The paradigm shift introduced by Industry 5.0 has intensified the focus on enhancing worker autonomy through new technologies within decision-making contexts. Despite the growing body of research on this phenomenon, there is a notable lack of guidance on the appropriate methods for investigating this issue. This literature review aims to identify and characterize existing research and evaluate the most commonly used methods. The findings reveal various methods, each with significant limitations, and uneven coverage of new technologies studied to enhance decision-making processes and autonomy at work. Case studies emerge as the predominant method, but their inherent characteristics and opportunistic nature present significant limits to the establishment of structured research, enabling a comprehensive and generalizable study of such issues. However, the adoption of use cases is now emerging as a promising alternative to meet these challenges, but the development of a structured and coherent research framework is still needed.
The implementation of Digital Twin in healthcare significantly contributes to the diagnosis of pathologies, the suggestion of treatment scenarios, and the simulation of surgical interventions. Our focus is specifically on the application of Digital Twins in ophthalmology. The integration of artificial intelligence in the processes of prediction, monitoring, and simulation of pathologies brings numerous benefits, but also presents ethical risks and challenges. These risks include patient data confidentiality, accuracy of simulations and compliance with professional ethical standards. This paper emphasizes the necessity of in-depth discussions regarding the implementation of ethical standards to ensure the responsible use of artificial intelligence in medicine.
This paper describes an agent-based approach addressing the unique challenges of mobile multi-robot order picking in the logistics sector. Given the significant growth in the courier, express, and parcel service market, there's an increasing need for systems that not only cope with a rising demand but also adapt to the dynamic logistics environment. This study introduces enhancements to an Agent-Oriented Software Engineering methodology, focussing on information-related aspects that are of increasing importance in the era of Industry 4.0. By systematically developing a Multi-agent System that employs a decentralized control strategy, this research offers a scalable solution to optimize multi-robot tasks in warehousing and order fulfillment processes.
In the last 60 years, Programmable Logic Controllers (PLCs) have become central to manufacturing systems, driving real-time control, data acquisition, safety management, and machine communication. Their reliability and deterministic behaviour contributed to their ongoing importance, even as Industry 4.0 evolves. Simplified programming languages such as GCODE, IEC 61131, and IEC 61499 were developed to make complex behaviour programming accessible to non-experts. Technology providers also created proprietary languages or adapted standards to match hardware advancements. These languages balanced expressiveness and lightweight execution to meet the limited computational power of 1970s technology. However, as intelligent industrial systems increasingly integrate concepts like collaboration, cognition, and self-management, traditional automation languages showing their limitations. Enhanced computational power is shifting the focus from lightweight languages to more expressive ones. This paper highlights the urgent need for new automation programming languages that align with the sophisticated requirements of modern intelligent manufacturing systems. Such languages must harness today's computational advancements to ensure future systems are robust, efficient, and adaptable to the complexities of contemporary industrial environments. Based on these statements, a research agenda is proposed, composed of three main steps: efficient machines-agents design, cognitive work analysis and definition of enriched languages.
Large Language Models (LLMs) are rapidly evolving and increasingly utilized in a wide range of applications. They offer significant advantages over traditional Natural Language Processing approaches for analyzing requirements written in natural language. When designing intelligent systems according to Industry 4.0 principles, these requirements are often written by stakeholders from different domains, possibly leading to defects and misunderstandings. In this paper, we explore the potential of LLMs for detecting various defects, such as ambiguities, inconsistencies and incompleteness in requirements. We present an approach for improving current requirements engineering processes by defining a conversational interface specifically designed for requirements engineering based on LLMs. This interface allows stakeholders from different domains to interact with requirements without the need for extensive knowledge about the underlying models used by Artificial Intelligence, i.e., mainly the LLMs. This will ultimately provide a bridge between stakeholders and developers to create a common understanding of the requirements. The evaluation results of this approach show that it is possible to detect defects in requirements with a promising level of accuracy and respond with valuable suggestions for improving the quality of requirements.
Generative Artificial Intelligence (GAI) presents significant potential for managing Industry 4.0 companies' performance, driving efficiency, reactivity and innovation and enhancing thus the performance. However, despite its advantages, the use of GAI also has several limitations, notably regarding the focus on the short-term techno-centric vision of performance and the associated ethical issues. Moreover, ethical issues that may concern numerous aspects such as data privacy, human dependence, motivation or loss of autonomy can be a threat to the overall vision of performance, particularly in the long term, as the other elements that underlie it are neglected, namely the social and the environmental ones. Aware of the importance and the benefits of using GAI, companies are looking for operational solutions that enable them to ensure long-term performance while avoiding ethical risks. Therefore, the idea put forward in this paper subscribes to the assumption that ethics in use of advanced technologies is a necessary condition for performance. Then, in keeping with performance, the use of GAI must be coupled with management of the associated ethical risks. In this sense, an exploratory analysis is proposed, from data management situations experienced by a bearing manufacturer, partner in this study. Digitalizing its manufacturing for a few years, the Company is facing cases where considering ethics becomes necessary. Hence, associated ethical risks are presented as well as their potential impact on the (efficiency-effectiveness-relevance) performance conditions, in the short, medium and long term, leading thus to the discussion of deontological rules to put in place when using GAI in performance management.
The advances brought by Industry 4.0 contribute to the digital transformation and green transition, accelerating advances toward a circular economy. In this context, some interconnected concepts share similarities and complementarities, that can be analyzed to understand the potential benefits of their associated use. This paper discusses the relationship between the Digital Product Passport (DPP) and the Digital Twin concepts, highlighting their alignment and exploring their similarities and differences, aiming to understand how they complement each other. Therefore, the typology of assets, data model, and functionalities were analyzed, as well as an architectural and component alignment based on the elements provided by ISO 23247 while considering that the DPP concept is still in development. A case study is also presented to illustrate the implementation of both concepts.
The paper discusses the problem of adaptive management of mobile resources using couriers as an example. To solve the problem, an advancement is proposed from traditional combinatorial and heuristic algorithms to multi-agent models and methods that adaptively solve conflicts within schedules. In the proposed approach, solving the scheduling problem is self-organized through the cooperation and competition of software agents of the system representing orders and resources on the virtual market. This process stops when the state of "competitive equilibrium" (consensus) is achieved, i.e., when none of the agents can further improve the results. The proposed multi-agent method is used to solve the problem of managing mobile resources. At the same time, it is suggested to customize the system for the customers' business specifics using ontologies and ontological models of enterprises. The main result of applying the developed models and methods is an increase in the efficiency of mobile resources via adaptive planning and fast return of investments. The results obtained and prospects for further development are discussed.
As the embodiment of the digital twin for Industry 4.0, the Asset Administration Shell (AAS) facilitates standardized data exchange throughout the entire production lifecycle. However, despite the hype surrounding the AAS, its industrial application is still in its infancy. Especially for small and medium-sized enterprises (SMEs) the variety of implementation possibilities, use cases, and the prevailing IT infrastructure can pose challenges. Successfully navigating their digital transformation in this landscape requires methodical approaches, digital collaboration, and leveraging best practices so companies can carefully manage the transition from hype to tangible results, considering their company characteristics and specific requirements. With a focus on data management systems (DMS), this research shows where the AAS finds its use cases in the context of the prevailing IT infrastructure of (SME) companies, how it can be integrated into the existing and running systems, and finally how companies can gain added value from the application of the AAS as an enabler of the associated digitization, using the example of a medium-sized company from the mechanical engineering industry.
The application of digital twins (DTs) and artificial intelligence (AI) in public transportation has significantly improved traffic management and efficiency. Techniques such as agent-based modelling, reinforcement learning, and multi-agent systems have been used to dynamically adjust traffic signals and reroute vehicles, reducing congestion and improving traffic flow. Additionally, DT-centric approaches for driver intention prediction and adaptive multi-agent networks have shown potential in managing large-scale IoT systems. This study investigates the integration of DT standards and advanced AI methods, such as multi-agent systems and predictive models, to enhance the decision-making processes in the TransMilenio transportation system. The findings demonstrate that the model proposed can reduce the waiting time of passengers within the system.
Companies are finding collaborative robots (cobots) increasingly appealing due to the latest developments in the field. Unfortunately, a lot of human operators are untrained in operating these sophisticated devices. Observations conducted on the shop floor revealed that humans usually communicate through task-based interactions (TBIs) instead than following precise instructions that robots need to follow. In order to tackle this issue, we suggest a TBI architecture that allows human operators to assign complex tasks to machines through the use of triplets in task specification. This approach uses a task breakdown and a reasoning mechanism based on the Soar cognitive architecture.
Today's companies are complex, interconnected entities that coordinate resources and processes across multiple decision-making levels. This coordination, traditionally led by humans, spans various scales-from workstations to entire production sites-adding layers of complexity. To address this, the paper introduces a framework for designing a decision-support system aligned with human cognitive functioning, considering multi-level and multi-scale settings. The framework aims to manage the company's multilayered characteristics while ensuring coherence in decision-making. Specifically, it seeks to: (1) support adaptable organizational decision-making; (2) provide insights into the impact of decisions at different levels; and (3) explore the synergy between human expertise and automation. The paper concludes with a case study demonstrating the framework's importance and proposes a solution for future validation.
The integration of data capture, analysis, monitoring, and control technologies is rapidly becoming the cornerstone of next-generation smart buildings. However, developing digital twins that dynamically interact with these buildings presents a significant challenge. In this paper, we study the most appropriate data models for leveraging a digital twin from data-driven smart buildings. We propose a framework that exploits a knowledge graph to directly address the challenges encountered in real-world building management systems, ensuring that the information is comprehensible as a preliminary step to intelligent decision-making. Furthermore, we validate this proposal for improving building performance and sustainability through a real-world use case. The experimental results, utilizing dynamic data streams from the Internet of Things (IoT), demonstrate promising outcomes. This research paves the way for using graph-based models and algorithms as digital twin enhancers for managing data-driven smart buildings.
In recent years, the topic of data spaces has been increasingly gaining traction. Multiple research initiatives, backed by industrial and government actors have introduced several data spaces for various industries and medicine, and beyond. In order to support value creation, public services and collaboration between a plethora of different participants of the data spaces, the sharing of information and building services and insights thereupon has been identified as key enabler. However, these initiatives differ in various properties from the initial idea for data spaces derived in research. In this article, we review the current state of literature regarding the concept of data spaces and the impact of European initiatives, namely International Data Spaces, Gaia-X and European Health Data Space, on the literature. Identifying characteristic properties of data spaces and its participants, we provide a general overview.
In response to the environmental repercussions of the linear economy model, the circular economy has emerged, aiming to reduce the environmental impact while balancing economic and social aspects. Industrial ecology is a pillar of circular economy, facilitating collaboration among local industries through industrial symbiosis. Industrial symbiosis involves exchanges and pooling of resources and services between industries. Despite their pivotal role, industrial ecology and symbiosis implementation pose challenges, prompting increased scholar attention. This study conducts a literature review and a taxonomy proposal for industrial symbiosis. Through an examination of recent surveys, critical gaps in current research are uncovered, and key criteria for evaluating industrial symbiosis practices are identified. The findings highlight the need for more holistic approaches that integrate economic, social, and environmental dimensions, focus on operational decision-making, and address uncertainties. This review aims to guide future research and enhance the practical application of industrial symbiosis.
The new industrial revolution has introduced digital twins as a cuttingedge technology that promises huge potential for bringing added value in smart industries. However, a generic and adaptable digital twin architecture is critically needed to meet the evolution requirements. A number of existing studies have proposed different architectures for designing digital twins. However, significant challenges remain in conceiving and deploying a scalable and reliable digital twin architecture in industrial systems. Considering this, how can we move towards an architecture that enables (1) real-time data collection and exchange between the physical world and the digital world, (2) real-time monitoring, (3) advanced analysis, (4) dynamic planning and adaptability, (5) automated execution, and (6) supporting human in decision-making processes? The main objective of this study is to develop a generic-to the extent possible- and adaptable digital twin architecture, designed using the MAPE-K framework combined with the holonic paradigm. This proposition is illustrated with two case studies from two different fields, healthcare and manufacturing, in order to demonstrate its capabilities.
In the healthcare sector, the use of simulation models is pervasive and multifaceted, serving a range of purposes, including but not limited to supporting decision-making, resource allocation, and reducing overcrowding of facilities. These models may be connected to the physical world in various ways, which allows them to be integrated into other systems, such as the Digital Twin (DT). Nevertheless, in the case of DTs, the simulation model needs to be updated quickly to ensure accuracy, a task that is not straightforward when a specific simulation model with a high level of detail is required. To deal with this stake reusable models with twinning features could be a solution. As a result, a review is conducted to study existing simulation models in the healthcare sector and to assess their twinning process as well as their reusability and abstraction level. The main conclusions of this review, including 26 papers are the following. Firstly, most of the models are specific and therefore difficult to reuse. Secondly, generalisable models are not being used to develop specific models. Thirdly, the updating phase of simulation models, being part of a DT, is never mentioned.
Underground mines are arduous work environments, where worker safety needs to be improved through the implementation of new mining technology. Better integration of workers with technology and the complex systems in an underground mine can help improving the safety and productivity of workers. To facilitate this integration, the paper presents a holonic architecture for a Human Cyber-Physical System (HCPS) to improve worker safety in underground mines. The HCPS aims to monitor and inform decisions concerning each worker's health and safety, while enabling the coordination of workers in the execution of mining operations. This paper applies a systems engineering approach to develop the architecture requirements and develops the architecture using the Activity-Resource-Type-Instance (ARTI) reference architecture and holonic design principles. An illustrative example is presented to show how the developed HCPS would perform the monitoring and coordination of workers in an underground mine.
The requirements for manufacturing systems that can manufacture individualized products are demanding. In addition to a large throughput while simultaneously supporting unique manufacturing processes for each product, high product quality, compactness, energy efficiency, expandability, and cyber resilience are essential. This paper examines two demonstrators that utilize an industrial PC-based control with incorporated Machine Learning (ML) and a flexible product transport system at the centre of the corresponding manufacturing system. The use of a multi-core processor demonstrates efficient exploitation of synergies, potentially extending the manufacturing system’s lifecycle. Both demonstrators incorporate neural network models to control the flow of individualized products. The inference time for image classification is less than 500 µs with a classification accuracy of 100
The demands of innovative production systems are shifting from mass production to the creation of smaller quantities with a focus on high quality. To achieve these evolving demands, Zero Defect Manufacturing has emerged as a key paradigm. This approach requires an innovative architectural monitoring tool where real-time data is continuously gathered and analysed to predict defects and assess their potential impacts. It also necessitates the seamless integration of diverse data sources, advanced processing algorithms, and Digital Twins to align with industrial requirements. In this paper we present a real-time, rule-based monitoring tool applied to a real-world car manufacturing use case. The tool successfully generated early alerts for quality deviations, enabling production engineers to shift from a reactive to a proactive approach by detecting potential quality issues early in the process.
