traditionally considered as human traits have been defined and modelled in cyber–physical systems and digital twins in the context of Industry 4.0/5.0, and what cognitive functions they implement. We explore their theoretical foundations, in particular in relation to cognitive psychology and humanities definitions and theories. Then we analyse how interoperability between cognitive systems has been considered, leading to cognitive interoperability, and we highlight the role of knowledge representation and reasoning.

As manufacturers are adopting data-driven decisions and processes, manufacturing is becoming more vulnerable to digital threats, making data security a major challenge for Industry 4.0. More specifically, data manipulation is considered a serious threat to organizations with significant and damaging consequences. Adopting a cybersecurity framework is essential to protect organizations against these cyber threats. These frameworks are often generic, unopinionated, and only provide high-level guidance for mitigating cyber risk. Our objective is to find a framework that offers a solid foundation we can customize to support our needs for addressing data manipulation risk. This paper aims to review cybersecurity frameworks and identify the most customizable, yet opinionated, option.