**1. Introduction**

Manufacturing processes are becoming increasingly automated and connected within companies. This means that the engineer has to acquire new competencies related to planning [1] and process managemen<sup>t</sup> in VUCA (volatile, uncertain, complex and ambiguous) contexts [2], in both the manufacturing field and in the market. Due to these challenges in Industry 4.0, it is necessary to develop new competencies and improve existing competencies, for engineering students, as well as for professionals who have to design, manufacture, and manage interconnected smart products and processes. This requires the identification and development of training activities associated to the complex and creative characteristics of environments 4.0. These training activities are made possible by the potential of the interactions of digital enablers, and are integrated into the tasks to be carried out, whose main feature is the exchange of knowledge in real time.

Manufacturing systems 4.0 [3] have to be provided, not only with adaptability to VUCA contexts in their technological, economic, environmental, and social aspects, but also in a sustainable manner, so that, depending on the purpose established by the market [4], the system can co-evolve in a stable way. This entails a continuous evolution of the competencies associated to Operators 4.0, in order to deal successfully with increasingly complex and creative problems [5]. The aforementioned evolution of the competencies associated with Operators 4.0 gives rise to the interest in conceiving socio-technical cyber-physical manufacturing systems (SCMS), in which the processes and relationships between human and technological factors are integrated and can co-evolve, which is crucial in the management, development, and growth of smart factories and learning factories. This interaction of the human factor with machines and robots acquires major importance in these factories through interfaces based on cognitive and neurocognitive technologies.

The adjustment in the field of design and production managemen<sup>t</sup> between manufacturing systems and Operators 4.0 [6] can be supported by cyber-physical systems, and from the possibilities offered by key enabling technologies (KETs) and new frameworks. The aforementioned lack of adjustment provokes the need for engineers and technicians to be trained as managers of the digital transformation by updating their knowledge and competencies with online support through interfaces that enable connection in workflows. The digital transformation process constitutes a key element that places the human factor at the centre of Industry 4.0, by creating integrated and co-evolutionary systems that take into account the work environment and the marketplace [7]; these systems are herein labelled SCMS.

The foregoing constitutes a research potential in which contributions have been made that refer to general aspects of the organisation of Industrial 4.0 [8–10], and others aspects related to the integration of Operator 4.0 within the socio-technical systems of Industry 4.0 [11–21].

This paper focuses on the life cycle engineering of manufacturing systems for Industry 4.0 [22–24], and the potential of KETs and the variety required thereof for the integration of Operators 4.0 towards the growth and development of such systems. In this paper, a new framework called Design for the Human Factor in Industry 4.0 (DfHFinI4.0) is therefore proposed, which allows the human factor to be placed at the core of Industry 4.0 and is based on the conceptual frameworks of the connectivist paradigm, the law of requisite variety, and on activity theory. Its main contribution lies in the modelling of SCMS from the consideration of the relationships between the human and technological factors (equipment and information system). This consideration enables the transfer of routine and smart competencies from human operators to the technical systems and takes full advantage of the engineer´s talent by encouraging competencies of greater scope derived from the incorporation of KETs. All this brings added value to the managemen<sup>t</sup> of the engineering processes, technologies, and competencies required in the different phases of the life cycle of smart and learning factories, and ensures the adaptation of products and processes to market dynamics [25].

The goal of this paper involves responding to the problem of conceiving cyber-physical socio-technical manufacturing systems from socio-cognitive conceptual frameworks under the perspective of life cycle engineering, in which the growth and development of the talent of Operators 4.0 is made possible. This research objective, ye<sup>t</sup> to be put into practice, constitutes exploratory research, with a qualitative approach, and uses deductive methodology, in which the DfHFinI4.0 framework is formulated from conceptual frameworks identified with bibliographic review techniques [26], whose feasibility is explored in the inclusion proposal for the Purdue Enterprise Reference Architecture (PERA) architecture and methodology.

In the design and application of the proposed methodology, the following steps can be distinguished:


The conception of SCMS considering the perspective of life cycle engineering enables the work systems of the engineer and technicians to be configured as Operators 4.0, through the integration of all aspects associated with professional competence for their operational and e ffective e fficiency [27], which permits the gap between human and technological factors to be bridged. Regarding these SCMS, the navigation by the Operator 4.0, through the cyber-physical space, is managed by the interfaces, and the associated training actions are adjusted for the acquisition of higher levels of experience and professionalism. All this fosters progressive adaptation from the academic to the professional field and career development. The proposed DfHFinI4.0 framework can be integrated into various life cycle engineering methodologies of smart and learning factories. A case study for the PERA methodology has been implemented in cyber-physical systems and is configured as PERA 4.0.

The organisation of the paper is structured as follows: Section 2 contextualises and considers the aspects related to smart manufacturing, while paying particular attention to the cyber-physical systems (CPS), for its projection in the conceptualisation of SCMS. Section 3 describes the conceptual domains that are employed to obtain SCMS. In Section 4, the conceptual domains are articulated in a DfHFinI4.0 framework, which allows the configuration of integrated and co-evolutive SCMS, and establishes the relationships between its elements. Section 5 applies the proposed framework to the PERA methodology, thereby transforming it into PERA 4.0. Section 6 lays out the discussion and proposes future work. Finally, Section 7 presents the conclusions.

#### **2. Background of the Literature**

In this section, from among the possible types of reviews characterised by Mayer [26], a status quo review is carried out that entails a description of the state of knowledge in smart manufacturing, especially regarding cyber-physical systems in Industry 4.0. The content of the review has been graphically represented, as shown in Figure 1. The result of the review will allow us to characterise di fferent aspects of SCMS and the gap associated to the human factor.

**Figure 1.** Background of the literature organisation.

#### *2.1. Life Cycle Knowledge- and Technology-Intensive Industry (KTI) Manufacturing*

The Organisation for Economic Co-operation and Development (OECD) taxonomy classifies the industries into five groups (high, medium-high, medium, medium-low, and low), and includes all the manufacturing industries in the high and medium categories [28]. Knowledge- and technology-intensive industries (KTIs) are those that have a particularly strong link to science and technology, and are classified by the OECD taxonomy as high- and medium-intensive R&D industries. Aerospace, computers and office machinery, testing instruments, pharmaceuticals, motor vehicles, chemicals, machinery and equipment, business, communications, and education constitute knowledgeand technology-intensive industries (KTIs) [29].

The evolution of automation, given the possibilities presented by digital enablers, connectivity and artificial intelligence, has made possible the inclusion of knowledge in a wide variety of cyber-physical elements [15], by proposing areas of research in which the distribution of knowledge and intelligence between Operators 4.0 and technological solutions is produced, not only with the assistance of KETs [30], but also by cognitive capabilities [9]. Another area of research [31] of grea<sup>t</sup> significance for the present work is related to the personalisation of technology and occupational environments that use subrogate models of Operators 4.0 to parameterise the adaptation of the technology, thanks to their ability to conceptualise and consider them as another cyber-physical system of the SCMS.

Within the framework of Industry 4.0, the life cycle engineering of manufacturing systems, together with the concept of cyber-physical systems of technical equipment, includes the concept of Operator 4.0 [32]. Its differential features include creative intelligence and expertise in the domain of knowledge that constitutes the field of responsibility. Their modes of operation in Industry 4.0 environments are performed cooperatively with robots and machines, and with cyber-physical resources, and employ advanced human-machine interaction technologies and adaptive automation to achieve a suitable degree of symbiosis [11–33]. Collaborative robots enable the creation of shared work environments where productivity can increase while minimising delivery response time [34], and developing tasks cooperatively to solve open or complex problems under creative approaches that are representative of VUCA contexts. This provides evidence that environments based on the combination of human and technological factors can successfully tackle such contexts [35].

On the one hand, all the challenges that characterise the smart manufacturing process require continuous innovation and learning [36]. Distributed manufacturing is supported by a real-time operation planning system that controls manufacturing networks [37]. Manufacturing processes are becoming increasingly automated and connected within organisations [1]. Complexity and flexibility in manufacturing require analytics, efficient problem-solving, and process improvements. Business intelligence (BI) analysis graphs represent expert knowledge on analysis processes [38]. Manufacturing involves collaborative information exchange from several sources under different working conditions [39]. The concept of collective intelligence has been applied in engineering within the field of cyber-physical systems (CPS). Knowledge is used by automated problem-solving methods to coordinate and supervise manufacturing systems. Ontology can thereby play a major role in the process of creating and managing knowledge [40]. On the other hand, lean manufacturing is a managemen<sup>t</sup> model that focuses on minimising losses and optimising the creation of value for the client. Enke et al. [14] take into consideration the combination of lean manufacturing and Industry 4.0. Lean-based methodologies can improve organisational capabilities and tools to facilitate the transformation of a company into Industry 4.0 [41]. The concepts of lean manufacturing and Industry 4.0 can be developed in an end-to-end value chain for the Learning Factory to learn how to carry out a digital transformation [42].

#### *2.2. Industry 4.0 Features*

Among the characteristics of Industry 4.0 systems, boundaries between operation technologies (OT) and information technologies (IT) are disappearing [43]. Visualisation technologies, and fundamentally augmented [16], virtual and mixed reality are incorporated into production processes and training, since it has been proven that these constitute useful tools for Industry 4.0 [17,35]. Previously established

immersion technologies [44], as well as brain-computer interfaces and brain-machine interfaces, improve manufacturing systems [45] and Operator 4.0 performance [35].

Several models and definitions, such as reconfigurable manufacturing systems, smart factories, and ubiquitous factories, are associated with Industry 4.0 [46]. The complexity and flexibility in which companies and engineers have to operate in these environments require analysis [10], e fficient problem-solving, and process improvement. The use of business intelligence (BI) analysis charts represents expert knowledge for analysis processes and provides support in the work carried out by the engineer [38], from the engineering perspective of the life cycle 4.0, in cyber-physical systems design, development, and managemen<sup>t</sup> environments [47].

In Industry 4.0 systems, the customisation of products, processes, and services requires flexibility in the manufacturing and intelligence for smart products. It is also possible to enable the integration of ontology-based web services for flexible manufacturing systems [48]. Semantic web technologies can be used with cybernetic systems to integrate the decision-making process into smart machinery [49]. This allows automated decisions to be made, to help in the configuration of the manufacturing system from its representation of a virtual or digital twin. Therefore, ontology can play a major role in the process of creating and managing knowledge of cyber-physical systems [40].

One critical aspect of Industry 4.0 relative to advanced manufacturing is that of the availability of real-time information to optimally program the objectives of manufacturing systems throughout systems that have edge, fog, and cloud architecture [50]. Qu et al. [51] propose an ontology-based framework to represent a synchronised and station-based flow workshop, and develop a multi-agent reinforcement learning approach for optimal programming.

In order to apply simulation solutions that improve the e fficiency and profitability of Industry 4.0 systems, digital twins are created to describe the behaviour of the system. Stark, Kind and Neumeyer [47] consider the digital twin as the digital representation of a product, machine, service, product service system, or other intangible assets, that alters their properties, conditions, and behaviour, through models, information, and data. This concept is not only restricted to the operational part, however: it is also transposed to the human component [18] by generating a digital twin for the Operator 4.0 [31]. Recent developments in machine learning and other big data techniques o ffer new possibilities in conjunction with the concept of a digital twin [52] and subrogated models.

#### *2.3. Application in Manufacturing*

Relevant themes that have emerged as a result of flexibility, customisation, optimisation (saving time and costs), and smartisation, and those that require increasing connectivity include: predictive maintenance, which serves the objectives of sustainable manufacturing in the three dimensions of 3E; virtual commissioning [53] and crowdsourced manufacturing organisations [54]; real-time online support for production operators; new opportunities for servitisation; co-design; and co-manufacturing, cloud manufacturing and social manufacturing [55].

The role of Industry 4.0 maintenance, especially that of predictive maintenance, presents a strategic factor in manufacturing [56]. Techniques, such as forecasting, health and safety managemen<sup>t</sup> (PHM), and condition-based maintenance (CBM), create a demand for Operators 4.0 with adaptive interfaces that allow suitable characterisation and development of the required maintenance, with the necessary connectivity and the appropriate decision support [19]. The implementation of such techniques requires the use of the Industrial Internet of Things (IIoT) [40], cloud computing [57], big data [58], machine learning, and augmented reality [59,60]. In this respect, Cachada et al. [56] describe the architecture of intelligent and predictive maintenance to support Operators 4.0 by providing guided intelligent decision support.

Virtual commissioning, through the creation of a simulation model in a virtual environment of a manufacturing plant, allows Operator 4.0 to propose the necessary changes for its subsequent implementation in the real plant. However, today´s lack of competencies and associated experiences hinder the full integration of this tool in manufacturing [53].

Crowdsourced manufacturing organisations share their manufacturing resources based on their demand and capacity. Kaihara et al. [54] have developed a manufacturing simulation model in collaboration with a resource model and a negotiation algorithm based on cyber-physical systems to evaluate the e ffectiveness of manufacturing. This enables Operator 4.0 to cover the task of resource manager and requires interfaces capable of providing real-time feedback on shared resources and on any needs that may arise.

#### *2.4. Smart and Learning Factories*

The new professional engineer profiles that have emerged in Industry 4.0 environments require a suitable characterisation of the competencies to be acquired [20,61], in order to interact with smart manufacturing agents. To this end, the learning factories have been developed and are employed to instruct and train engineers through an approach between learning and professional practice, thereby contextualising Industry 4.0 environments. Furthermore, learning factories enable applied research to be carried out, both in engineering areas and in other areas of interest [8,21], and also foster collaboration between companies, students, and universities [62], with dual training models.

For successful operations in smart and learning factories, not only must training in technical competencies be taken into account, but also training in solving complex problems with uncertainty and decision-making in real time, under time pressure [63]. Special attention should be given to the acquisition of creative competencies, innovation, multicultural teamwork, and the ability to solve complex problems, all as enablers for their operation in VUCA contexts [61]. Research studies are currently being carried out in which transformation processes are developed in a manufacturing workshop, led by training in a learning factory, which involves instruction, integration, and engineering. This instruction is related to training strategies and objectives [64], in which serious gaming instructional techniques are incorporated, which can be employed to develop competencies related to new technologies in Industry 4.0 [65].

A smart factory [66] is made up of cyber-physical systems consisting of a physical part and an associated digital twin, with grea<sup>t</sup> possibilities for connectivity, intelligence, and data processing in the cloud and in the fog, and for operation with subrogate models [50]. These operational environments determine the need for tools that integrate the potential of the competencies acquired by the engineer with the use of technologies 4.0 [64]. In this context, Operator 4.0 constitutes a cyber-physical system with the possibility of multiscale and multilevel connectivity, with grea<sup>t</sup> analytical, calculation, and simulation capacity.

The development of learning factories, which allows the introduction of the concept of a digital twin and its applications, is necessary not only at the level of large companies, but also for SMEs (small and medium-sized enterprises) [13,67]. This leads to the integration of all the possible business fabric within the concept of Industry 4.0, whatever the size of the company.

#### *2.5. Research Gap*

Once the review of the most specific research parameters of Industry 4.0 and its manufacturing systems has been carried out, these parameters can then be characterised in a more detailed way as shown in Table 1, in order to establish a way in which to conceive the adaptive manufacturing systems of Industry 4.0, through the operation, growth, and development of human talent in VUCA environments. In this table, those papers that are most relevant for their contribution towards adaptive manufacturing through synergistic actions of processes, technology, and human factors in SCMS, characterise the research gap.

The contributions of those various studies whose objective involves the integration of technological enablers and their potential for adaptive manufacturing have been examined, both for exogenous changes (external client and contextual factors) and for internal changes (internal client and contextual factors). Identification has been made of the need for: conceptual resources and frames that allow the analysis of the productive activity located at any level of granularity; the search for allostasis (dynamic balance) between technology, processes, and Operator 4.0; and for the leverage of human talent and its affective coupling to manufacturing systems. In the following section, conceptual frameworks are presented that lead to the synergistic and adaptive manufacturing system framework proposed in the paper.


**Table 1.** Specific research parameters of Industry 4.0 and its manufacturing systems.

The levels considered are: (✗) minor level, (-) medium level and (✓) mayor level. D1: Modelling and simulation of SCMS. D2: Assistance in decision-making and navigation strategies. D3: Soft and hard skills. D4: Affective interaction. D5: Cognitive, socio-cognitive and cultural ergonomics. D6: Competence managemen<sup>t</sup> and talent development throughout the professional life cycle. 9

After this review of the literature, observation is made of the effort being carried out in the field of digital transformation for the application of key enabling technologies (KETs), as shown in Figure 2, and of the implementation of SCMS, which enables successful operations in VUCA environments. The preceding situation modifies the bases of productivity and the competencies required to adapt to this new situation, in which the life cycles of products, manufacturing technology, and knowledge are becoming shorter and more volatile.

**Figure 2.** Application context for DfHFinI4.0.

Life cycle engineering 4.0 must be equipped with tools to integrate experience and to project human talent in the growth and development of SCMS from the opportunities of the operational environment. To this end, it is necessary to identify a set of conceptual frameworks that enable tools to be derived: for the conception of the work in the SCMS environment, in the most complete and possible way; to integrate the various elements therein, both individually and socially; and to assign tasks to the human and technological factors so that they work collaboratively by establishing adaptive interfaces. For this purpose, Vigosky's activity theory [68,69] has been considered.

Complementary to the decision to identify a conceptual framework for the formalisation of work in all dimensions of its complexity, its conception is required to integrate elements and solutions that allow its adaptation and self-regulation, depending on the operational, cognitive and a ffective variety of the Operators 4.0 that can undertake the job, thereby allowing its coupling to the established work system. Ashby's law of requisite variety has therefore been selected [67,70], since it is integrated in the activity theory, which enables the self-regulation of tasks in accordance with the Operator 4.0 who develops it and with the a ffective connection required.

Finally, the further aspect of the human factor integration that requires special attention is that of the co-evolutionary development of the joint action of the socio-technical system which enables growth and development. This aspect provides online support strategies in the development of tasks and assistance in the growth and structuring of the experience acquired in the form of lessons learned, for their subsequent reuse in analogous situations, and consequently enables navigation strategies in the information system under semantic websites that allow the innovation and growth at both the individual and collective levels. To this end, the instructional framework of connectivism has been selected to derive strategies and tools [71,72].

Subsequently, these conceptual areas are structured in the proposed framework which is oriented towards the integration of the human factor within the environment of a smart company [73] which configures an SCMS, based on the search for the best available techniques, with the aim of seeking manufacturing excellence. All these techniques integrate, on the one hand, the potential of emerging conceptual frameworks; on the other hand, they integrate key digital and technological enablers, by merging the human and technological factors that correspond to the environment that produces them.

In the following section, the conceptual approaches and the KETs that will integrate the framework are presented, and an analysis is given of the most characteristic features that make them value drivers for the conception of the DfHFinI4.0 framework.

## **3. Conceptual Frameworks**

The adaptation by companies to the requirements of Industry 4.0 entails the appropriate training of both engineers and technicians [42] that is focused on handling problems of a greater complexity and open situations of a wider scope, as depicted in Figure 3. The acquisition, training, and improvement of these competencies, defined in accordance with the updating of the tasks to be carried out Operator 4.0 and their relationship with the other elements of the SCMS, enable professional profiles to be generated in accordance with the requirements in the smart and learning factories [74]. This represents a major training challenge, both in the academic and professional fields [75]. The engineer, as Operator 4.0, has to be capable of handling interoperability, virtualisation, decentralisation, service orientation, modularity, and key technology enablers (KETs) [76].

Along with the new demands of competencies required for workers, it is necessary to conceive the technology and associated processes of interaction with operators in Industry 4.0, as cognitive and socio-cognitive systems with a ffective connectivity. These characteristics make possible the integration and co-evolution of the SCMS formed by the human and technological factors under an organisational framework in the cultural context of a given company. Both factors, human and technological, are coupled [77] through adaptive interfaces that co-evolve as a cyber-physical system, together with Operator 4.0 and industrial equipment.

**Figure 3.** Smart Manufacturing requirements.

As shown in Figure 4., the potential from KETs triggers the search for a framework articulated in a toolbox that enables affect-socio-cognitive SCMS to be obtained, that respond to the different situations in which the engineer as an Operator 4.0 must be competent within an Industry 4.0 environment. This potential is articulated in conceptual frameworks derived from other areas of research: the theory of action [78]; the activity theory [79]; the required variety of Ashby for affective integration and coupling; and support in obtaining objectiveness and intentionality in the domains of cyber-physical entities through the connectivist paradigm:


The proposed DfHFinI4.0 framework requires the use of high-frequency trading concepts and technologies (HFTs) that are useful in modelling cyber-physical systems 4.0. A descriptive review of their state of knowledge based on Squires [80] is therefore proposed, which generates a proposal adapted to the field of smart manufacturing. In the following subsections, these various conceptual approaches are developed and articulated within this field.

**Figure 4.** Conceptual frameworks that configure DfHFinI4.0.

#### *3.1. Activity Theory*

According to Vigotsky's activity theory, the action of the engineer as an Operator 4.0, which is represented in Figure 5, takes place in a SCMS [81], where the engineer with the interface solves problems or, with the help of a tool, transforms the raw material or the product [82] as a result of its application. The interaction between the human factor and machines and robots is mediated through interfaces [12], whereby the latter constitute the tools within the activity system [83]. The adoption of Vigotsky's activity theory as a unit of analysis for containing all its elements involved in work situations was reformulated by Yrjo Engeström [84]. This author has considered its evolution through three generations of research. Its use is especially significant as an instrument to model the activity and knowledge, that, in Industry 4.0, is largely transferred from technology to the human factor.

In the second generation of activity theory, Engeström expands the Vigotsky triangle to represent the collective elements of the activity system. From the Engeström model [85], it is possible to develop a general structure of the activity based on Operator 4.0, machines, robots, work equipment, interfaces, rules, and division of labour, as well as on the knowledge implicit in the technology (tool) and in the operator, and even in the evolution of their competencies as a consequence of the development of work. The rules and regulations, explicit and implicit, define the course of action to accomplish the task. The tasks are carried out according to the organisational structure of the company, through a division of labour. In the third generation of activity theory, tools are developed for dialogue, diversity of perspectives and networks of interacting activity systems. Modelling operational activity with activity theory enables the analysis to be performed with different granularity [86], and also the incorporation of the analysis of the different dimensions of the elements for the study of their integration and of the effect that arises from articulated solutions.

**Figure 5.** Activity theory and Law of requisite variety to model characteristics of Operator 4.0.

This theory has been applied, among other areas, in training organisation [87], human-computer interaction (HCI) [88], and information systems (IS) [89], since it has been proven to be a highly useful tool in the establishment of the way in which ICTs and other technologies interact with their context [90]. In this system of analysis of the proposed activity, the operator and the technological and organisational elements are interrelated [91], thereby enabling a holistic analysis of the socio-technical system. Its articulation in engineering has grea<sup>t</sup> potential for engineers to adapt to the requirements of smart and learning factories [92], when it comes to modelling and responding to training and co-evolution needs [93,94]. All this reduces the static and dynamic complexity in manufacturing systems regarding cyber-physical systems.

Activity theory [95] facilitates interactivity and assertive navigation, and encourages the engineer's creativity as an Operator 4.0, in the resolution of open and complex problems. This is made possible by enhancing not only cognitive competencies [96], but also affective coupling, since the latter enables better responses to be made to possible feedbacks, thereby facilitating the resolution of problems in collaborative work. In this respect, the same objective as that which the video game industry [97] demands from designers can be applied to smart manufacturing; an innovative design that attracts the attention of the internal customer (Operator 4.0), where there is constant feedback at the cognitive and affective level, which encourages constant engagement. This can be employed to design interfaces and the work system as a whole, with the aim of promoting its use through motivation [98], affectivity, dependency, and feedback generated by the engineer.

#### *3.2. Law of Requisite Variety*

The work analysis requirements focused on the explicit and implicit knowledge that cyber-physical manufacturing systems must support, both for human and technological factors, to justify the incorporation of Vigotsky's activity theory as a tool to formalise the elements that integrate a work system, both in its individual and social dimension, with the aim of obtaining socio-cognitive manufacturing systems. By taking this theory into consideration, it is possible to identify various elements of the work and their features. Such is the case of Operators 4.0 and their associated competencies, the tools, products and problems on which they take part, the organisation and specialisation required for the work accomplished, in addition to the associated operational culture. Nevertheless, the aforementioned analysis suffers from the inclusion of mechanisms to identify, among others, sensory, cognitive, and affective capacities, and experience of the Operators 4.0 that will be

assigned to a workstation and the required demands of the work system. This situation considers the need to identify a conceptual framework that enables the characterisation of the variety of competencies of workers and the work system demand with which they interact, and to determine their discrepancy and variety adaptation mechanisms. The aforementioned mechanisms and the a ffection that comes from their implementation must be represented in the activity theory for the articulation of mechanisms that allow the adaptation of the required variety of the technological system to the variety established in the worker.

In relation to this, given the need for a conceptual framework to adapt variety on the structure of the elements of the activity theory, it is worth considering Ashby's law of requisite variety [67]. This law establishes that a regulator-regulated system is one that is made up of a regulatory subsystem that exercises its action based on the information collected from the regulated subsystem [70]. In this kind of system, the regulatory part (work station, machine, robot, process) must have at least the same variety as the regulated part (Operator 4.0 or work equipment 4.0) so that the system reaches stability, which necessitates the establishment of a one-to-one correspondence between the varieties on each side. In order to achieve this adjustment, adapters are employed to reduce or amplify the variety, depending on what is required [99]. This adapter is itself a system or part of a system, and can act in either direction, increasing the variety by means of amplifiers or reducing it by means of reducers, until the regulator and regulated subsystems reach the same variety and determine an a ffective occupational use experience.

The adjustment of the collaborative environment to the engineers as Operators 4.0 [100] can be carried out by taking into account the law of variety (law of requisite variety), and will depend both on the constraints of the tasks to be performed [101] and on the cognitive [102] and a ffective [103] features possessed by the engineer and the technicians who operate in Industry 4.0, as illustrated in Figure 5 above.

Regarding the adaptation elements of the variety, in the context of Industry 4.0, it is necessary to consider the characteristics of the Operator 4.0 tasks, limited to situations of complex, creative problem-solving, with uncertainty and deadlines. This requires online support regarding knowledge, embodied in Operators 4.0, from the company's knowledge bases and other stakeholders in the project. All this determines that, as adaptation strategies of the variety of technology 4.0 (through interfaces, mobile devices, tablets, and wearables), the potential of KETs and the managemen<sup>t</sup> of the variety required are utilised, so that the navigation strategies provide the online support required by the Operator 4.0.

As mentioned earlier, this enables the personalised occupational activity of Operators 4.0 to be modelled, who, in their implementation of achievements, determine the support of navigation strategies and of the dynamic managemen<sup>t</sup> of the operational requirements, of their learning and improvement of competencies and of the reconfiguration of the technological environment 4.0 for the variety required according to the designated worker. In order to answer these questions, it is considered as a conceptual framework to manage the variety in the navigation strategies, learning support, and systematisation of lessons learnt in the connectivist paradigm [104], which will be presented with KETs.

## *3.3. Connectivist Paradigm*

In the field of smart manufacturing, connectivity acquires major importance in a cyber-physical environment, with the hybridisation of the physical and digital world, and with artificial intelligence and knowledge, not only in the cloud, but also integrated into production, equipment and tools in the edge. This has resulted in Operator 4.0 competencies moving towards complex and open creative, social, and problem-solving competencies, with opportunities for continuous learning. Support in the navigation strategies of Operator 4.0 is necessary for the resolution of problems; it is here where the connectivist theory of Stephen Downes and George Siemens intervenes, which a ffects the process of connecting specialised knowledge [105,106], under the theoretical principles of the Clark and Chalmers concept of the extended mind [107].

A crucial component of this connectivity falls on the semantic web as the informational dimension of the extended mind [108]. On the one hand, the semantic web is an intelligent entity that generates, shares, and connects content, capable of being interpreted by operators, machines and robots, to work collaboratively through specific languages such as XML, RDF, RDFS, OWL. It enables the articulation of connectivist schemes for the online support of navigation strategies in problem-solving, learning, collaboration, and systematisation of the experience. This semantic web enables the meaning of this data to be interpreted in a similar way to that of the semantic analysis of the operator's language and therefore to support the workers in their various tasks. It provides communication content between Operator 4.0 and machines and robots with semantic content, and allows the information to be processed based on a semantic assessment of its content, which permits its best coupling through interfaces so that it can be optimally interpreted by the engineer [49]. On the other hand, the use of ontologies in the semantic web enables the correct identification of the meaning of instructions according to a given situation and context [109]. Together, these facilitate the establishment of interactive and navigation strategies to support problem-solving and to enable the learning process and the improvement of competencies [110].

Connectivism integrates the IIoT, cloud computing, and virtual reality technologies, among others, which enable connection, accessibility, and data sharing [111], as depicted in Figure 6. The engineer has to perform tasks, both training and professional, by collaboratively exchanging information in real time with machines and robots, in order to adapt to different working conditions [39]. Furthermore, connectivism seeks to create collaborative environments, connected not only between Operator 4.0 and machines and robots, but also between end customers, vendors, suppliers, and all those agents involved [112]. These strategies, in the form of instructional knowledge and subrogate instructional models, will be carried out from the cloud.

**Figure 6.** Connectivity in the Smart and Learning Factories.

Connectivism is based on theories of neural networks, chaos, complexity, self-organisation, and non-linear systems, from which navigation strategies are structured. Due to these influences, it integrates tools that increase, from connectivity and virtuality, the ability to facilitate interaction between the elements that shape the smart and learning factories. Connectivism has been described as the amplification of knowledge and understanding through the extension of the network, and is called the theory of knowledge for the digital age [106].

Thanks to these characteristics, the connectivist paradigm has been established as the basis for the cyber-physical socio-technical system, since it facilitates the real-time network support of cyber-physical systems as part of DfHFinI4.0, which allows constant and dynamic network workflow and training.

The navigation strategies, in conjunction with a variety of filters and amplifiers, are established in an integrated manner regarding the activity theory, which in turn, as support for activities of less granularity, constitute fractal elements [113] of top-down and button-up analysis of the SCMS for Industry 4.0, and, through the KETs, determine its greater significance by acting as lever arms.
