Exploring the Role of Digital Twin and Industrial Metaverse Technologies in Enhancing Occupational Health and Safety in Manufacturing

Abstract

The evolution of Industry 4.0 and the emerging paradigm of Industry 5.0 have introduced disruptive technologies that are reshaping modern manufacturing environments. Among these, Digital Twin (DT) and Industrial Metaverse (IM) technologies are increasingly recognized for their potential to enhance Occupational Health and Safety (OHS). However, a comprehensive understanding of how these technologies integrate to support OHS in manufacturing remains limited. This study systematically explores the transformative role of DT and IM in creating immersive, intelligent, and human-centric safety ecosystems. Following the PRISMA guidelines, a Systematic Literature Review (SLR) of 75 peer-reviewed studies from the SCOPUS and Web of Science databases was conducted. The review identifies key enabling technologies such as Virtual Reality (VR), Augmented Reality (AR), Extended Reality (XR), Internet of Things (IoT), Artificial Intelligence (AI), Cyber-Physical Systems (CPS), and Collaborative Robots (COBOTS), and highlights their applications in real-time monitoring, immersive safety training, and predictive hazard mitigation. A conceptual framework is proposed, illustrating a synergistic digital ecosystem that integrates predictive analytics, real-time monitoring, and immersive training to enhance the OHS. The findings highlight both the transformative benefits and the key adoption challenges of these technologies, including technical complexities, data security, privacy, ethical concerns, and organizational resistance. This study provides a foundational framework for future research and practical implementation in Industry 5.0.

1. Introduction

1.1. Background

The ongoing fourth industrial revolution (Industry 4.0) is being propelled by rapid advancements in digitalization and information technology [1,2,3]. Digital technologies such as the Internet of Things (IoT), Big Data, Cloud Computing, Artificial Intelligence (AI), Machine Learning (ML), Cyber-Physical Systems (CPS), Virtual Reality (VR), Augmented Reality (AR), and robotics are crucial aspects of Industry 4.0, which are transforming the manufacturing environment [2,4,5,6,7,8]. These technologies are greatly transforming global socio-cultural and economic environments by interacting across physical, digital, and biological domains, impacting various industries, including manufacturing [1]. Industry 4.0 leverages digital technologies to create smart, connected, and automated systems, optimizing operations, cutting costs, and enhancing sustainability and product quality through real-time insights [3,6,9,10]. The core principle of the Fourth Industrial Revolution is that robots should adapt to humans rather than vice versa [11].

On the other hand, the fifth industrial revolution (Industry 5.0) builds upon Industry 4.0 by emphasizing human-centricity, sustainability, and resilience, and seeks to ensure worker well-being while promoting effective human-machine collaboration, that is, shifting from technology-driven Industry 4.0 to value-oriented Industry 5.0 [4,6,9,12,13,14,15,16,17,18]. Industry 5.0 marks a shift toward human-centered manufacturing that leverages technologies like Extended Reality (XR), Collaborative Robots (COBOTS), and wearable sensors to boost flexibility, safety, and interactive human-machine collaboration [12,19,20,21]. For instance, wearable sensors are used to monitor worker well-being and adapt activity schedules accordingly in real time [22].

Ensuring worker safety, especially in high-risk industries, including manufacturing, is vital for both well-being and workplace efficiency and is legally mandated by the Occupational Health and Safety (OHS) Act of 1970 [23]. OHS remains a persistent concern in industrial sectors, especially manufacturing, where complex environments and high-risk operations increase the potential for accidents and long-term health conditions. Manufacturing environments are inherently hazardous, exposing workers to physical strain, repetitive motion, poor posture, chemical risks, and high-stress conditions, all of which contribute to accidents, injuries, and illnesses, such as work-related musculoskeletal disorders (WMSDs) [7,13,24,25,26,27,28]. Exposure to ergonomic risk factors, such as noise, vibrations, physical and mental strain, and poor posture, can compromise worker safety and lead to WMSDs [13].

Globally, work-related accidents, including fatal and non-fatal injuries, affect millions of workers annually [29]. For instance, in 2018, WMSDs in the manufacturing sector accounted for 14.16% of reported cases across all private sectors [27], and in 2020, manufacturing had the highest rate of non-fatal accidents (18.2%) and the second highest rate of fatal accidents (14.6%) in the EU [17]. WMSDs are prevalent in the manufacturing sector, especially in European countries, where machine operators and assembly workers experience some of the highest rates, that is, 66% [25]. These health risks not only cause work disability [26] but also result in higher absenteeism and presenteeism, reduced motivation, increased employee turnover, and significantly negatively impact productivity and costs [30].

Additionally, workplace accidents can arise from various manufacturing processes [7]. For example, chemical exposure, overheated components, and faulty equipment interactions contribute to workplace accidents and production losses [7,31,32]. Chemical-related accidents account for two-thirds of manufacturing injuries in the United Kingdom, costing the manufacturing industry over £627 million annually [32]. Furthermore, Industry 4.0 introduced advanced digital technologies, but its implementation has primarily focused on productivity and automation, sometimes overlooking the human aspect. Additionally, traditional safety systems typically adopt a reactive approach, addressing incidents only after they occur, rather than proactively identifying and mitigating potential risks.

Hence, in the manufacturing context, it is crucial to prioritize workers’ well-being and longevity to ensure and sustain manufacturing for future generations, industrial revolutions, productivity, and business continuity [20]. OHS is vital in manufacturing, particularly for small and medium-sized enterprises (SMEs) [33], since a safer, healthier workplace supports well-being and boosts worker productivity [34]. Safety 4.0 is a concept that applies Industry 4.0 technologies to proactively prevent accidents, enhance safety and productivity, and reduce work-related stress and mental health risks [7,27,35]. Digital solutions can effectively address hazardous situations and enhance safety performance [33]. For instance, Operator 5.0 is an emerging concept that envisions a smart operator using advanced interfaces and data systems to ensure sustainable manufacturing operations and workforce well-being [13,16]. Moreover, businesses are increasingly recognizing OHS as a strategic priority essential for enhancing productivity, reducing costs, and achieving overall financial success [26,36]. Effective employee health management is cardinal for enterprise development and sustained growth [37]. Thus, safety is a top priority, and companies are digitally transforming it through both technological and managerial approaches [38].

Despite advances in immersive and intelligent technologies, workplace accidents, ergonomic injuries, and mental stress remain prevalent in the manufacturing industry. Many existing safety protocols and training methods are often passive and outdated. The emergence of the Digital Twin (DT) and Industrial Metaverse (IM) enables the transition from reactive to predictive safety management. These technologies offer real-time insights, immersive training environments, and intelligent decision support, creating opportunities for enhanced ergonomics, reduced risks, and improved workforce engagement. This study is motivated by the urgent need to explore how the combination of Digital Twin (DT) and Industrial Metaverse (IM) technologies can actively address these ongoing safety issues. This research aims to bridge the gap between traditional safety approaches and the emerging paradigm of human-centric, predictive safety approaches enabled by advanced digital ecosystems. Hence, this study is motivated by the pressing need to bridge the current gap in OHS practices using these transformative digital capabilities.

1.2. Study Objectives

Maintaining the health, safety, and well-being of workers is vital for achieving social sustainability in manufacturing [39], and safety is a top priority in the industrial sector [40,41]. Effective safety training is essential for minimizing and preventing accidents during human-machine interactions [14]. Given that humans are central to any business, effectively managing their resources and capacity, particularly their health, safety, and well-being, has become increasingly crucial in recent decades [26]. Moreover, safety training is legally mandated in many countries and industries as a vital accident-prevention measure [42].

While safety training is crucial for human-machine interactions, current methods in manufacturing fall short by not adapting to individual learning needs, knowledge levels, and operational habits, and often rely on passive and unengaging methods like text, images, videos, and regulations [14,42]. Traditional training methods are often ineffective due to low engagement, poor knowledge retention, high costs, and limited ability to simulate real hazards [43]. Furthermore, despite technological advancements, the consistently high rates of WMSD in manufacturing highlight the urgency of better assessment methods and regular updates to safety protocols [27].

Innovative emerging tools like immersive VR and AR platforms have demonstrated promise in industrial education by combining gamified experiences with simulations to improve knowledge retention and emotional preparedness. For example, manufacturing and other industrial sectors have become prime areas for applying immersive learning environments via virtual technologies, which combine gamification with educational design to improve safety training [44]. Immersive visual technologies create exciting opportunities to enhance the effectiveness of safety training programs [45]. These emerging immersive technologies are central to the shift towards human-centered smart manufacturing, which entails safe and reliable human-robot collaboration (HRC) [46].

Hence, this study aims to explore the integration of DT and IM technologies to enhance OHS in the manufacturing sector. These technologies hold significant promise by enabling real-time monitoring, analysis, and optimization of working conditions through immersive digital simulations. While previous studies have primarily examined their applications in industries such as construction and healthcare, there is limited research that comprehensively addresses their combined integration in manufacturing for OHS purposes, highlighting the novelty and necessity of this study. Thus, the main objectives of this study are to investigate the definitions, characteristics, and applications of DT and IM technology in manufacturing. Furthermore, we analyze how these technologies contribute to improving OHS through immersive training and real-time risk management. In addition, we identify the key enabling technologies, benefits, and barriers to implementation. Based on these objectives, we posed the following research questions (RQs):

• RQ 1: What is a DT, and how is it used to create virtual replicas of plants, machinery, and production processes?
• RQ 2: What is an IM, and how can virtual worlds be used to simulate real industrial environments?
• RQ 3: How does the integration of DT and IM technologies create an advanced digital ecosystem that enhances safety training, incident management, and workplace monitoring?
• RQ 4: What are the benefits of using immersive technologies in safety training and incident management?
• RQ 5: What are the challenges and limitations of the large-scale adoption of these technologies in manufacturing?

This study supports the shift toward Industry 5.0 by demonstrating how immersive and predictive technologies can transform OHS in manufacturing. By addressing these key RQs, a roadmap for moving from reactive safety measures to proactive, human-centered strategies is offered. This study highlights the unique value of integrating DT and IM technologies to create safer and smarter work environments and support the vision of Industry 5.0 in the manufacturing sector. Among the research questions, RQ3, which focuses on integrating DT and IM technologies to build a proactive safety ecosystem, represents the core contribution of this study. This integration is underexplored in the current literature and reflects the research value of our study, which lies in proposing a unified, immersive, and human-centered safety framework for the manufacturing sector.

2. Methods

2.1. Research Strategy

This study adopted a systematic literature review (SLR) approach to explore the roles of DT, IM, and other interconnected advanced digital technologies in enhancing OHS in manufacturing environments. The SLR methodology is characterized by a rigorous, structured, and transparent process for identifying and evaluating relevant academic literature [47], following a replicable and scientific methodology [48]. Additionally, this review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, which outline a detailed, step-by-step framework for literature identification, screening, and inclusion.

2.2. Data Collection

The research literature was sourced from SCOPUS and Web of Science databases, which are among the largest repositories of peer-reviewed abstracts and citations, ensuring high-quality and reliable material. While several databases index literature on digital and immersive technologies, SCOPUS and Web of Science were chosen for their broad interdisciplinary coverage, strong indexing in engineering, health, safety, and digital technologies, and their comprehensive inclusion of peer-reviewed documents.

The initial search strategy targeted the Title, Abstract, and Keyword fields using a combination of terms related to DT, IM technologies, Industry 4.0/5.0, OHS concepts, and industrial scope, connected by Boolean operators “AND” and “OR”. Figure 1 shows the search string employed in both the SCOPUS and Web of Science databases. The initial search generated 219 research documents from both databases, published between 2000 and 2025. The research items were then screened.

2.3. Screening Process

The literature selection followed a three-stage filtering approach aligned with PRISMA: identification, screening, and eligibility phases. In the identification phase, the initial search yielded 219 documents. At this stage, non-English publications, duplicate studies, publications before 2010, and conference proceedings and reviews were excluded, resulting in the removal of 82 documents. The remaining 137 studies, including articles, reviews, conference papers, and book chapters, were retained for further screening, as shown in Figure 2.

During the screening phase, the titles, abstracts, and conclusions were reviewed to assess the relevance of each study to the research focus. This step aimed to identify literature situated within the context of emerging advanced digital and immersive technologies, including DT and IM to enhance worker safety in the manufacturing sector. As a result, 43 studies that were not focused on manufacturing or relevant technologies were excluded, leaving 94 studies for full-text evaluation.

In the second step of the screening phase, that is, the eligibility process, the remaining studies were thoroughly reviewed to determine their alignment with the scope of this study. Research documents that addressed interconnected immersive technologies aimed at improving OHS were included. However, studies focusing solely on policies and regulations, algorithm optimization, and not addressing OHS were excluded. This process resulted in 75 articles being selected, ranging from 2016 to 2025, and eligible for this study.

To ensure the relevance of the findings, it was necessary to establish eligibility criteria definitions, as automatic search results could yield research items that are not directly related to the research questions or database search criteria. Thus, it was deemed important to set both exclusion and inclusion criteria. The eligibility criteria for including the studies are summarized below:

• Published in peer-reviewed international journals, conferences, or book chapters.
• Focused on the use of immersive or digital technologies (e.g., DTs, IM, XR, VR, AR, and CPS) within the manufacturing sector.
• Explicitly addressing the implications for OHS in industrial settings.

Similarly, the studies that were excluded were based on the following exclusion criteria:

• Focused solely on technical algorithm development or optimization.
• Discussed regulatory policies and regulations.
• Pertained to unrelated sectors such as healthcare, construction, and education.
• Duplicate documents.
• Conference proceedings.

3. Descriptive and Bibliometric Analysis of Selected Studies

3.1. Publication Trends over Time

The selected research documents (75) in this study, published annually between 2016 and 2025, are shown in Figure 3. The figure shows an increasing research trend related to advanced digital technologies for improving OHS in Industry 4.0/5.0 settings. From 2016 to 2019, publication activity remained relatively low for the research theme of this study. However, since 2020, there has been an increase in relevant research publications that will continue into 2025. In particular, from 2023, a notably significant upward trend can be observed. This growth may be attributed to the rise in emerging advanced digital technologies such as VR, AR, and DT, and the transformation into a human-centered Industry 5.0. Additional factors contributing to this growth could be technological maturity, affordability, and easier availability of immersive technologies, as well as post-pandemic digitization of safety training systems.

3.2. Document Types

The analysis of the selected documents by type revealed the following distribution: journal articles (59%), conference papers (38%), and book chapters (3%), as shown in Figure 4. This mixed distribution is important to the aim of this study, as it reflects both the academic strength of the field and the evolving nature of the research landscape. For example, the high proportion of peer-reviewed journal articles underscores the scientific credibility and growing maturity of certain areas of the application of DTs and immersive technologies in OHS. At the same time, the large share of conference papers indicates a fast-moving and still-evolving field, where many innovations are in the early stages.

This distribution is important for several reasons. First, it highlights the need to bring together existing knowledge to support future research and practical use, which is the main goal of this study and the transition to Industry 5.0.

First, it supports the need for this review by showing that the field is still growing and is somewhat fragmented. Second, it justifies the use of an SLR, as such an approach is especially valuable in rapidly developing areas and brings together scattered findings. Finally, the limited consolidation of knowledge across sources highlights existing research gaps and confirms the novelty and necessity of this review.

3.3. Subject Distribution and Citations

Figure 5 shows the subject area distribution of research on DT, IM, and other interconnected technologies in creating an immersive interactive environment in the manufacturing sector, highlighting their multidisciplinary nature. The most prominent fields include Engineering (30.1%), Computer Science (19.7%), Other Disciplines (10.6%), Social Sciences (6.2%), and Chemical Engineering (6.5%). Additionally, contributions from Mathematics, Physics, Medicine, Business Management and Accounting, Environmental Science, and Decision Sciences are also important, reflecting the broad relevance of leveraging digital transformation research across domains.

Citations from other studies enhance the impact and credibility of the published work. Of the 75 research documents, there are (28) research studies with at least 10 citations, followed by (20) research studies with citations between 10 and 50, (4) studies with citations between 50 and 70, (2) studies with more than 100 citations, and (1) study with citations greater than 300.

Table 1. Top 10 publication sources.

Sr. No	Research Study	Year	Citations	Source
1	[36]	2018	307	Safety Science
2	[39]	2020	128	Sustainability (Switzerland)
3	[49]	2022	66	Computers and Industrial Engineering
4	[50]	2018	59	IFIP Advances in Information and Communication Technology
5	[51]	2022	56	Safety Science
6	[52]	2018	52	2018 IEEE 16th Student Conference on Research and Development, SCOReD 2018
7	[3]	2023	48	Process Safety and Environmental Protection
8	[1]	2021	40	International Journal of Computer Integrated Manufacturing
9	[5]	2023	37	Administrative Sciences
10	[8]	2023	35	Heliyon

shows the top 10 most cited publications, covering themes like human-centric Industry 5.0, safety, and immersive digital systems. These studies highlight the role of DTs, IM, and relevant digital technologies, which can aid in creating an advanced digital ecosystem to enhance OHS in the manufacturing sector.

3.4. Co-Occurrence of Keywords

A keyword co-occurrence analysis was conducted using VOSviewer software v1.6.20.0 on the final dataset of 75 studies. This analysis was performed to reveal how certain keywords frequently appear together across different documents, highlighting their thematic connection and linking related research items. In the VOSviewer software, this analysis was conducted using “All Keywords” as the unit of analysis and “Full Counting” as the counting method, which assigns equal weight to each keyword co-occurrence. This approach was chosen over fractional counting for co-authorship and bibliographic coupling analyses. From a total of 743 keywords, a keyword map was generated using a threshold of at least five occurrences per keyword, resulting in a visualization of 33 keywords. Figure 6 presents this keyword co-occurrence map, where “Industry 4.0” appeared most frequently (25 times), followed by “accident prevention” (22 times), “virtual reality” (19 times), “occupational risks” (15 times), and “worker safety (12 times)”. Moreover, the results (Figure 6) reveal four dominant research clusters:

• Cluster 1 (Red): This cluster is comprised of 14 keywords and focuses on topics related to Industry 4.0 technologies, OHS, and accident prevention. It emphasizes Industry 4.0 technologies such as IoT, AI, and CPS, and how they can be utilized to prevent accidents, enhance OHS, and workers’ wellbeing.
• Cluster 2 (Green): With nine keywords, this cluster centers on Industry 5.0, smart manufacturing, human-centricity, and worker safety. It highlights the shift to Industry 5.0 that marks a change toward human-centered smart manufacturing that utilizes technologies like DT, AR, COBOTS, or HRC, with a focus on worker safety.
• Cluster 3 (Blue): This cluster includes seven keywords related to safety and personal training of the workers by using virtual environments and VR technologies.
• Cluster 4 (Yellow): Containing only three keywords, human, ergonomics, and occupational health, this cluster solely reflects the importance of workers’ health that can be monitored in real-time using ergonomic analysis via DT, IM technologies towards a smart, human-centric, safe, and sustainable manufacturing ecosystem.

The presence of these research themes highlights the growing focus on immersive and real-time technologies for proactive safety solutions in manufacturing. Future research should delve deeper into these areas to uncover more impactful insights.

3.5. Bibliographic Coupling (Documents)

The findings of this analysis provide valuable insights into emerging trends within the field and help outline potential directions for future research. A bibliographic coupling map was generated using VOSviewer based on shared references among the research documents included in the study. It analyzes the thematic structure of the selected research documents and how they are grouped based on what they cite. By setting a minimum citation threshold of three, 39 out of 75 research items met the criteria and were included in the visualization shown in Figure 7. These documents were grouped into eight distinct clusters, reflecting the diversity and interconnectedness of the research through citation relationships. Key metrics in bibliographic coupling include co-citation strength and total link strength, which indicate the influence and connectivity of individual documents.

Table 2. Research documents with the highest total link strength.

Sr. No	Research Document	Total Link Strength	Cluster
1	[8]	26	4 (Yellow)
2	[53]	19	4 (Yellow)
3	[40]	17	1 (Red)
4	[54]	16	1 (Red)
5	[36]	12	4 (Yellow)
6	[27]	11	2 (Green)

lists the documents with the highest total link strength based on bibliographic coupling. For instance, the study by [8] (Cluster 4—Yellow) has the highest total link strength (26), indicating a strong thematic alignment with numerous other studies in the dataset. This study explored the impact of Industry 4.0 technologies on OHS, but it has a modest citation count of 35. This highlights the role of [8] as a core integrative study, particularly in linking Industry 4.0 technologies to OHS frameworks. In contrast, the research study [36], also in Cluster-4 (Yellow), is the most highly cited document in the review (307), but its total link strength is only 12. This suggests that while it is broadly influential, it shares fewer references with the other documents, possibly due to its broader framing or foundational nature. This study examined whether Industry 4.0 technologies adequately address OHS challenges, highlighting the critical need for integrated OHS strategies to mitigate emerging risks.

Similarly, research item [40] in Cluster-1 (Red) has a total link strength of 17 and focuses on the effectiveness of immersive VR training in improving safety behaviors in confined space environments. Meanwhile, study [27] in Cluster-2 (Green) demonstrates a total link strength of 11 and investigates the use of computer vision, specifically for human and PPE detection, and Mixed Reality (MR)-based training. Both these studies represent emerging themes in Safety 4.0, with strong linkages to the evolving body of research. These findings show that studies with high link strength often occupy thematic centers within the ecosystem of DT and IM applied to OHS, offering practical alignment and relevance even if they have fewer citations. This supports the argument that thematic connectivity, not just citation count, should guide the identification of key works in emerging interdisciplinary fields like smart manufacturing safety.

4. Results and Discussions

4.1. Emerging Digital Technologies

Digital transformation in manufacturing is driven by integration of advanced and immersive technologies, aimed at increasing efficiency, safety, and human-machine collaboration. This section explores emerging technologies such as DTs, IM, immersive environments, and their enabling key technologies like IoT, AI, and CPS, that are revolutionizing the OHS in manufacturing, and answering RQ1, RQ2, and RQ3, respectively. As mentioned by [55], the essence of intelligent manufacturing lies in digitizing the physical world and connecting all systems. These technologies serve as the backbone of advanced digital ecosystems like smart factories, enabling capabilities like real-time monitoring, predictive analytics, ergonomic analysis, and immersive virtual safety training, fostering a smart, human-centric, safe, and sustainable manufacturing ecosystem.

4.1.1. Digital Twin

DT is a pivotal technology that enables the shift from traditional to smart business systems [56]. DTs are real-time digital replicas of physical entities, such as machines, production systems, or human workers, capable of continuously updating data from physical entities to the virtual world to attain optimized goals through data analytics, simulation, and predictive techniques [4,18,19,41,56]. DTs consist of three main components: a physical object, its digital counterpart, and the data link between them for continuous information exchange [9,46]. Moreover, unlike static digital models and one-way digital shadows, DTs provide a dynamic, two-way data exchange between the physical and virtual systems, allowing fully automated and real-time updates [13,57]. According to [58], DTs exist in diverse forms, such as physics-based, data-driven, hybrid, and those focused on maintenance or process optimization. Similarly, the study by [56] highlights the significance of DT technology in enhancing the competitiveness of the shipbuilding industry, particularly through the improvement of advanced outfitting, which reduces the danger of work injuries.

The development of dynamic systems and predictive DT platforms has been advanced through the application of real-time sensor data and AI algorithms. In DT systems, numerous IoT sensors facilitate real-time digital replication, which optimizes the workflow, enhances throughput, and ensures quality [59]. Similarly, a DT leverages the IoT, simulation, and data analytics to create a virtual replica that mirrors an asset’s data, processes, operational states, and lifecycle [3]. AI-powered, IoT-driven DTs optimize production by utilizing real-time sensor data to improve efficiency and decision making [3,6,60]. For example, in a 3D DT environment, physical object data are integrated into a digital model via sensors and cameras, enabling simulations to predict behavior and performance, with the resulting insights fed back to optimize real-world operations and decision-making [7,41].

DT technologies are foundational in modern manufacturing, offering a wide range of benefits and applications. They enable real-time monitoring, remote control, and predictive maintenance, while also supporting virtual risk assessments, improved team collaboration, faster decision-making, and better communication through enhanced documentation [61]. DTs are essential for virtualizing and digitizing key physical components, functions, and systems, providing deeper insights into manufacturing processes [61,62]. DTs seamlessly integrate virtual and physical manufacturing environments using real-time data (from sensors, machines, and other sources) to enable smart, adaptive, and sustainable processes by optimizing production, minimizing downtime, enhancing quality, and supporting predictive maintenance [61,63]. They are also critical for understanding system resilience and detecting potential risks [9]. For instance, DTs can enhance workplace safety and boost operational efficiency. As shown by [18], DTs can simulate workplace airflow and the concentration of airborne contaminants released during manufacturing, thereby enabling the optimization of hazard control systems such as ventilation. Moreover, VR serves as an enabling technology for DTs. According to [56], when combined with AR tools, DTs streamline the alignment of virtual and physical environments, cutting communication time between design and production teams.

In addition to asset monitoring, human DTs are rapidly gaining traction in manufacturing due to their ability to provide real-time monitoring of machines, products, and worker conditions, such as health, stress levels, and posture, within an integrated ecosystem [9,22]. DTs support system design by predicting performance and adapting to workers’ needs [59]. Unlike conventional training simulations, DTs support closed-loop feedback systems that continuously adapt to operator states and machine interactions, thus aligning with the human-centric design of Industry 5.0.

Compared to traditional manufacturing management tools, such as static CAD models and standard simulation platforms, DTs provide a more dynamic and interactive connection between the physical and virtual environments. While legacy systems typically offer reactive insights based on historical or static data, DTs enable predictive modeling, real-time risk assessment, and continuous feedback loops. This makes them particularly valuable for proactive safety and operational monitoring.

Figure 8 highlights the key insights related to DTs in terms of industrial applications, strategic benefits, enabling technologies, safety, and design. The influence of DT technology is categorized into four principal domains. First, technological enablers, which include IoT, AI, data analytics, simulation, and real-time data integration, form the foundation of DT systems. Secondly, industrial applications which encompass workflow optimization, predictive maintenance, quality assurance, system resilience, and adaptive manufacturing, demonstrating DT’s role in enhancing operational processes. Thirdly, safety and design benefits are realized through real-time monitoring, hazard simulation, ergonomic adaptation, and virtual risk assessment, contributing to safer and more efficient work environments. Lastly, strategic benefits highlight the impact of DTs on improving efficiency, decision-making, collaboration, sustainability, and responsiveness to dynamic operational needs. This diagram illustrates how DT acts as a pivotal transformative technology driven by advanced digital tools to deliver substantial operational, safety, and strategic advantages across various industries.

To conclude, DTs are not just digital copies but dynamic, self-updating systems that allow for real-time simulation, prediction, and intervention. They serve as a core platform for replicating real-time physical environments, enabling immersive simulations and predictive analyses for process monitoring, hazard identification, and ergonomics monitoring. This review found that DTs are central to enabling proactive safety management, validated across high link strength papers. DTs also serve as the backbone for integrating IoT and sensor data, making them essential for conducting predictive analytics. Their integration with IoT and AI lays the foundation for proactive, human-centric safety management in Industry 5.0. DTs have shifted OHS from static compliance-based models to continuous ergonomic optimization and incident forecasting. For example, [56] demonstrated how shipbuilding DTs reduce pre-construction hazards through airflow simulation and equipment layout optimization. However, the effective implementation of DTs depends on access to high-quality real-time data, specialized technical expertise, and seamless integration across platforms. These requirements can pose challenges, especially for smaller or less digitized organizations, such as SMEs.

4.1.2. Industrial Metaverse

The growing emphasis on human-centric values, immersive experiences, and economic flexibility is driving the development of innovative manufacturing models and the emergence of IM. IM is currently poorly defined, with gaps in research on its enablers [4]. Although still an emerging concept, it is described as a social manufacturing environment that integrates cyberspace, physical systems, and human interaction, enabling collaborative value creation among stakeholders [4,64,65]. It consists of three interconnected spaces: physical space (real-world objects, humans, and environments), cyberspace (virtual data and simulations), and social space (human-machine interactions). In the term metaverse, “meta” refers to the virtual world, and “verse” refers to the real world [4,64].

Although IM relies on various enabling technologies, its realization is incomplete without DTs [4]. IM creates immersive and collaborative environments that extend the functionality of DTs. While DTs provide precision simulations, IM enables experiential learning and collaboration, allowing users to practice coordinated safety responses and decision-making remotely. Moreover, recent studies have confirmed its growing relevance for enabling cyber-physical-social integration, cross-functional collaboration, and immersive training. For example, DTs are essential for enabling IM, where humans, COBOTS, and process management systems can interact seamlessly, both locally and across different locations [65]. Similarly, DTs are essential for CPS, which is a core element of the IM [4]. IM and DTs create immersive digital environments for visual interactions and operational purposes. More specifically, IM creates immersive digital environments with unique identities for interaction, while DTs focus on replicating the physical world for optimization, emphasizing technology over human engagement. IM basically represents the expansion of DTs into human and societal domains [64].

The IM architecture comprises five layers: perception (data collection), networking (data processing), fusion (integration of physical and cyber systems), interaction (3D visualization), and configuration (decision-making) [4]. Unlike VR or DTs alone, IM enables multi-user interactions within virtual environments. For example, engineers in different locations can enter the same immersive space to analyze equipment, simulate production lines, and conduct ergonomic assessments [65]. Moreover, IM combines technologies like blockchain and AI to create immersive AR experiences [64]. The integration of IM with blockchain and edge computing enables real-time updates of safety protocols and predictive task planning in smart factories. For example, an Extended Reality (XR) based HRC assembly system in the IM, integrating blockchain and edge-cloud networks, can enhance real-time operations through improved planning, task optimization, and strategic decision-making [6]. Edge computing enhances computing by improving efficiency, lowering latency, and bolstering data security through localized processing [37]. As shown in the literature, IM’s layered architecture of IM and XR interfaces supports enhanced decision-making and strategic planning in manufacturing safety systems.

The benefits of the industrial metaverse lie in interactivity, synergy, and economics. For example, integrating IM into manufacturing systems creates substantial value by reducing scrap and rework, minimizing downtime, lowering compliance costs, increasing throughput, and enhancing workforce training. More broadly, the implementation of IM technologies offers a range of benefits, including real-time multidisciplinary collaboration and training within immersive 3D virtual environments free from geographic or language constraints. Additional advantages include reduced training expenses, improved knowledge retention, real-time machine interaction and synchronization, enhanced spatial awareness, and greater system usability [64].

Figure 9 provides a concise overview of the enabling technologies, key applications, and core functions of IM. The core functions represent the foundational capabilities of IM, including the development of immersive and interactive virtual environments, facilitation of value co-creation and collaboration through shared digital workspaces, and integration of physical, cyber, and social spaces to achieve virtual-real symbiosis. Similarly, the key applications demonstrate how these functions are operationalized, such as in the collaborative design, commissioning, and execution of manufacturing processes, enhancement of human-machine interaction involving humans, COBOTS, and process management systems, and optimization of operations through improved planning, task execution, and strategic decision-making. Supporting these applications are enabling technologies, which include DTs, CPS, Big Data Analytics, Cyber Infrastructure, Human-Machine Interfaces, Metaverse Configuration tools, AI, and Edge Computing. These technologies collectively enable immersive experiences, seamless integration between real and virtual environments, operational efficiency, low latency, and robust data security, forming the technological backbone of IM.

Table 3. Key distinctions and synergies between DTs and IM in manufacturing OHS.

Aspect	DT	IM
Focus	Real-time replication & simulation	Immersive interaction & collaboration
Core Components	Sensors, data analytics, physical asset	Cyber-physical-social layers, avatars, XR
Maturity	Technically mature	Conceptually emerging
Human Engagement	Limited (focused on data insights)	High (immersive, multi-user)
Main Contribution to OHS	Predictive analytics, hazard modeling	Training, simulation, collaborative planning

shows the key differences and synergies between DTs and IM in manufacturing OHS.

To sum up, IM builds upon the foundations of Industry 4.0 and DTs by integrating CPS, immersive environments, and intelligent automation into unified, human-centric digital spaces. These environments support remote safety training, scenario-based planning, and cross-functional engagement. It creates a seamless bridge between physical production environments and immersive virtual realities, enabling users to simulate, plan, and manage complex operations through real-time interactions. While DTs provide precision simulations, IM enables experiential learning and collaboration, allowing remote users to rehearse coordinated safety responses and decision-making. It delivers immersive multi-sensory experiences that strengthen real-time decision-making, enhance virtual collaboration, and improve situational awareness. For instance, its convergence with DT and Extended Reality (XR) technologies drives the evolution of intelligent safety ecosystems. It enhances the effectiveness of traditional DTs by embedding human presence and interactions into virtual models. In summary, IM enhances DT functionality by enabling shared immersive experiences for training and hazard response. Its integration into OHS frameworks represents a shift toward collaborative and experiential safety designs.

4.1.3. Immersive Technologies

Immersive technologies, such as VR, AR, and XR, are transforming traditional OHS in high-risk industrial environments. According to the literature, these technologies play a central role in enhancing safety training and operational simulations. Advancements in VR, AR, and MR technologies are enhancing industrial applications both ergonomically and economically [66]. For example, immersive technologies provide interactive and controlled environments in which users can safely simulate and prepare for dangerous scenarios without real-world consequences [12,17,29]. Likewise, immersive environments enhance design, monitoring, programming, and training by combining digital technologies like DTs, robotics, and XR technologies [6,51].

VR is a computer-generated technology that creates immersive, interactive, three-dimensional (3D) digital environments that allow users to visualize and interact with simulations in real time using their natural senses and abilities [10,32,67,68]. Immersive VR technology is typically delivered through head-mounted devices with audio and haptic feedback, creating an absorbing experience by transporting users beyond physical reality with an enhanced sense of presence, realism, and high interactivity via handheld controllers [10,29,44,45,57]. For instance, VR headsets and handheld devices deliver interactive, step-by-step training for assembly and maintenance tasks [10].

Originally developed for gaming, VR has evolved into a valuable industrial tool for enhancing operations, product design, and especially real-time safety training through hazard identification and accident reconstruction in controlled environments [17,67]. In manufacturing, VR plays a transformative role by visualizing product development, consolidating information, aiding decision-making, and optimizing processes through early issue detection, improved assembly time, efficiency, and quality, leading to cost and time savings [1,69,70]. Likewise, a 3D VR environment in manufacturing serves as an immersive simulation tool that enables interactive safety training for operators by safely replicating real-world scenarios in a controlled environment [17,71].

VR is particularly effective for training operators in complex tasks, enhancing decision-making, and improving worker performance [10,68]. Its ability to reproduce environments and the safe management of risk scenarios make it a highly effective training tool for immersive safety training, including hazard identification and accident reconstruction [54,67,71]. Furthermore, it includes interactive step-by-step training for assembly and maintenance tasks [10]. Beyond technical applications, immersive VR nature experiences in the workplace during breaktime show significant promise for reducing psychological and physiological stress of workers [30].

AR is an emerging technology that overlays digital content, such as holograms, images, information, or instructions, onto the physical world, creating a seamless and interactive experience that enhances user perception and blurs the lines between the real and virtual environments [31,59,72]. AR enhances a real-time view of the physical world by overlaying virtual computer-generated information, relying on displays, input devices, tracking, and computers as its core components [52]. Users interact with a projected 2D or 3D image by pointing their device, such as a smartphone, tablet, or wearable device like smart glasses, at a specific image, which the system then processes [72]. AR can be divided into immersive experiences using wearable displays like AR glasses for high realism, and non-immersive experiences using screen-based devices like smartphones and tablets [42]. Depending on the hologram visualization technology used, AR can be categorized into mobile, spatial, and wearable forms [31], with AR devices primarily existing as head-mounted displays or handheld devices [59].

AR’s capacity to integrate virtual data with real-world environments offers significant applications across various fields, including enhancing human-robot interaction and providing cognitive support. For instance, AR enhances HRC by displaying overlay instructions directly onto real objects, boosting human trust, managing fear during accidents, and easing anxiety in HRC by visually delivering contextual information and enhancing the overall user experience [69,73]. This is complemented by simulation-based training, which improves the safety of HRC by enhancing training for working alongside robots [67]. Moreover, smart cognitive support tools, such as AR and intelligent interfaces, use AI to deliver real-time adaptive decision support tailored to an operator’s cognitive load and task demands [50]. In industrial environments, AR and VR tools facilitate applications such as assisted worker guidance, asset monitoring, visual guidance, and security risk warnings [2,69]. For instance, an automated guided vehicle may deviate from its expected path [2]. Wearable devices like smart glasses are particularly impactful, enhancing worker efficiency and reducing errors by overlaying real-time information directly onto the user’s view, enabling hands-free operation, and making them promising future workplace tools [38,72].

Extended Reality (XR), encompassing VR, AR, and MR, provides immersive and interactive learning that enhances engagement, retention, and motivation while lowering training costs [43]. These immersive technologies effectively fuse the digital and physical worlds to simulate reality (VR), overlay digital information on the physical world (AR), or create environments where physical and digital objects interact in real time (MR) [12,54]. According to [66], MR technology has recently been adopted to improve maintenance and production systems. Moreover, XR serves as a powerful tool to overcome the shortcomings of conventional training methods, utilizing VR for full immersion, AR for digital overlays on the real world, and MR for the interactive fusion of virtual and real elements [12]. For example, applications in manufacturing include using devices such as HoloLens for monitoring and controlling CNC milling machines, visualizing spacecraft cable assembly, and employing DT technology for complex product design, simulation, and virtual debugging [14].

A brief overview of the applications and functions of immersive technologies related to OHS in industrial settings is shown in Figure 10. The figure highlights the distinctive yet complementary roles of VR, AR, and XR in industrial training. VR is utilized for immersive safety training, offering realistic 3D simulations that facilitate hazard identification, accident reconstruction, complex task rehearsal, and emergency preparedness. This ultimately supports better decision-making and reduces worker stress. Conversely, AR enhances real-time worker support by overlaying digital information onto the physical environment, guiding tasks, improving perception, enabling human-robot collaboration, and providing cognitive assistance, while also aiding in asset monitoring and security alerts. Finally, XR serves as a holistic immersive learning framework that integrates VR, AR, and MR to deliver engaging and cost-effective training experiences. By blending digital and physical realities, XR enables full immersion (VR), contextual overlays (AR), and the interactive coexistence of virtual and real elements (MR), thereby enhancing knowledge retention, motivation, and overall training effectiveness.

In short, immersive technologies such as VR, AR, and XR are transforming training by enabling immersive, hands-on learning and realistic hazard scenario simulations, as supported by the reviewed literature. These technologies facilitate experiential safety training and environmental simulation. Moreover, they offer safer and more engaging alternatives to conventional training methods, improving hazard awareness, decision-making, and skill transfer. Their use enhances risk awareness, skill retention, and cognitive learning, especially in hazardous or confined environments. Recent studies have confirmed that VR scenarios improve hazard recall and learning retention, while AR enhances cognitive support and task accuracy in real-time settings. Additionally, their interactive and scalable nature makes them highly effective for safety training in diverse industrial settings.

4.1.4. Internet of Things, Sensors, and Motion Capture

The IoT functions as the digital backbone of Industry 4.0 and 5.0, connecting factory elements such as machines, tools, production lines, workers, and infrastructure through smart sensors that enable continuous real-time data collection and communication [3,9]. IoTs act as a crucial interface between the physical and cyber worlds, enabling technologies like DTs, which are virtual representations of physical assets. The IoT is considered the primary driver of Industry 4.0, serving as the foundational technology for CPS [61], enabling physical objects to sense, communicate, compute, and actuate, thereby possessing both physical and virtual identities [3]. Integrating IoT and AR further expands the capabilities of traditional Information Communications Technology (ICT) systems by merging physical and virtual realities with ambient data, creating context-aware ambient intelligence, and enabling multi-modal interaction [2].

Adverse conditions in industrial settings often require monitoring of factors like dust, noise, and heat, which can harm workers’ health. In manufacturing, IoT and sensor networks are revolutionizing operations by providing real-time monitoring of environmental factors, such as temperature, humidity, noise, and air quality [18,20]. This is exemplified by [74], who implemented a five-layer IoT-based architecture for real-time environmental monitoring in smart manufacturing. Industry 4.0 integrates IoT into manufacturing, driving digitization, optimization, and customized, knowledge-driven value creation. This continuous data collection from sensors, including gas and particulate matter, supports hazard detection, ergonomic assessment, and safety compliance through real-time alerts and notifications [20,33,34]. For example, as mentioned by [38], smart sensors and IoT devices can quickly identify fire hazards or other accidents, enabling automatic machine shutdowns and immediate alerts to local authorities for a rapid response. By connecting factory elements to the Internet, the IoT provides real-time data insights into production processes, helping to identify areas for efficiency improvement [9]. Additionally, IoT can provide technical solutions for preventing workplace injuries and illnesses [34]. IoT technologies can help prevent human errors and accidents, enhance operational efficiency, improve process safety, and mitigate the impact of severe incidents. For example, smart monitoring devices, such as smart watches and smart weight scales, within a human DT framework can track body composition, location, stress, and sleep, thereby improving worker safety in industrial settings [19].

Furthermore, motion capture (MoCap) systems, which use cameras and sensors to track the real-time movements of people or objects, play a crucial role in ergonomic analysis by converting movements into digital data [22,75]. The MoCap system continuously gathers and interprets human posture data in real time [28] and is increasingly used to program and control robots by enabling them to replicate human movements and interact seamlessly with their environment. This is accompanied by specialized sensors and vision systems that enable immediate responses to human actions [16] and the combination of multiple sensors that monitor various physical and behavioral biometric parameters with ML technologies [9]. Hence, MoCap technologies, as shown by [22] and [28], contribute significantly to ergonomic analysis and human-robot programming by digitizing movement data in real time.

Figure 11 provides a concise overview of the integrated applications and functions of IoT, sensor technologies, and MoCap systems in industrial environments to enhance industrial intelligence and safety. For instance, the IoT facilitates real-time data exchange between physical and cyber systems, supporting the implementation of DTs, CPS, and smart manufacturing. This connectivity enhances process efficiency, safety, and decision-making, while also enabling ambient intelligence and multimodal interactions. In contrast, sensors play a critical role in monitoring environmental and biometric parameters, contributing to hazard detection, safety compliance, and ergonomic assessments through continuous data collection and real-time alerts. Finally, MoCap technology enables real-time tracking of human movement and posture, supporting ergonomic analysis, robot programming, and seamless human-robot collaboration. Together, these technologies form a foundational layer for intelligent, responsive, and safe industrial environments.

To sum up, IoT, sensors, and MoCap technologies function as the digital nervous system of smart factories. They enable the continuous monitoring of workers and the environment through real-time data collection and communication. Several studies confirm the role of the IoT in integrating real-time physiological and ergonomic data into OHS. For example, IoT devices and sensors continuously collect environmental and physiological data, while MoCap systems support posture and movement tracking. These technologies provide the data required for real-time DT updates and AI-driven analyses. When this continuous stream of data is fed into DTs and AI systems, it aids in enabling advanced predictive analytics, ergonomic assessments, risk detection, and adaptive safety interventions. Together, these technologies create a responsive and intelligent workplace that prioritizes worker well-being and operational efficiency.

4.1.5. Artificial Intelligence & Big Data Analytics

AI and robotics are indispensable in modern industries, and operations without them are almost unimaginable today. AI-driven architecture for Industry 5.0 enhances manufacturing by integrating active learning, explainable AI, and simulated reality, prioritizing human-machine collaboration and safety [6,35]. AI enables machines to learn from past data, adapt to new inputs, perform complex human-like tasks, and retain information for future optimization [52]. Integrating historical and real-time data enhances manufacturing decisions, leading to improved performance, safety, reliability, and sustainability of industrial systems [36]. Furthermore, ML and big data analytics are key to predicting and preventing equipment failure in production systems [37]. Big data analytics aggregates and analyzes diverse information to enable superior real-time decision-making compared to traditional methods [8], providing organizations with a competitive advantage [43]. The study by [38] identified IoT, cloud computing, AR, and big data as key technologies closely linked to safety management.

Data acts as an essential connection between the physical and digital worlds, especially in virtual HRC scenarios, driving the HRC DT [46]. Data can be collected using various methods, such as sensors, data acquisition cards, and similar technologies. Human behavioral data can be collected efficiently via apps on smartphones or smartwatches, as well as through specialized wearable devices [76]. Effective data acquisition, including communication platforms, storage, and management, is essential for linking the physical and cyber domains, allowing feedback and decision support from the cyber world to the physical realm [15]. Wearable devices are compact electronic tools equipped with sensors and computational power that can be worn on various body parts to store diverse information [34,72]. For instance, wearable sensors can monitor human physiological data like heart rate and blood pressure, while a myoelectric cuff analyzes electromyographic signals to assess fatigue [46]. Digital transformation using IoT and big data analytics in health and safety systems has improved employees’ heart health, as seen in Shanghai’s BYD Manufacturing Enterprise [37]. Big Data enhances the ability to analyze human behavior and predict errors, thereby improving safety. Additionally, these wearable devices offer significant advantages such as high precision, flexibility, and adaptability [28].

When integrated with DTs, AI enhances automation, such as visual object recognition and decision-making through data analysis, and enables systems to interpret operator behavior and environmental conditions to deliver personalized robotic assistance [22,59]. For example, AI integrated with DTs can automate detection tasks, such as text extraction, damage assessment, inspection, and inventory management [59]. DTs, as real-time virtual models of physical assets, rely on AI and big data for diagnostics, forecasting, and optimization to inform decision-making and enhance performance [3]. Additionally, AI and data analytics are vital for identifying patterns in workplace accidents and illnesses, enabling companies to continuously improve their health and safety protocols [19]. Combined with AR, these technologies facilitate quick data exchange, proactively minimizing occupational risks during operational, maintenance, and complex tasks [18].

In summary, AI and big data analytics transform the way industrial systems forecast hazards, manage stress, and automate tasks. They are central to Industry 5.0, making modern manufacturing systems more intelligent, adaptive, and human-centric than before. AI and big data analytics play crucial roles in improving workplace safety by enabling advanced hazard prediction, recognizing worker behaviors and activities, and supporting informed decision-making. These technologies are widely documented as central to intelligent OHS that analyze sensor and behavioral data to enable predictive and personalized interventions. Furthermore, these technologies form the cognitive layer of digital safety ecosystems, allowing systems to learn from vast amounts of data, adapt to changing conditions, and proactively respond to potential risks. By analyzing patterns and trends, AI can anticipate unsafe situations before they occur, while also helping safety managers make faster and more accurate decisions. In addition, big data analytics can uncover hidden patterns that traditional assessments might miss. However, the use of AI also introduces challenges, particularly around transparency, for instance, how decisions are made, and trust, as workers and managers need to understand and rely on these systems. Addressing these concerns is essential for ensuring responsible and effective implementation.

Table 4. Key aspects of AI and big data analytics in the modern industry.

Aspect	AI	Big Data Analytics
Primary Role	Enables intelligent automation, learning, and human-machine collaboration	Provides the foundation for informed, real-time decision-making and predictive analytics
Impact	Transforms manufacturing (Industry 5.0), optimizes operations, and enhances worker support and safety	Improves manufacturing performance, prevents failures, and provides competitive advantage through data-driven insights
Data Interaction	Leverages data for analysis, decision-making, and system optimization (e.g., DTs)	Collects, aggregates, and processes vast datasets for insights and operational linking (physical-cyber)
Key Benefits	Adaptability, enhanced safety, personalized support, and automation	Superior decision-making, predictive maintenance, improved health and safety, and operational efficiency

shows some key aspects of AI and Big Data Analytics in Industry 5.0 settings in terms of their roles, impacts, and benefits.

4.1.6. Cyber-Physical Systems

CPS are intelligent, interconnected, and self-regulating systems that integrate computational and physical elements to monitor and control real-world processes in real time [3,9,21,36]. They form the technological foundation of industry 4.0 and 5.0 by enabling continuous interaction between physical operations and their DTs, facilitating decentralized decision-making and adaptive manufacturing [20,26,55,61]. They are complex, large-scale, and heterogeneous systems that rely on the interconnections of numerous networked devices [9]. Composed of sensors, actuators, controllers, and smart units, CPS connect machines, storage, and production systems into autonomous and self-regulating networks [26,53].

In manufacturing, CPS are increasingly adopted to enhance HRC, production efficiency, and automation [21]. Smart wearable devices, such as sensor-equipped smart soles, can effectively monitor shop-floor workers’ movement and well-being, contributing to safety in factory settings, and exemplifying CPS applications [26]. CPS also enables real-time synchronization between physical and virtual environments, forming the core of smart factories through integration with IoT, cloud computing, and big data analytics for predictive maintenance and operational optimization [61,77]. Likewise, CPS defines the core of the smart factory concept through its extensive reliance on intelligent robots as the primary production framework [11].

Cyber-Physical Production Systems (CPPS), a manufacturing application of CPS, are designed for automation, predictive maintenance, and collaborative robotics. CPPS rely on computing technologies like cloud, fog, and edge to perform tasks such as computation, modeling, and analysis, which are selected based on latency, bandwidth, and security requirements [15]. Cloud technology enables real-time information management and remote access, eliminating the need for physical proximity to computers. Cyber-Physical Social Systems (CPSS) and Human-Centric Cyber-Physical Systems (HCPS) place human needs at the center, integrating physiological and behavioral data via wearable sensors to enhance safety, personalization, and resilience in industrial environments [4,9,15,16,20]. These systems integrate human factors directly into the CPS design to enhance operator interaction, safety, and well-being, thereby making industrial plants more human-focused, resilient, and sustainable.

HCPS systems use smart sensors and modeling to simulate human-machine cooperation, enabling real-time feedback and tailored interfaces [16,19,20]. These systems also integrate edge computing and thermal imaging analysis to enhance safety management [9] and rely on 5G for high-speed communication, enabling real-time VR/AR task execution, remote control, and health monitoring [20,31]. Similarly, edge computing is widely used in AR image processing applications because it reduces latency and boosts performance by positioning computing devices near the data source [31]. Moreover, modern CPS are supported by advanced human-machine interfaces and immersive interaction technologies, such as AR, VR, and MR, and the utilization of these technologies can enhance worker safety and operational efficiency [4,15]. For example, VR enables virtual pre-assembly and training, AR provides real-time guidance and maintenance information, and MR offers versatile support across all manufacturing stages. Similarly, Cyber-Physical IM Systems (CPIMS), built upon CPS and immersive technologies, offer a virtual interface that enables collaboration between humans and CPS, creating an immersive experience [64]. Hence, as supported by multiple references, CPS enables real-time decision-making, predictive maintenance, and intelligent feedback loops, and is increasingly applied to integrate physiological feedback and AR/VR interfaces for safety optimization.

Figure 12 presents an overview of CPS, highlighting its key applications, functions, and benefits in manufacturing. These applications highlight the diverse contexts in which CPS is deployed, including smart factories, Industry 4.0/5.0 environments, HRC, predictive maintenance, health and safety monitoring, immersive training through AR, VR, and MR, and CPIMS. The functions section outlines the core operational capabilities of CPS, such as real-time monitoring and control, decentralized decision-making, integration of sensors, actuators, and intelligent units, adaptive learning and responsiveness, and modeling and simulation of human-machine interactions. Finally, the benefits emphasize the value of CPS in industrial systems, including enhanced safety and worker well-being, increased automation and efficiency, personalized and human-centric production, improved cyber resilience, operational optimization, and seamless integration between the physical and digital domains.

In summary, CPS are technologies that connect physical machines, such as robots or factory equipment, to digital systems that can monitor and control them. These systems collect data from the physical world in real time, process it quickly, and send back instructions to adjust how the machines operate. This constant exchange of information allows factories to respond quickly to changes, helping prevent accidents and improve safety. CPS forms the backbone of Industry 5.0 by linking virtual models with physical operations. Its ability to act and react in real time is vital for safe human-machine collaboration and decentralized decision-making. In smart manufacturing, CPS makes it possible to create flexible and responsive environments in which safety measures can adapt instantly to new situations. CPS provides a responsive layer to the digital ecosystem, translating data into real-world actions and enabling closed-loop safety control. The integration of digital technologies with physical systems is vital for CPS, cyber resilience, predictive analysis, and worker safety in industrial environments. CPS enables the realization of Industry 5.0’s vision by providing a technological foundation for adaptive, safe, and intelligent manufacturing operations where robots adapt to human needs, rather than the other way around.

4.1.7. Collaborative Robots

COBOTS or HRC play a pivotal role in Industry 5.0 by enabling safe, adaptive, and ergonomic collaboration between humans and machines. Unlike traditional industrial robots, COBOTS are designed to work safely side-by-side with human operators, particularly in assembly, packaging, and quality control [6,16,17,78]. They are particularly favored for assembly operations, assisting operators with complex tasks [41,69]. Moreover, COBOTS support social sustainability by improving physical ergonomics while maintaining production efficiency [39].

Effective and safe HRC relies critically on the monitoring and real-time recognition of human emotional and physiological states [21]. Safety design is a key priority for improving the safety of HRC assembly, which can be achieved by utilizing DT [41]. DTs play a crucial role in ensuring the safety and effectiveness of HRC. DTs use multi-dimensional models and fused data for the real-time control and optimization of physical processes, enabling new simulations and optimizations of HRC processes and strategies [46]. Safety is paramount in HRC, and DT-based simulations are key to proactively enhancing workplace safety by predicting and mitigating risks in real time, often prior to physical production [19,41,46]. This approach enhances cognitive functions and decision-making while addressing safety within a human-centered framework. DTs, combined with AI, XR, and advanced control techniques in Industry 5.0, improve HRC through real-time monitoring, intuitive interfaces, and adaptive control [6]. For instance, a DT model of the HRC environment can be integrated with VR for safety testing, using machine vision and deep learning to monitor human-robot distances and provide feedback via AR glasses for real-time adjustments [46].

Industrial COBOTS are vital for smart manufacturing because they boost output while maintaining safe work environments [58]. They offer a cost-effective solution for large industrial operations, such as plant maintenance, where full automation is impractical or too expensive [11]. COBOTS, equipped with advanced sensors and control methods, can independently perceive and analyze their surroundings and activities, ensuring safe human interaction while boosting productivity and product quality and preventing occupational health issues and injuries [53]. This is achieved through their ability to use visual data from cameras to understand their surroundings, make decisions, and react to changes [41]. Effective and safe HRC critically relies on the monitoring and real-time recognition of human emotional and physiological states. COBOTS also address ergonomic challenges by reducing physical strain, thereby improving workplace ergonomics, safety, quality, and productivity [24,39].

The HRC physical scenario comprises three layers: physical entities layer (robots, humans, and environment), perception layer (devices for data collection), and network layer (for data transmission), with DT models implemented in a software-based virtual scenario using game engines, such as Unity and Unreal Engine, analyzing and optimizing the physical entities [46,65]. Simulation software utilizes real-time data to model manufacturing systems, allowing engineers to virtually test, analyze, and optimize settings before implementing physical changes [52]. This enables the creation of a virtual production line with COBOTS by simplifying CAD models, animating them, and importing them into 3D engines to simulate the user interaction [25].

To conclude, COBOTS are designed to work safely alongside humans in shared workspaces, offering physical support and helping to reduce risks in industrial environments. COBOTS enhance collaborative ergonomics, safety, and operational efficiency through real-time sensing, adaptive behaviors, and virtual safety validation. Compared to traditional robots that operate in isolated zones, COBOTS are built to interact directly with people, that is, human-centricity, assisting with tasks that may be repetitive, heavy, or hazardous. Moreover, ensuring safety in these human–robot interactions requires careful ergonomic design to minimize strain and discomfort, as well as behavior modeling to predict and respond to human actions. Technologies like DTs and CPS play a key role by providing real-time data and feedback, allowing COBOTS to adapt their behavior instantly and maintain a safe and efficient working environment. The literature review shows that DTs, XR, and machine vision are crucial for COBOT simulation, monitoring, and risk mitigation. These studies highlight COBOTS’ importance in promoting proactive safety and intelligent cooperation within advanced manufacturing environments. Although COBOTS are advancing human-machine interactions in OHS by enabling shared tasks and physical assistance in industrial settings, their effective implementation still faces challenges, particularly regarding worker trust, system complexity, and the need for contextual awareness to ensure safe and intuitive interactions.

Table 5. Overview of COBOTS in smart, advanced manufacturing environments.

Aspect	Description
Core Function	Advanced robots (COBOTS) designed for direct, safe collaboration with humans, pivotal for Industry 5.0
Primary Use Cases	Assembly operations, packaging, quality control, and ergonomic support
Safety Paramount	Emphasizes real-time risk prediction, mitigation, and safety-by-design, heavily leveraging DT simulations and human state monitoring
DT Integration	DTs model, simulate, and optimize HRC processes in real-time, enhancing decision-making, safety, and operational efficiency through virtual environments
Enabling Technologies	AI, XR, advanced sensors, machine vision, and real-time data analytics for intelligent perception, control, and interaction
Key Benefits	Increased efficiency, productivity, product quality, improved ergonomics, and prevention of OHS issues, while offering a cost-effective solution for complex industrial tasks

presents an overview of COBOTS in smart manufacturing, highlighting their role in enhancing HRC, safety, and efficiency through intelligent connected systems.

4.1.8. Integration of Digital Technologies for Creating Advanced Digital Ecosystem

This section integrates emerging digital technologies into a cohesive framework designed to support immersive and intelligent OHS management. While each enabling technology offers standalone benefits, their true value emerges through integration. Figure 13 shows the advanced digital ecosystem (conceptual diagram) that can be visualized as a layered system with the DT at its core, surrounded by the broader IM and other interconnected enabling technologiesto enhanceg OHS in manufacturing. In a modern Industry 4.0/5.0 OHS ecosystem, DT is central, enabling real-time replication and risk simulation. It is surrounded by IM, which uses DT to build immersive hybrid workspaces. The integration of DT and IM technologies in manufacturing signifies a paradigm shift from reactive to proactive safety and operational strategies. DTs enable real-time monitoring and dynamic simulation, thereby supporting data-driven decision-making for OHS. In parallel, IM bridges the gap between physical and virtual environments through immersive technologies, enhancing worker training and hazard visualization. These tools not only mitigate risks but also contribute to ergonomic and psychological well-being, aligning with the human-centric vision of Industry 5.0.

DT is powered by inputs from the IoT, sensors, MoCap, AI, big data analytics, and CPS, with a reciprocal relationship with CPS. These technologies enable real-time data into the DT, allowing for accurate simulations and predictive modeling. For example, wearables, MoCap, and sensors send physiological, biomechanical, and environmental data to the DT, enriching its fidelity and responsiveness. The DT serves as a foundational element of the IM, contributing to its virtual representation and operational intelligence. A key finding is the synergistic effect of combining technologies. For instance, DTs enhanced by AI and IoT provide predictive analytics, while immersive tools offer practical training environments without real-world consequences. CPS and COBOTS ensure safe human-machine collaboration using real-time physiological data. For instance, COBOTS receive real-time input to adjust the grip, speed, and proximity, and simultaneously feed performance data to the DT and AI layers for continuous adaptation. These integrations culminate in a responsive and adaptive digital ecosystem capable of continuous improvement.

IM integrates immersive technologies such as XR, including VR, AR, and MR, for interaction, as well as AI and big data for processing, and COBOTS for physical-digital synergy. Immersive technologies interact with both DT and IM, supporting 3D safety training and ergonomic evaluation. They serve as a key interface layer, enhancing the data flow across all nodes and powering advanced safety training, monitoring, and decision-making. COBOTS are interconnected with all major components (IoT, AI, CPS, DTs, and immersive technologies) to enable intelligent, safe, and ergonomic collaborations. They rely on AI and big data for intelligent and safe operation and can be monitored and controlled remotely in immersive virtual environments. IoT, sensors, and MoCap feed real-time data into DTs to support COBOTS. CPS integrates physical and digital systems, linking both DTs and COBOTS. Moreover, the immersive layer also visualizes HRC scenarios, simulates adjustments, and provides AR/VR guidance to human workers, enhancing situational awareness and decision-making.

The arrows in Figure 13 illustrate the flow of data, influences, and interactions between the key components of the DT-based IM ecosystem. For instance, arrows flowing into the DT represent real-time input sources such as IoT, sensors, MoCap, AI, big data analytics, and CPS, which feed real-time data and intelligence into the DT. The feeding of real-time data aids in the simulation and monitoring of the environment. Similarly, the arrows from DT to IM show that DT is a foundational building block of the IM and indicate how DTs enable the construction of immersive spaces, contributing digital intelligence to the IM. Likewise, arrows into the IM represent technologies like immersive technologies, AI, and big data analytics that enable interaction and functionality within the metaverse. For instance, the arrows from AI and big data to DT and IM reflect their roles in real-time analytics and predictive modeling. Furthermore, bidirectional arrows with COBOTS indicate interactive feedback loops. For example, bidirectional arrows with COBOTS reflect their dual role of receiving real-time input for adaptive behavior and feeding back performance data to the DT and AI layers. These interactions enable COBOTS to operate safely, intelligently, and collaboratively with other components and humans. For instance, in IM, COBOTS can be monitored and controlled remotely using immersive virtual environments. Similarly, bidirectional arrows from and into immersive interfaces such as VR and AR support interaction in both directions. They function as two-way communication, that is, both input mechanisms (for training actions or commands) and output channels (for visualizations and scenario feedback), helping connect the physical and virtual worlds.

From an OHS perspective, this digital ecosystem fosters a proactive and predictive environment. It supports hazard prediction and prevention, immersive safety training, real-time monitoring, ergonomic optimization, and continuous improvement of safety protocols, ultimately transforming manufacturing safety into a dynamic and data-driven process. In short, the arrows map the ecosystem’s interdependencies, showing how data flows, technologies interact, and systems co-evolve to support a smart, immersive, and human-centric industrial environment from an Industry 5.0 perspective, addressing RQ1, RQ2, and RQ3. The integration of DT, IM, and enabling technologies into an interconnected and unified digital safety ecosystem represents the central contribution of this study. As shown in Figure 13, each technology plays a distinct and synergistic role in facilitating immersive training, ergonomic design, hazard forecasting, and collaborative safety. These technologies form an integrated safety system in which human data continuously shapes machine behavior in real time. This holistic framework (integrated approach) advances beyond siloed applications by enabling real-time, human-centered decision-making in smart manufacturing environments.

Moreover, the novelty of this study lies in its holistic synthesis of the interaction between DT and IM technologies and their collective enhancement of worker safety, ergonomics, and training effectiveness. The integration model shown in Figure 13 represents a novel contribution to the literature, particularly in addressing RQ3. While existing studies examine DT and IM separately, this work highlights how their convergence, along with supporting technologies such as CPS, AI, and big data analytics, can drive a shift from reactive safety strategies to a predictive, immersive, and adaptive human-centric OHS ecosystem. Unlike previous research, this study explores this interaction and presents a conceptual framework and technology taxonomy to guide both future research and practical implementation.

Additionally, insights from bibliographic analysis further support this integrated model; for instance, research items related to immersive training, predictive ergonomics, and DTs consistently cluster together, indicating that these themes are thematically central to OHS innovation in Industry 5.0. The high link strength clusters consistently focused on immersive safety training, DT-driven ergonomics, and AI-based risk forecasting. These findings confirm the ecosystem’s relevance to current research priorities and its originality in combining these aspects. The proposed advanced digital ecosystem addresses RQ 3 directly and offers a roadmap for safety transformation in data-rich environments. It also acknowledges potential challenges, such as cybersecurity, transparency, and trust in AI-driven systems, which must be addressed to ensure the responsible and effective adoption of AI in manufacturing environments.

4.2. Applications in Occupational Health and Safety in Manufacturing

Immersive and digital technologies play a significant role in modern safety training and incident management. The metaverse and other technological advancements are enabling innovative training methods with immersive technologies, offering a compelling alternative to traditional training [40]. The convergence of DT and IM technologies is reshaping the management of OHS in manufacturing. By integrating real-time data, virtual environments, and intelligent analytics, these technologies enable continuous monitoring, dynamic risk assessment, and immersive safety intervention. These technologies support the transition from reactive to predictive and proactive safety strategies, addressing ergonomic risks, hazard recognition, training limitations, incident management, and decision-making inefficiencies. This section explores how their combined use enhances workplace safety and provides examples of their implementation in real manufacturing contexts. Furthermore, this thematic synthesis reflects the bibliometric results (Figure 6 and Figure 7), linked to predictive monitoring and ergonomics, immersive safety training, and human-robot collaboration, providing a structured perspective on how these technologies are being applied across key manufacturing safety challenges.

4.2.1. Continuous Monitoring and Risk Prediction

Advanced emerging technologies are revolutionizing continuous monitoring and risk prediction in workplace safety and ergonomics. These include DTs, AI, big data analytics, wearable technologies and sensors, and immersive technologies. As detailed in

Table 6. Advanced emerging technologies for continuous monitoring and risk prediction in workplace safety and ergonomics.

Technology	Key Applications for Continuous Monitoring and Risk Prediction	Benefits	References
DTs - Digital Exoskeleton Models - Human DTs	Comprehensive ergonomic and safety assessment and prediction: Monitoring worker movements and WMSDs risks, simulating human activities in virtual environments for ergonomic analysis (including HRC), proactive fault prediction, and predicting workplace stress/burnout.	- More precise and comprehensive risk assessment - Improved efficiency in workplace planning and design - Real-time insights into worker well-being and system status - Enables proactive prevention of ergonomic issues and human errors - Reduces the need for human presence in hazardous environments during design and validation	[3,9,13,19,22,25,39,43,46,62,65,75,76,77,79]
AI & Big Data Analytics - AI-based algorithms, - Emotion AI, - Predictive Analytics, - Big Data Analytics	Intelligent hazard identification and predictive safety management: Analyzing vast data for risk evaluation, predicting equipment failures, understanding human emotions for mental health support, and identifying environmental threats (e.g., noise, hazardous gases) for proactive intervention.	- Automation of risk detection and alerts - Proactive mental health support and intervention - Enhanced predictive capabilities for equipment maintenance and incident prevention - Informed decision-making and comprehensive safety performance monitoring - More efficient risk management through data-driven insights	[4,6,7,8,16,18,19,46,76]
Wearable Technologies & Sensors (leveraging IoT) and MoCap - Smart PPE (helmets, wristbands), - IoT devices (gas sensors), - Motion sensors, - Mocap systems, - VR ErgoLog system	Real-time worker well-being, environmental monitoring, and ergonomic assessment: Tracking physical and mental well-being, detecting unsafe behavior and environmental threats, monitoring vital signs, analyzing posture/movements for ergonomics, providing timely warnings for fatigue/falls, and automated/objective ergonomic assessments via motion capture	- Continuous, unobtrusive data collection on worker health and environment - Immediate alerts and proactive risk management - Enhanced emergency response capabilities - Improves ergonomics and reduces physical strain with real-time feedback - Facilitates proactive maintenance based on machinery conditions - Enables accurate and objective ergonomic evaluations	[7,8,16,18,19,25,27,33,34,50,51,53,55,74,75,80]
Immersive Technologies - VR - AR - XR	Immersive risk visualization and real-time guidance: Cost-effective body tracking for ergonomic risk evaluation, simulating human-robot interactions and confined spaces, overlaying safety information and real-time data for enhanced situational awareness, displaying warnings on equipment, and guiding operators through procedures	- Safe, immersive environments for ergonomic assessment and risk visualization - Enhanced situational awareness and perception for workers - Real-time guidance and alerts to prevent incidents - Improved communication and collaboration in hazardous environments - Supports comprehensive virtual prototyping and layout design	[1,6,7,14,31,62,67,71,72]
Human-Robot Collaboration (HRC) Systems - Bionic Exoskeletons, - DT-powered HRC systems	Optimized human-robot interaction for safety: Enhancing workplace accessibility by reducing physical demands, and modeling/simulating human-robot interactions for ergonomic assessments and workload reduction	- Improved worker safety and well-being in hybrid workplaces - Optimized human-robot interactions for productivity and safety - Enables safer and more efficient workstation design - Reduces physical strain on workers	[11,13,22,39,41,46]

, these technologies offer comprehensive solutions for improving worker well-being, enhancing safety, and optimizing workplace design through continuous monitoring and risk prediction.

DTs play a central role in enhancing workplace risk management by improving both safety and effectiveness. As indicated in Table 6, DTs are continuously updated with real-time data to provide enhanced safety monitoring and visualization. DTs, when integrated with IoT and sensor technologies, enable the continuous monitoring of human posture and ergonomic stressors to predict musculoskeletal risks. For example, DTs support comprehensive ergonomic and safety assessments by monitoring worker movements and WMSD risks, often integrating with digital exoskeleton models [79], which enhance workplace accessibility and human-machine collaboration. Similarly, human DTs are crucial for ergonomic assessments, simulating human activities to reduce fatigue and workload, and monitoring, optimizing, and predicting processes for human well-being. AI and multi-source data further enhance these models by enabling real-time simulation and assessment of worker movements. For example, this integrated approach enables proactive safety interventions, triggering preventive actions to assist workers during emergencies [53]. Moreover, DTs also contribute to workplace design, product and process improvement, and reduction of WMSDs and stress by predicting fatigue and suggesting timely actions. This reflects a shift from reactive compliance to anticipatory safety, aligned with the Safety 4.0 framework. Additionally, integrating DT-based live monitoring with AR-powered notifications significantly improves process safety and enhances hazard prevention.

AI-based algorithms and big data analytics are pivotal for intelligent hazard identification and in predictive safety management. Big data enables real-time OSH monitoring through continuous data collection and processing and helps organizations track safety performance and make informed decisions to prevent accidents. These tools analyze vast datasets to evaluate risks, predict equipment failure, and identify environmental threats. They also support the mental health of workers by interpreting emotional data to proactively trigger support mechanisms, preventing severe mental health issues.

Similarly, wearable technologies and IoT sensors, such as smart PPE, IoT devices, motion sensors, and MoCap systems, enable continuous and seamless data collection on worker health and environmental conditions. These digital tools provide real-time alerts for unsafe behavior and support proactive risk management. For instance, these technologies can be used to monitor vital signs, detect falls, and analyze posture to improve workplace ergonomics [8,50]. Wearable trackers monitor and analyze human performance in demanding environments and issue timely warnings [80]. By integrating wearables, sensors, and ML, predictive modeling can create smart personal protective equipment (PPE) to prevent workplace accidents [18]. For instance, Operator 4.0 leverages digital technologies to enhance productivity, well-being, and safety [50]. Additionally, IoT devices, such as gas sensors, further improve industrial safety by detecting environmental threats and alerting operators in real time [7,34]. This is also supported by [55], where smart factories use intelligent sensors, cloud computing, and IoT to promptly detect fires and other accident risks. In summary, the benefits of leveraging these technologies include continuous data collection, immediate alerts, enhanced emergency response, improved ergonomics, reduced physical strain, facilitated proactive maintenance, and accurate ergonomic evaluations.

Immersive Technologies, particularly VR and AR, provide immersive risk visualization and real-time guidance, enhancing situational awareness and training. They offer cost-effective body tracking for ergonomic risk evaluation, simulate human-robot interactions and confined spaces, overlay safety information and real-time data for enhanced situational awareness, display warnings on equipment, and guide operators through the procedures. For instance, VR enables accurate ergonomic evaluations through virtual environments and MoCap and simulates confined spaces for safety analysis. For example, the VR ErgoLog system supports the analysis and improvement of human movement in virtual settings [75]. Additionally, VR also facilitates ergonomic studies in confined spaces and HRC. Furthermore, AR tools overlay critical safety information and real-time data to enhance the situational awareness of workers. They help prevent individual, collective, and environmental risks by detecting hazards and supporting workers in real time. For example, AR goggles with multispectral and infrared vision extend sensory perception, while AR headsets provide real-time task instructions to prevent errors and accidents [7,31], such as environmental contamination, unsafe equipment positioning, or harmful physical fields.

Finally, HRC systems, such as bionic exoskeletons and DT-powered interfaces, optimize the interaction between humans and robots by reducing physical demands, simulating ergonomic workflows, and enhancing safety and productivity in hybrid workspaces. Collectively, these technologies form an integrated ecosystem that advances safety, efficiency, and human-centric design in modern industrial environments.

4.2.2. Safety Simulations and Incident Management

Advanced emerging technologies are transforming safety simulations and incident management in manufacturing, shifting the focus from reactive responses to proactive preparation and real-time interventions.

Table 7. Advanced emerging technologies for safety simulations and incident management.

Technology	Key Applications	Benefits	References
DTs - Real-time DTs, - Virtual simulation (simulation platforms) - DT-powered HRC systems	Proactive risk mitigation and incident preparedness: - Facilitating dynamic risk assessment and fault prediction for machinery and production processes - Enabling virtual development and validation of safe HRC assembly systems - Simulating human worker ergonomics in various scenarios (e.g., confined spaces, robot coexistence) for safety decision-making - Reducing human presence in hazardous environments through virtual simulation	- Proactive risk mitigation and prevention - Enhanced operational efficiency and reduced downtime - Improved safety in human-robot collaboration - Enables safe testing and optimization of processes - Reduced need for human exposure to real hazards	[1,3,6,9,16,18,41,49,58,63,76]
AI & Big Data Analytics - AI-equipped sensors - AI-based algorithms - Intelligent risk assessment - Smart surveillance - Big Data Analytics	Intelligent incident detection, prediction, and response: - Real-time monitoring of the production floor for safety dangers and violations - Predicting equipment failures and enabling precise diagnoses - Automating hazard identification and generating preventive information for CPS - Ensuring PPE compliance and alerting workers to safety breaches in hazardous areas	- Enhanced real-time hazard monitoring and threat detection - Automated incident prevention and proactive maintenance - Improved compliance with safety protocols - More informed and data-driven decision-making in safety management - Reduced human error and improved overall workplace safety	[7,11,16,17,38,55,76]
Immersive Technologies - VR - AR - XR	Realistic incident simulation and real-time operational support: - Training employees for safe operations and evaluating human responses in emergencies through immersive VR scenarios - Simulating assembly line operations and providing step-by-step guidance for complex procedures (AR) - Overlaying critical safety information (e.g., warning signals on malfunctioning equipment, robot paths) directly onto the worker’s field of vision (AR) - Enabling risk-free practice of dangerous real-world scenarios - Facilitating decision-making for operator movement and emergency evacuations	- Safe, cost-effective training for high-risk scenarios - Enhanced situational awareness during operations - Real-time operational guidance and error prevention - Improved emergency preparedness and response - Reduced risk of actual accidents and injuries	[6,7,10,12,14,18,23,27,28,29,31,32,37,40,42,44,45,52,54,59,60,65,67,68,69,71,81]
Smart Sensors, Wearable Technologies, and Human-Machine Interfaces - Wearable devices - Environmental sensors - Smart sensors, - IoT devices	Real-time data collection for situational awareness and alerts: - Measuring and collecting real-time data on environmental threats and worker physiological states - Alerting operators to danger zones via audio-visual signals - Providing feedback on worker behavior and environmental conditions to trigger safety interventions	- Continuous monitoring of workplace environment and worker status - Immediate detection of hazardous conditions - Proactive alerts to prevent incidents - Supports data-driven decision-making for safety	[7,15,16,18,34,53,69,81]

provides an overview of these emerging technologies, highlighting their key applications and benefits in manufacturing environments.

At the core of this transformation are DTs and simulation platforms like Unity and Unreal Engine, which enable the creation of virtual replicas of physical systems and processes. These technologies are applied across various domains, including real-time DTs, virtual simulation platforms, and DT-powered HRC systems, to support proactive risk mitigation and incident preparedness. These tools enable dynamic risk assessment, fault prediction for machinery and production systems, and virtual development and validation of safe HRC assembly processes. By simulating human ergonomics in diverse scenarios, such as confined spaces or collaborative robot environments, DTs support informed safety decision-making and reduce the need for human presence in hazardous areas during system planning and design [9,18]. This virtual testing environment allows for the safe optimization of procedures, significantly enhancing operational efficiency, reducing downtime, and preventing both system and human errors [3,76]. Furthermore, DTs contribute to worker well-being by identifying ergonomic risks and enabling the evaluation of HRC scenarios, ultimately leading to improved safety, reduced physical strain, and greater resilience in hybrid workspaces.

AI and big data analytics complement DTs by enabling real-time incident detection, prediction, and automated responses [7,11,16,17,76]. These technologies leverage AI-equipped sensors, smart surveillance systems, and predictive algorithms to monitor production environments for safety violations, predict equipment malfunctions, and ensure compliance with PPE protocols. In addition, AI-powered simulations and virtual models analyze large volumes of data to generate preventive insights for CPS, thereby enhancing safety standards and enabling proactive management strategies. Applications include intelligent risk assessment, automated hazard identification, and real-time alerts for safety violations in hazardous zones. The integration of these tools leads to enhanced hazard monitoring, automated incident prevention, proactive maintenance, and data-driven decision-making in safety management. Ultimately, this reduces human error, improves compliance with safety protocols, and strengthens the overall workplace safety.

Immersive Technologies, including VR, AR, and XR, play a pivotal role in enhancing safety training and real-time operational support. These technologies offer realistic simulations of emergency scenarios, allowing workers to practice responses in risk-free virtual environments, thereby improving their decision-making and preparedness. For example, VR simulators are widely used for training in hazardous scenarios, improving decision-making, and evaluating emergency responses without real-world risks [45,67]. VR simulators are widely used to train employees for hazardous situations, while AR enhances situational awareness by overlaying critical safety information, such as warnings on malfunctioning equipment or robot paths, directly onto the worker’s field of view [6,7,31,37,60]. Similarly, XR technologies provide cost-effective training for high-risk tasks and support real-time hazard identification. In addition, AR facilitates step-by-step guidance for complex procedures and visualizes safety measures in shared workspaces to prevent collisions and improve maintenance efficiency. In the context of Industry 5.0, immersive technologies are integrated with AI and smart surveillance systems to enable intelligent factory safety management, including real-time hazard monitoring and PPE compliance alerts when workers enter hazardous areas. Collectively, these tools contribute to enhanced emergency preparedness, real-time operational guidance, error prevention, and reduced risk of accidents and injuries, while also supporting collaborative and adaptive safety strategies in modern industrial environments.

Finally, sensory devices and human-machine interfaces play critical roles in providing real-time data inputs. These systems continuously monitor environmental threats and worker conditions, trigger immediate alerts, and support data-driven safety interventions. For instance, smart sensors and devices like wearables and AR devices aid in measuring KPIs, tracking objects and personnel, and monitoring bio-sensory data for improved ergonomics and OHS [15]. Ensuring a safe, healthy, and effective future workplace relies on integrating sensory devices into the work environment and processing data using AI [16]. Likewise, virtual simulations and real-time data from wearable safety gear can aid decision-making for operator movement and emergency evacuations when chemical concentrations exceed the safe exposure limits [18]. Together, these digital technologies form a comprehensive and highly responsive safety ecosystem for modern manufacturing and Industry 5.0 settings.

4.2.3. Immersive Safety Training

Emerging immersive technologies (VR, AR, and XR) are revolutionizing immersive safety training in manufacturing by offering effective, safe, and engaging alternatives to traditional methods. These technologies are essential for providing safe, risk-free training, reducing accidents, and enhancing situational awareness in hazardous environments.

Table 8. Advanced immersive technologies for safety training.

Technology	Key Applications	Benefits	References
VR - VR simulators - VR-based training systems	Simulating hazardous scenarios and complex operations: - Recreating high-risk facilities (e.g., chemical plants, steel plants) and dangerous scenarios for risk-free practice - Simulating complex machinery operation and assembly line procedures - Enabling virtual interaction with machinery and objects - Allowing workers to safely experience realistic scenarios without real-world consequences - Improving decision-making and evaluating emergency responses - Reducing psychological stress through immersive nature experiences	- Safe practice in high-risk environments without actual exposure - Cost-effective alternative to traditional training (reduces equipment costs, travel) - Improved skill transfer, retention, and training effectiveness - Accelerated comprehension and confidence building - Potential for stress reduction and improved well-being	[7,8,10,12,14,23,28,29,30,32,40,44,45,54,56,62,67,68,69,71]
AR - AR glasses/headsets - AR-based safety assistants	Real-time operational guidance and hazard identification: - Superimposing digital information onto the real world to create interactive learning environments - Guiding operators with real-time instructions for tasks (e.g., running machines, product creation) - Providing remote assistance - Overlaying critical safety information, warning signals on malfunctioning equipment, and visualizing safety measures in shared workspaces - Enhancing hazard identification skills and extending workers’ sensory perception	- Enhanced situational awareness during real-world tasks - Real-time guidance reduces errors and accidents - Improved knowledge transfer and decision-making - Safe simulation that blends virtual and real elements - Prevents collisions and enhances industrial safety	[7,14,31,37,38,42,52,55,72,81]
XR - XR systems (encompassing VR, AR, MR)	Broad spectrum immersive training and cost-efficiency: - Providing safe and cost-effective training for dangerous tasks and industrial maintenance - Offering new opportunities for worker safety across various industrial environments - Enabling virtual operation training systems for safe and effective machinery interaction	- Broad applicability across diverse industrial training needs - Highly effective, customizable, and scalable learning experiences - New avenues for improving worker safety	[7,12,43,53]

provides an overview of the key benefits and applications of using these immersive technologies for interactive safety training in industrial settings.

VR provides a compelling platform for simulating hazardous scenarios and complex operations, allowing workers to safely experience realistic situations and build confidence without real-world consequences [32,40]. This includes recreating high-risk environments, such as chemical or steel plants, for risk-free practice. For example, an immersive safety training program uses a 3D virtual replica of a steel plant’s casting bay, where operators learn through VR experiences involving material transfer with electric overhead travelling cranes [71]. Furthermore, VR significantly improves skill transfer, retention, and overall training effectiveness, leading to fewer accidents than traditional methods [8,62]. For example, VR can digitally replicate hazardous facilities for proactive safety training [32] and even reduce psychological stress, contributing to improved workplace well-being [30]. Similarly, VR enables a comprehensive evaluation of trainee behavior in simulations [23,29,40,54], aids in hazard identification [45,54], and reduces the need for physical facilities, making training more cost-effective. By immersing workers in interactive, real-world-like scenarios, VR fosters decision-making and emergency response skills [42], resulting in successful skill transfer and training outcomes that are comparable to traditional methods [62]. In summary, immersive VR has the potential to transform safety training and evaluation by reducing accidents, enhancing skills, and promoting safer practices in high-risk industries.

AR enhances safety training by overlaying digital information onto the real world and creating interactive learning environments that improve training outcomes. For instance, AR glasses provide real-time safety alerts, remote assistance, and guided instructions for hazardous situations [72]. This dual-view capability of viewing both real and digital environments simultaneously improves tasks such as repair, maintenance, and training, increasing efficiency and reducing risk [18]. Moreover, AR supports safety compliance by enabling operators to pre-train on high-risk tasks in virtual environments. The primary goal of AR-based training is to improve knowledge and decision-making by immersing workers in interactive, hazard-free simulations that blend virtual and real-world elements [42]. Dynamically collected safety data can alert operators to danger zones through audio-visual signals or guided motion paths, thereby enhancing situational awareness [69]. Wearable devices and environmental sensors further enrich AR training by collecting real-time worker data, leading to more personalized and effective learning experiences.

XR, which encompasses VR, AR, and MR, offers a comprehensive suite of immersive training solutions that are cost-efficient and scalable. XR systems deliver customizable, effective, and scalable learning experiences [12], opening new possibilities for improving worker safety in diverse industrial settings [43]. The development of virtual operation training systems using XR enables safe and effective training without the risks associated with using actual machinery [12]. Additionally, organizing safety training in immersive virtual environments is a sustainable approach that reduces environmental impact and costs while enhancing behavioral skills and safety awareness [44]. The integration of smart sensors, wearables, and VR and AR technologies supports employee safety training and hazard detection, helping to shape effective safety measures and control strategies [18]. In summary, immersive technologies provide a powerful alternative to traditional training by simulating real-world scenarios in a safe, resource-efficient, and sustainable manner.

While traditional systems, such as 2D manuals, e-learning modules, checklist-based audits, and video or PowerPoint training, have long supported industrial safety through structured, rule-based approaches, they often fall short in adapting to real-time conditions or addressing human-centered risks. In contrast, emerging technologies, such as DTs and XR-enabled metaverse platforms, offer predictive and immersive tools that enhance training, situational awareness, and ergonomics. When integrated with DTs and immersive simulations, COBOTS and CPS enable adaptive, real-time human-robot interaction. As demonstrated in [46], DT and VR systems can dynamically adjust robot behavior based on worker position and fatigue, enhancing trust, reducing cognitive load, and ensuring ergonomic alignment in shared workspaces. However, these advanced systems also introduce challenges related to implementation complexity and data integration, which are discussed in the following sections.

To conclude, this study synthesizes multidisciplinary insights (Figure 5) into a unified, advanced digital ecosystem for OHS, consolidating diverse digital technologies into a cohesive safety innovation framework for smart manufacturing. This highlights how emerging technologies, such as DT and IM, are not only reshaping industrial operations but also enabling new paradigms and innovative approaches to workplace well-being. While adoption challenges remain, the convergence of DT and IM technologies provides a forward-looking blueprint for predictive, immersive, and human-centric safety strategies that are aligned with the principles of Industry 5.0.

More importantly, this integration goes beyond the existing literature by linking predictive, immersive, and real-time systems into a human-centric digital ecosystem, offering practical pathways to advance OHS beyond compliance toward resilience and sustainability. Additionally, this study serves as a foundation for future empirical validation and real-world implementation of intelligent OHS frameworks in smart manufacturing environments.

Furthermore, the key insights from the previous sections underscore the maturing landscape of enabling technologies in OHS applications. This study identifies the convergence of immersive, predictive, and integrative technologies that facilitate real-time monitoring, adaptive training, and personalized safety interventions. These findings align with recent research advocating for a proactive, data-driven, and human-centric redesign of OHS strategies. These findings align with recent research advocating the proactive, data-driven, and human-centric redesign of OHS strategies [3,8,36,52,59,76]. Finally, bibliographic coupling analysis reinforces the centrality of immersive training, digital ergonomics, and intelligent system integration in shaping the current discourse on OHS transformation.

4.3. Advantages of Using Immersive Reality for Safety Training and Incident Management

The integration of interconnected digital immersive reality technologies, such as XR, VR, and AR, into OHS training offers a wide range of advantages. These technologies create dynamic, real-time training environments that significantly enhance user interaction, engagement, and cognitive learning. Studies have shown that such immersive digital tools foster a deeper understanding and retention of safety protocols, making them highly effective for modern training needs [12,53,71]. The rapid advancement of affordable and versatile VR and AR headsets has unlocked new applications across industries, including manufacturing. For instance, AR enhances operator safety in the food industry by forecasting the necessary cleaning and replacement of porous sieves in hot-break juice extractors [72]. Similarly, VR technology enables operators to remotely perform hazardous procedures without compromising performance [45]. VR is gaining significant recognition in safety training due to its potential to offer numerous benefits over traditional training methods. For example, VR offers benefits like remote presence, flexible training schedules, and instant, risk-free, interactive learning experiences [62].

Another key benefit of these technologies is their ability to enhance cognitive resilience and awareness. AR supports operators in maintaining mental agility and reducing the likelihood of errors by overlaying critical information in real time. Meanwhile, VR provides a safe, controlled environment in which users can practice crisis management scenarios [16], thereby building mental resilience and improving decision-making speed and accuracy. For example, AR and VR multimedia instructions improve operator cognitive abilities and decision-making and reduce cognitive workload compared with paper-based methods [10]. Additionally, AR helps workers work better with robots by showing helpful information on real objects, building trust, and reducing fear or stress. For example, AR improves HRC by overlaying real-time instructions on physical objects, fostering trust, reducing fear, and enhancing the user experience [69,73].

XR technologies also contribute to broad-spectrum safety improvements in various industrial settings. They enable the development of virtual operation training systems that allow workers to engage in realistic and risk-free simulations. Furthermore, this advanced approach not only minimizes the dangers associated with handling actual machinery but also opens up new opportunities for safety innovation and workforce preparedness [12,14,43]. Moreover, AR-based safety training enhances workers’ knowledge and decision-making capabilities through immersive simulations that blend virtual and real-world elements. These simulations provide a highly effective alternative to traditional training methods, offering customizable, scalable, and safe learning experiences that can be tailored to specific workplace scenarios [12,42]. In short, XR overcomes the traditional training limitations by using VR for immersion, AR for real-world overlays, and MR to merge the physical and virtual environments interactively.

Finally, the use of immersive virtual environments for safety training promotes sustainability and cost efficiency. Immersive technologies such as VR offer a cost-effective, safe, risk-free, realistic, and comprehensive approach to safety training compared to real-world training environments [40]. Real-world testing is often costly and time-consuming, with limited design iterations, whereas VR enables faster and more affordable testing with more iterations to optimize design, safety, and efficiency [70]. Furthermore, by reducing the need for physical resources and minimizing environmental impact, these technologies support greener training practices. Additionally, collecting real-world data in hazardous manufacturing is costly and risky; however, immersive VR can simulate environments and procedures using synthetic data generated via game engines like Unity, using AI and ML, reducing time, cost, and effort while enabling tailored, scenario-specific datasets [17].

Compared to traditional safety systems, such as paper protocols or static training videos, emerging technologies like DTs and XR-based simulations provide dynamic and adaptive safety environments. However, these advantages are not universal. Although immersive systems can enhance training retention in complex scenarios, they require a robust digital infrastructure and user proficiency. In contrast, conventional technologies, although less flexible, are often more cost-effective, reliable, and easier to scale, especially within legacy systems. Therefore, transitioning to digital ecosystems requires a careful balance between innovation and practical feasibility, particularly in the case of SMEs.

To conclude, embracing technological innovation provides businesses with substantial benefits, such as reduced costs, technological progress, better management, and improved employee safety. Among these innovations, immersive reality technologies stand out for their ability to help lower operational costs while simultaneously enhancing participants’ behavioral skills and safety awareness, making them a smart strategic investment for forward-thinking organizations and a key enabler in transforming to Industry 5.0. The integration of DT, IM, XR, IoT, AI, CPS, and COBOTS forms a proactive OHS ecosystem that shifts safety management from reactive to predictive. This advanced digital ecosystem enables the continuous monitoring of physical and cognitive states, virtual scenario-based training, and early risk detection. Recent studies [6,16,36,52,59,71] support the benefits of this approach, including improved worker engagement, predictive incident prevention, and ergonomic optimization. Moreover, these advanced digital ecosystems not only mitigate operational risks but also align safety innovations with the broader goals of Industry 5.0.

Table 9. Key advantages of using immersive reality technologies.

Key Advantages	Description	References
Cost and Time Efficiency	- Significantly reduces time and costs associated with training, product development, and testing. - Eliminates the need for physical setups, expensive equipment, and on-site training sessions - Enables faster, more affordable testing with increased design iterations - Reduces time and effort by utilizing synthetic data generated via game engines	[4,8,12,14,17,40,44,45,62,63,70,71,72]
Enhanced Safety and Risk Mitigation	- Offers a safe, risk-free, and comprehensive approach to training compared to real-world environments - Allows for practicing dangerous scenarios without actual exposure to risk - Improves overall safety through realistic accident scenarios and plant simulations - Enables early detection of deviations (e.g., in machinery or production) and allows remote performance of hazardous procedures	[38,40,44,45,54,55,63,67,71,72]
Improved Learning and Performance	- Boosts decision-making, enhances realism, and increases knowledge retention. - Builds trainee confidence, fosters high engagement, and improves safety performance - Enhances risk and hazard recognition skills - Improves operator cognitive abilities and reduces cognitive workload compared to paper-based methods - Facilitates acquiring knowledge and skills to reduce injuries, improve PPE usage, and build familiarity with emergencies	[10,12,32,43,45,54,55]
Operational Optimization and Efficiency	- Enables the evaluation and modification of complex systems and optimization of strategies - Reduces product development time and prevents errors through integrated knowledge - Aids in optimizing production processes, reducing downtime, and enhancing quality assurance - Allows predictive maintenance and helps in boosting production and lowering costs	[38,56,63,72]
Flexibility and Accessibility	- Provides remote presence and flexible training schedules - Eliminates the necessity for physical training sessions and employees leaving their usual workplace - Offers a safe and repeatable solution for industrial robotics training, overcoming challenges like expensive equipment and complex programming - Accommodates multiple workers with programs tailored to individual characteristics, experiences, needs, and roles	[12,44,45,54,62]
Enhanced Situational Awareness and Guidance	- Overlays crucial safety information and warnings onto the worker’s field of vision in real-time - Enhances user experiences by merging real-world information with the virtual world - Forecasts necessary cleaning/replacement in industrial equipment to enhance operator safety	[2,55,66,72]

provides an overview of the key benefits of using these interconnected immersive reality technologies for safety training and incident management, answering RQ4.

Finally,

Table 10. Key applications and OHS benefits of DTs, IM, and other interconnected digital technologies in manufacturing.

Technology Group	OHS-Related Advantages	Key Applications
DT and IM	- Real-time monitoring of workers and environmental conditions - Predictive analysis of risks (e.g., fatigue, ergonomic strain, hazards) - Human-centric simulation for planning and decision-making - Remote collaboration and immersive industrial safety design	- Human-robot safety modeling - Digital exoskeleton integration - Simulation of process hazards and workspace ergonomics
Immersive Technologies (XR, VR, AR)	- Risk-free training environments for safety scenarios - Real-time instructions and hazard alerts - Ergonomic assessment and cognitive load visualization	- VR steel plant simulations - AR-based guidance for machine operation - Confined-space risk training
Wearables, IoT, and MoCap	- Continuous monitoring of vitals and posture - Early detection of fatigue, stress, or physical risks - Support for smart PPE and fall detection	- IoT wristbands and helmets - MoCap-based posture analysis - Sensor-driven alerts for unsafe conditions
AI, ML, and Big Data Analytics	- Detection of abnormal patterns in worker behavior or environmental conditions - Prediction of fatigue, burnout, and injuries - Adaptive responses and automated alerts	- Emotion AI in Human DTs - Data-driven safety dashboards - Real-time noise/hazard environment mapping
HRC Systems	- Ergonomic collaboration strategy development - Simulation and validation of safe HRC workflows	- VR-integrated HRC testing - Layout optimization for collaborative workstations

concludes the roles of DTs, IM, and interconnected advanced digital technologies in enhancing OHS. It shows the key applications and OHS-related advantages of these technologies in a manufacturing context.

4.4. Implementation Challenges

RQ5 aimed to identify the challenges and limitations of the large-scale adoption of DT, IM, and other interconnected immersive technologies in developing an advanced digital ecosystem and integrating them into OHS practices within manufacturing. Technological advancement is widely regarded as a catalyst for sustainable development, driving innovation and improved solutions [82]. However, advanced digital transformation introduces significant challenges, including high costs and scalability, data privacy and data management complexities, and cultural and organizational resistance. The implementation of IM in manufacturing faces significant challenges, such as cybersecurity, cost, standards compatibility, and computational demands, as metaverse technology continues to evolve and mature [64]. Technological immaturity is a barrier faced by several industries, such as textiles [83] and construction [84]. The challenges and barriers identified through the review of research items addressing RQ5 are discussed in the following section.

4.4.1. Scalability and Cost

The adoption of advanced immersive technologies in traditional workplaces faces significant challenges related to their scalability and cost.

Table 11. Key challenges related to scalability and cost.

Barrier	Description	References
High Costs and Investment Uncertainty - Initial investment - Compatibility and integration costs - Cost justification and ROI uncertainty	- Significant financial outlays for hardware, software, setup, and specialized personnel. - Integrating new technologies into existing legacy infrastructure requires expensive system upgrades. - High upfront investment often leads to uncertainty regarding ROI, despite potential long-term benefits, making adoption challenging, especially for SMEs	[6,12,33,37,43,45,57,64,68,85]
Technical and Infrastructure Limitations - Computational and storage requirements - Network infrastructure limitations - Hardware limitations & design inefficiencies	- VR, DT, and metaverse applications demand high computational power, storage capacity, and ongoing resource-intensive maintenance. - Immersive technologies require ultra-high network reliability, low latency, and high throughput, often exceeding current 5G capabilities. - Hardware design inefficiencies such as poorly designed VR setups, limited brightness, lack of haptic feedback, and physical discomfort hinder usability and effectiveness.	[6,12,33,34,37,54,56,57,64,68,69,74,86]

outlines the critical obstacles related to scalability and cost in integrating advanced digital and immersive technologies, such as XR, VR, AR, DT, and IM, into OHS practices. These barriers are categorized into two main areas: (1) high costs and investment uncertainty, and (2) technical and infrastructure limitations.

In the high costs and investment category, the main factor is the significant expenses for hardware, software, setup, and skilled personnel. High initial investments and costs related to hardware, software, and professional personnel for maintenance and support create barriers [68]. This is particularly burdensome for SMEs with limited budgets [6,12,64]. Moreover, integrating these advanced technologies into existing systems often necessitates costly upgrades, further increasing the financial burden. For example, integrating XR into existing real-world production lines may require substantial upgrades to legacy systems [12], which are old or outdated computer systems or technologies. In addition, the inherent uncertainty surrounding the return on investment (ROI) for these technologies further complicates the decision-making process for many companies, making them hesitant to invest due to perceived risks [37,85]. Furthermore, the significant financing and technical support required for building the IM could lead to large enterprises dominating development, potentially fostering industrial monopolies [4]. This could further prevent SMEs with already constrained budgets and a lack of capital investment from adopting these immersive technologies, along with the uncertainties and risks involved.

In the technical and infrastructure limitations category, several key challenges are highlighted, including the demanding computational and storage requirements of immersive technologies and limitations in the current network infrastructure. Such challenges are highlighted by [74], where an IoT-based environmental monitoring system in manufacturing encountered packet loss under variable network loads and faced trade-offs between affordability and data stability. For instance, VR, DT, and metaverse applications require substantial processing power and storage, demanding continuous updates and expert involvement, which can be time-consuming and challenging for SMEs [6,54,62,68]. Furthermore, IM necessitates ultra-high network reliability, extremely low latency, and high throughput, which existing 5G networks may not fully support, thereby impacting critical applications like environmental monitoring, precise control, and immersive experiences [4]. Hardware design inefficiencies, such as inadequate headset brightness, can also hinder the immersion experience, leading to eye strain, reduced hazard recognition, and training ineffectiveness [43]. Likewise, the lack of haptic feedback hindering intuitive interaction with virtual objects [12] also impedes user adoption and overall effectiveness. Ensuring compatibility across diverse technology standards and integrated software tools is another significant hurdle for metaverse systems [4,64]. Moreover, human factors play a crucial role, as traditional skills may not directly translate to using innovative technologies, and there is a concern that workers’ natural instinctive responses to abnormalities might be delayed or ignored in virtual environments [4,7].

To conclude, lowering the implementation costs of metaverse technology is crucial for its adoption, especially by SMEs. Although immersive environments involve high upfront costs, these are justified by benefits such as reduced safety risks and increased efficiency [12].

4.4.2. Cybersecurity, Privacy, and Data Management

Digitally driven industrial systems face increasingly complex and evolving security threats as technology advances.

Table 12. Key barriers related to cybersecurity, data management, privacy, and ethical concerns.

Barrier	Sub-barrier	Description	Reference
Cybersecurity and System Vulnerabilities	Evolving and complex threats	- Advanced digital systems (CPS, DTs, IoT, XR) face expanding security concerns from equipment failures to software and control vulnerabilities.	[3,9,12,18,35,81]
	Cyber-attacks and attack surface	- Attacks (e.g., ransomware) compromise automated safety systems, risking workers and operations, and causing financial and societal losses. - Rapidly produced, insecure IoT devices increase the attack surface when integrated into large industrial systems.	[7,8,9,52,53,81,85]
Data Management and Protection	Data integrity and management challenges	- Complexities include lack of standardized protocols, unprotected remote configs, and integrating socio-technical systems. - DTs require high-quality, real-time data, which is difficult to collect and manage, leading to flawed predictions if inaccurate.	[4,9,15,18,33,53]
	Limitations of security measures	- Advanced systems like VR, XR, and DTs demand high cognitive effort, cause memory retention issues, and increase stress.	[4,9,15,18,38,53,81]
	Cloud and edge computing trade-offs	- Cloud computing risks data leakage and is limited by internet reliability - Edge computing offers decentralized processing to enhance security, but requires specific infrastructure	[2,4,9,12,18,34,72]
Human and Ethical Concerns	Employee privacy and surveillance	- Advanced technologies (XR, wearables, sensors) collect sensitive personal data (biometrics, emotions), raising ethical and privacy concerns regarding autonomy and misuse (internal and external)	[4,12,18,22,33,34,79]
	Psychosocial stress and trust	- Monitoring technologies can increase workplace stress due to pressure to share personal data and adapt to new digital environments, impacting employee trust and privacy.	[2,5,12,16,53,64,80]

outlines the various barriers to the adoption and secure operation of digitally driven industrial systems, categorizing them into (1) cybersecurity and system vulnerabilities, (2) data management and protection, and (3) human and ethical concerns. These advancements have broadened security concerns beyond simple equipment failures to encompass vulnerabilities in crucial operational layers, including process control, monitoring, encryption, and software systems [3]. The increasing sophistication and interconnectedness of digital systems, such as CPS, DTs, IoT, and XR technologies, pose escalating cybersecurity, privacy, and ethical challenges across industrial environments [9,12,18,35,81]. For instance, cyberattacks, such as ransomware, directly jeopardize automated safety systems, leading to unintended operations, compromised protective measures, endangering both workers and operations, and substantial societal and financial losses [7,8,9]. In addition, many IoT devices are produced rapidly without sufficient attention to security vulnerabilities, risks, and potential threats [3], and their integration into large-scale industrial systems heightens the attack surface, contributing to broader cybersecurity threats and compromising worker health and safety [53]. Hence, unauthorized access to industrial control systems can cause production halts, data leaks, and safety risks for workers in automated settings.

The second overarching barrier category, data management and protection, addresses the intricate challenges associated with ensuring data integrity, security, and efficient handling in smart manufacturing environments. Cybersecurity in this context is complex due to the lack of standardized protocols, unprotected remote configurations, and integration of diverse socio-technical systems. For instance, DTs are critically reliant on high-quality, real-time data, which is often difficult to collect, manage, and scale securely. The integrity of this data is very important, as inaccuracies can lead to flawed predictions and hazardous situations [6]. Cyber threats are commonly addressed through cybersecurity measures like data encryption, robust enterprise security architectures, authentication protocols, and blockchain technologies [15,18,53]. Although encrypting industrial data can prevent leakage and tampering, the encryption and decryption processes can delay data transmission, impacting its timeliness and stability, and affecting performance and real-time decision-making [4]. Furthermore, reliance on cloud computing for centralized data management introduces the risk of data leakage [4] and limitations in consistency and reliability [9]. On the other hand, edge computing offers a decentralized alternative to mitigate some of these concerns, but it requires a specific infrastructure, which can be challenging.

Finally, human and ethical concerns explore the profound implications of advanced digital technologies for the workforce. Beyond technical risks, workforce monitoring through advanced digital technologies also raises serious ethical and privacy concerns regarding employee autonomy and data use. The deployment of XR systems, wearable technologies, and intelligent sensors enables the collection of highly sensitive personal data, including biometric signals, emotional states, and motion profiles [4,12,18,22,79]. This raises serious ethical questions about employee autonomy, privacy, and consent, with concerns extending to the potential for corporate misuse of such data and vulnerability to external data breaches. For example, data in the IM is exploited by major corporations, with personal information and biometric data already exposed online, raising both privacy and ethical concerns [4]. Constant monitoring of workers can also induce psychosocial stress, particularly when employees feel pressured to share personal information or adapt to unfamiliar digital environments [53]. Thus, the confidentiality and appropriate use of the collected health data raise important concerns.

In short, data security poses a significant challenge and is a major obstacle due to perceived vulnerability to hacker attacks and challenges in managing data consistency and integrity. Ensuring respect for fundamental workers’ rights, including privacy, independence, and human dignity, is of significant importance. Therefore, robust security and privacy protocols, encompassing encryption and strict access control, are not only essential for maintaining employee trust and legal compliance but also for fostering a safe and secure production environment that prioritizes worker well-being, aligning with the principles of Industry 5.0.

4.4.3. Organizational and Human-Centric Resistance

The implementation of advanced digital technologies is frequently obstructed by internal organizational resistance and the inherent technical and systemic complexities involved.

Table 13. Key organizational, cultural, and human-centric barriers.

Barrier	Sub-barrier	Description	Reference
Organizational and Technical Resistance	Skill gaps and workforce readiness	- Digital transformation is hindered by a shortage of skilled workers, low digital literacy, and the rapid pace of technological change. - New technologies demand continuous upskilling, which can be time-consuming and lead to occupational exhaustion.	[4,5,16,36,37,53,61,85]
	Technical complexity and infrastructure demands	- Advanced digital systems (CPS, DTs) require deep understanding, complex modeling, integration, and robust infrastructure. - Challenges include data quality, computational demands, cybersecurity, human factors, and the need for new work environments.	[6,9,16,18,35,85]
Health and Safety Risks from Digitalization	New OHS risks from technologies	- Digital transformation introduces new OHS risks by altering work organization and execution. This includes physical and mental health risks requiring updated assessments beyond traditional frameworks.	[3,85]
	VR/AR hardware and user experience limitations	- VR/AR devices can cause health issues (motion/cyber sickness, nausea, headaches, dizziness, fatigue) and physical discomfort, limiting long-term use. - Heavy, uncomfortable hardware negatively impacts user experience	[8,12,42,43,45,54,57,66,68]
	Cognitive overload and mental strain	- Advanced systems and immersive technologies (VR/XR, DTs) demand high cognitive effort, leading to memory retention issues, cognitive overload, and increased stress. - New knowledge demands, revised work methods, and unclear human-machine interfaces contribute to mental and emotional strain.	[6,9,12,16,18,50,54,81]
Cultural and Employment Impacts	Fear, anxiety, and job insecurity	- Employees fear job loss, diminished autonomy, and increased monitoring - Automation may lead to unemployment in the short/medium term - DTs’ ethical and socioeconomic impacts (workforce displacement, job role changes) are underexplored	[5,6,7,8,12,16,43,85]
	Dehumanization and alienation	- Technological progress can lead to worker alienation and exploitation through machine-paced labor and continuous monitoring. - Reduced human interaction due to technology prioritization can cause feelings of devaluation, impacting mental health and inclusion.	[11,16,22]
	User discomfort and resistance to change	- Negative experiences with immersive systems (simulator sickness, confusion between virtual/real) hinder adoption. - Older workers may resist VR due to unfamiliarity and discomfort - Employees may resist cooperation due to stress and fear of job loss	[8,18,42,43,45,54,68]

provides a comprehensive overview of the organizational and human-centric resistance that hinders the adoption and effectiveness of advanced digital technologies in industrial settings. This resistance is not singular but multifaceted, arising from technical complexities, workforce unpreparedness, and health and safety concerns.

Organizational and technical resistance often arises when digital transformation advances faster than the workforce can adapt. New technologies demand new knowledge and skills, which workers may struggle to acquire quickly [5]. Similarly, a shortage of digitally skilled workers is a major challenge that can increase workload pressure (additional strain) on existing employees [85], especially older ones [36], who struggle to acquire the new knowledge and competencies required, leading to increased workload pressure and potential occupational exhaustion. This issue is further compounded by the technical complexity and infrastructure demands of advanced systems like CPS and DTs, which necessitate a deep understanding, complex modeling, and robust infrastructure, presenting considerable challenges for data quality, computational demands, and overall secure and efficient implementation [6,9]. Additionally, the limited grasp of these complexities by top management and their narrow focus on basic ICT and automation further impede enterprise-level transformation [37].

However, the integration of digital technologies introduces new OHS risks by altering traditional work organization and execution, thereby impacting employee health and safety. For example, VR and AR technologies, despite their training benefits, can induce physical discomfort and user experience limitations such as motion and cyber sickness, headaches, and physical strain from wearable devices, making them unsuitable and uncomfortable for prolonged use [43,45,54,57,68]. Additionally, it can induce anxiety from heightened virtual interaction, along with physical discomforts like dizziness, tiredness, and potential loss of balance with intensive use [8,12,66], negatively impacting the user experience [42]. This leads to the ineffectiveness of using VR and AR for safety training, posing a significant obstacle to the use of these technologies [43,68]. Moreover, the increased cognitive demands of advanced systems, leading to cognitive overload and mental strain, contribute to heightened stress levels, reduced memory retention, and mental fatigue among workers [12,43,50,54,81]. For instance, managing the heavy cognitive load from rapidly increasing data causes high-stress levels among employees [16], as well as the difficulty of wearing virtual headsets with safety helmets [66].

Culturally, concerns over automation, such as fear, anxiety, and job insecurity, pose significant barriers, with older workers especially affected due to their higher susceptibility to simulator sickness and stronger resistance to change. Some workers resist cooperation due to stress and fear of job loss [8]. Older, more experienced workers, who are often less familiar with modern visual technologies, are more prone to experiencing simulator sickness when using VR, acting as a resistance to its adoption [45]. These anxieties, coupled with a sense of dehumanization and alienation fostered by machine-paced work and continuous monitoring, can erode trust, lower morale, and weaken social support, ultimately impacting worker well-being and dignity. For example, the adoption of advanced technologies has led to growing concerns about dehumanization and unemployment as automation increasingly replaces industrial workers [7,16]. User discomfort and resistance to change also reflect broader resistance, where skepticism and negative experiences with immersive technologies, especially among those accustomed to traditional, hands-on approaches, hinder adoption, illustrating the profound human and ethical dimensions that must be addressed for successful digital transformation. For example, traditional professionals still prefer hands-on physical training methods and are cautious about the long-term effectiveness and side effects of VR/AR tools [45].

To better organize the wide range of barriers identified in the literature, Figure 14 introduces a tiered framework that categorizes implementation challenges into four levels: (1) Technology and Infrastructure, (2) Operational, (3) Human and Organizational, and (4) Strategic and Economic. Each level represents increasing complexity and broader organizational impact, offering a structured lens through which to understand the multifaceted obstacles to adopting DTs, IM, and enabling technologies to enhance manufacturing OHS. Each level captures a distinct set of challenges that must be addressed for successful adoption.

At the base, Technology & Infrastructure challenges are especially prevalent among SMEs, which often lack the digital maturity required for DT and IM integration. Inadequate IoT networks, limited real-time data capabilities, and outdated legacy systems contribute to fragmented data environments and hinder synchronization efforts, as highlighted in the literature. These foundational gaps create significant barriers to the deployment of responsive, real-time safety systems.

Moving up the framework, operational challenges involve the practical complexities of integrating advanced technologies into existing workflows. Cybersecurity risks, system instability, and high resource demands for XR and CPS are common concerns. Many older industrial setups also struggle with the scalability and latency requirements needed for predictive analytics, making seamless integration difficult without substantial upgrades.

At higher levels, Human & Organizational and Strategic & Economic barriers become more pronounced. Resistance to change, digital skill gaps, and cognitive overload from immersive systems are frequently cited persistent issues. Strategically, many organizations lack a clear roadmap for immersive safety technologies and are hesitant to invest due to high upfront costs and uncertain returns on investments. As a result, many initiatives remain in the pilot phase. Addressing these layered challenges requires a coordinated approach that combines affordable pilot projects, targeted workforce training, and strong leadership commitment to drive digital transformation in safety practices.

To overcome these adoption challenges and barriers, especially for SMEs with limited resources, budgets, and investment challenges, several affordable strategies can be used. For instance, cloud-based DTs eliminate the need for expensive on-site systems by offering scalable simulations through flexible, subscription-based access. Similarly, simple lightweight XR tools, such as AR on smartphones or tablets, provide immersive experiences without costly headsets. These are particularly effective for AR-based training and work instruction overlays. Furthermore, open-source platforms such as Unity and Unreal Engine allow flexible, low-cost development, making it easier and cheaper to build customized virtual training environments or monitoring tools without the need to pay licensing fees. This is also supported by [74], whose use of wearable and wireless sensor nodes illustrates practical deployment scenarios of IoT for environmental monitoring that align with the goals of OHS. Low-cost solutions are attractive but can suffer from latency, limited processing power, and unstable communication. Finally, starting with small pilot projects for single use cases like ergonomic risk assessment or VR safety training, that is, phased adoption strategies, also helps teams learn and build skills and confidence before scaling up. These approaches reduce financial and technical barriers, enabling organizations to explore the benefits of immersive technologies without facing overwhelming initial costs, while supporting the human-centered and adaptable goals of Industry 5.0.

In summary, successfully navigating digital transformation in industrial systems requires a holistic approach that moves beyond purely technical solutions. Addressing the complex interplay of scalability and cost factors, cybersecurity threats, data management complexities, and profound organizational, cultural, and human-centric resistance is crucial for fostering a truly secure, efficient, and human-centered industrial future. Addressing these challenges requires not only technical solutions but also organizational change strategies and policy-level support for digital safety innovation.

To conclude, the five RQs posed in this study are addressed across multiple technological domains, each linked to specific enabling technologies and supported by the synthesis in Section 4.1. RQ1 was examined through the application of DTs, IoT, and sensors, which enable real-time modeling of work environments and ergonomic risk assessment. RQ2 is addressed by exploring how the IM, along with immersive technologies such as VR and AR, supports immersive safety training and virtual collaboration. RQ3 focuses on the integration of DT, IM, IoT, AI, CPS, and COBOTS into a unified advanced digital ecosystem that enhances OHS outcomes using predictive intelligence. RQ4 highlights the benefits of immersive technologies in improving worker engagement and enabling real-time adaptation to dynamic environments. Lastly, RQ5 addresses the challenges associated with implementing these advanced systems, particularly those related to technological complexity, worker trust, and financial constraints. Together, these responses provide a comprehensive understanding of how emerging technologies can be strategically aligned to support intelligent, human-centric safety management in smart manufacturing and advance towards Industry 5.0.

5. Practical Implications

The findings of this study provide several useful practical insights for stakeholders in the manufacturing industry, emphasizing the transformative role of advanced digital technologies in improving workplace safety, training, and operational efficiency.

First, it highlights the shift from traditional reactive safety measures to proactive safety management through the use of DTs and IM technologies, which enable predictive simulations and risk assessments to identify and mitigate potential hazards before their occurrence. Secondly, immersive training solutions using VR/AR provide safer and more cost-effective alternatives to traditional training methods, especially for tasks involving high risk, by allowing workers to practice in realistic controlled virtual environments without exposure to actual real-world danger. Furthermore, the integration of IoT devices, wearable sensors, and CPS facilitates continuous, real-time monitoring of environmental conditions and human factors, enabling the early detection of unsafe situations and improving incident prevention. Additionally, AI-driven analytics derived from human DTs support ergonomic optimization by tailoring work environments and tasks to individual workers’ physical and cognitive capabilities, thereby promoting both safety and productivity. Finally, this study highlights the importance of scalability, recommending the adoption of lightweight, cloud-based XR platforms that reduce the cost and complexity of the initial implementation, making advanced technologies more accessible to organizations of varying sizes, especially SMEs with limited financial resources and digital maturity.

6. Conclusions, Limitations, and Future Work

6.1. Concluding Remarks

This study provides a comprehensive review of the evolving landscape of OHS in manufacturing through the lens of advanced digital technologies, particularly DTs, and IM technologies. The findings highlight a significant transformation in safety management, propelled by the integration of immersive technologies, real-time monitoring tools, and data-driven decision-making systems enabled by the Industry 4.0 and 5.0 paradigms.

DTs offer predictive insights and virtual simulations to proactively identify and mitigate risks. IM technologies enhance this by enabling immersive collaborative spaces for training and safety evaluations without real-world exposure. Together, these technologies foster a paradigm shift from reactive safety management to proactive and predictive strategies that prioritize human-centric design and well-being.

Moreover, enabling technologies such as IoT, AI, XR, CPS, and COBOTS have been shown to play critical roles in this advanced digital ecosystem, contributing to real-time hazard detection, personalized safety interventions, and seamless human-machine interaction. These innovations support the broader vision of Industry 5.0, which emphasizes sustainability, resilience, and worker empowerment in manufacturing environments.

However, despite their promise, the implementation of DT and IM technologies faces several barriers, including technical limitations, high costs, data security concerns, and organizational and human-centric resistance. Addressing these challenges is crucial for enabling large-scale adoption and realizing the full potential of smart, immersive safety systems.

To make these technologies more accessible, especially in places with limited resources, such as SMEs, it is recommended to employ scalable cloud-based DTs and lightweight XR systems. This approach will allow organizations to benefit from digital safety and training tools without the need for expensive equipment requiring large investments in infrastructure, making Industry 5.0 more inclusive.

6.2. Limitations and Future Work

While this study offers a comprehensive synthesis of the current literature on the integration of DT and IM technologies to enhance OHS in the manufacturing sector, several limitations must be acknowledged.

First, the study is based on the SLR approach, which is limited by its exclusive reliance on peer-reviewed literature and does not incorporate empirical or experimental data. Consequently, the findings are based on existing interpretations, case studies, and technological descriptions found in the literature, which limits their applicability in real-world industrial contexts. The real-world validation of DT models is still limited, as most studies often focus on theoretical frameworks and case studies, in contrast to extensive empirical testing. To address these limitations, future research should incorporate empirical case studies and cross-sector benchmarking to enhance the robustness and applicability of these insights.

Second, the selected studies spanned diverse technologies and geographic regions, resulting in considerable heterogeneity. Thus, the findings of this study should be interpreted with caution regarding their generalizability. However, the emerging technologies identified present promising opportunities for enhancing OHS, their applicability and effectiveness are likely to vary across organizational scales and geographic regions. This diversity complicates the generalization of the findings across all manufacturing settings. Moreover, some technologies, in particular, IM, remain in conceptual or early developmental stages, with varying levels of maturity and readiness across different sectors and countries.

Finally, the analysis primarily emphasizes the technological potential and conceptual framework of DT, IM, and other interconnected enabling technologies rather than quantifying specific OHS outcomes, such as reductions in workplace accidents, absenteeism, or ergonomic risks. Future research should build on this conceptual foundation by incorporating empirical case studies, user-centered evaluations, and longitudinal assessments over an extended period. Additionally, future research could also investigate the long-term psychological and physiological impacts of immersive training environments on workers.

Nevertheless, the integration of DT and IM technologies marks a pivotal advancement in enhancing the OHS in manufacturing. These limitations, while not undermining the relevance of the findings, highlight the need for further research to validate and expand the proposed advanced digital ecosystem in real-world OHS environments. This study lays a foundational framework for future research and practical implementation, encouraging further exploration of human-centered digital ecosystems that support safer, smarter, and more sustainable industrial operations and the transition to the Industry 5.0 vision.