Defining stages of the Industry 4.0 adoption via indicator sets

A keyword search was performed in four core databases to gather relevant articles in the literature, which potentially discusses the indicator system for I4.0. As the development of I4.0 is yet an emerging domain and is still at its early stages (Issa et al., 2018), the search was expanded from I4.0-specific applications to the general technology and innovation adoption. The primary keywords used were: “digital transformation”, “industry 4.0”, “industry 4.0 adoption”, “innovation”, “innovation adoption”, “technology”, “technology adoption”; together with supplementary keywords such as “indicators” and “predictors” The study used the following databases: Elsevier's ScienceDirect and Scopus, Taylor & Francis’ www.tandfonline.com, and Springer's SpringerLink. To reach a comprehensive coverage of publications related to the indicators of the I4.0 adoption, journal articles, and conference proceedings were also obtained from these databases. In the following step, a content analysis was performed to extract prospective indicators on identified articles. Articles that do not ultimately provide related indicators were excluded.

A comprehensive list of indicators was generated from a variety of innovation studies, having numerical metrics as part of their methodology. However, in the context of I4.0, no present study was able to develop a set of indicators to assess the I4.0 implementation at different stages. Thus, a significant challenge was to select an appropriate set of indicators from the general system of innovation and technology adoption. Addressing the challenge, several criteria were used for the selection and construction of indicators. Four criteria were used in the screening process to select appropriate indicators. Miremadi, Saboohi and Jacobsson (2018) developed general criteria for the selection of indicators in the context of innovation systems. They used this set of criteria as it covered approximately all factors in the relevant literature (Miremadi, Saboohi & Jacobsson, 2018). To measure the I4.0 implementation, indicators must be understandable, available, relevant, and measurable. First, an indicator is considered understandable if it is straightforward, simple, and provides ease of understanding. Second, an indicator is available if data and information are accessible. Availability ensures that the value of a specific indicator is obtainable from a company's information system. Third, indicators are deemed relevant if they satisfy the goal of assessing the level of the I4.0 implementation and if they point to the characteristics or nature of activities per stage. Fourth, indicators must be measurable following an existing scientific measurement approach (e.g., surveys).

An initial list of indicators was generated from a literature review on a variety of technology and innovation adoption applications, as shown in Table 1. In this work, an indicator was defined as a source of information, from which problems could be detected in the application of innovation (Borras & Edquist, 2013). A total of 469 indicators were collected. At the outset, these indicators contained literal redundancies of terminologies in different sources. Consequently, such redundant indicators were excluded, and this process yielded 90 candidate indicators. Afterwards, an appropriate description of each indicator was provided, indicating primary sources, from which they were extracted. In cases of insufficiency, supplementary or secondary sources (i.e., related journal articles and scholarly books) were used. From the initial list of indicators with descriptions, specific terms were found to be synonymous. Indicators implying a synonymous meaning were treated as redundant, thus, excluded. Focus meetings were then conducted to qualify a final list of indicators.

Descriptions of each indicator were carefully assessed. Each indicator was then assessed using the four criteria, focusing on its understandability, availability, relevance, and measurability. Subsequently, following the process of a thorough assessment, indicators that did not meet the four criteria were rejected. This process generated a final list of 62 indicators of the I4.0 adoption. An initial list was then categorised according to stages of adoption from initiation, adoption-decision, and implementation (i.e., pre-adoption, adoption, and post-adoption, respectively) (Hameed et al., 2012). These stages were an essential determinant to reflect the entire innovation process and to control the applicability of each indicator (Table 2).

Table 1 presents the number of extracted indicators, their application, and sources to provide an overview of the initial listing of I4.0 indicators used in this paper. The first column indicates authors from whom candidate I4.0 indicators were extracted. Papers listed under the label “reference indicators” indicate sources used to collate respective innovation or I4.0 indicators. Hence, “reference indicators” are the sources of performed compilation. For instance, to explore and discuss I4.0 technologies, Lu (2017) collated I4.0 indicators from Jazdi (2014), Stock and Seliger (2016) and Wang et al. (2016), among others. Moreover, the second column comprises the field of application used by a corresponding author (e.g., Chor et al. (2014), Lu (2017), and Hameed et al. (2017)) to demonstrate roles of indicators. The third column displays the number of indicators extracted from corresponding works to comprise the initial list of I4.0 indicators in this paper. The information displayed in Column 3 (Table 1) demonstrates that most of the extracted indicators came from the general innovation literature. For example, 116 indicators were extracted from Chor et al. (2014). The result stems from the attribution of I4.0 to innovations in the current literature, as pointed out by Morrar et al. (2017), Liao (2017), Brettel et al. (2014), and Almada-Lobo (2016), among others. Based on such claims in the literature, placing general innovation indicators in the context of I4.0 is validated by its innovation status.

To illustrate the use of the developed indicator system to assess the I4.0 implementation, case studies involving two manufacturing firms in the Philippines were conducted. The developed indicator set for each stage of the I4.0 adoption intended to assess the degree, to which a firm was positioned, given its current I4.0 implementation. By using the indicator sets, this work offered a general methodological approach, which attempted to generate the value indicating the performance or maturity of the firm at any given stage. The performance or maturity value, now denoted as a general index, provides a snapshot of the performance of the firm at an I4.0 adoption stage at a given time. Note that the quality of this snapshot is highly dependent on the completeness and quality of the information used in the evaluation process, and the level of information granularity of a specific applied methodology. The methodological framework starts with the assignment of weights for the indicators of a given stage. Weight assignment could be carried out using multiple criteria decision-making methods (e.g., analytic hierarchy process, best–worst method, simple average weighting), expert opinion, Delphi method, group decision-making techniques, etc. Once the appropriate I4.0 adoption stage is determined for the firm, the performance or maturity of the firm is assessed against each indicator of the stage, using a specified evaluation scale. The results of the second process are the performance values of the firm for all indicators. Then, using a specified aggregation technique, these performance values are aggregated homogeneously. The aggregation process provides a general dimensionless index that describes the overall performance of the firm at a given stage. Although this work offers the general methodological framework, the specific methodology that embodies the framework is left at the discretion of the firm or its analyst. A detailed procedure in assessing indicators under each I4.0 adoption stage is presented as follows. Note that the proposed procedure is recommendatory, not absolute. A thorough analysis of the most appropriate methodology that maximises the quality of information used in the evaluation process is out of the scope of this work. The following procedure is presented to demonstrate the use of the proposed indicator system in a real-life application:

Step 1: Attain the performance of the indicator. Industry experts firstly identify the current stage of adoption (i.e., pre-adoption, adoption, and post-adoption) and further elicit their judgment on each indicator's performance with respect to the perceived adoption stage, using linguistic scales “very poor”, “poor”, “fair”, “good”, and “very good”, whichever is applicable.

Step 2: Translate the performance of the indicator into a numerical value, according to Xu (2006). For a given linguistic set, S, a corresponding s_α, the numerical value is attained for each indicator as in (1), (1)S={sα|α=−t,…,−1,0,1,…,t}S = \left\{ {{s_\alpha }|\alpha = - t, \ldots , - 1,0,1, \ldots ,t} \right\} that is (2), (2)S={s−2=very poor, s−1=poor, s0=fair, s1=good, s2=very good}S = \left\{ {{s_{ - 2}} = {\rm{very}}\,{\rm{poor}},\,{s_{ - 1}} = {\rm{poor}},\,{s_0} = {\rm{fair}},\,{s_1} = {\rm{good}},\,{s_2} = {\rm{very}}\,{\rm{good}}} \right\} correspondingly, these indicator indices will be used to obtain the overall performance of firms at a particular I4.0 adoption stage.

Step 3: Define the overall performance of a firm as regards the I4.0 adoption. The indices of previously generated indicators are then aggregated as shown in (3) to arrive at a general index on the performance of firms as regards the I4.0 implementation, (3)LWAAω(sα1,sα2,…,sαn)=ω1sα1⊕ω2sα2⊕…⊕ωnsαn=sα¯{LWAA_\omega }\left( {{s_{\alpha 1}},{s_{\alpha 2}}, \ldots ,{s_{\alpha n}}} \right) = {\omega _1}{s_{\alpha 1}} \oplus {\omega _2}{s_{\alpha 2}} \oplus \ldots \oplus {\omega _n}{s_{\alpha n}} = {s_{\bar \alpha }} Where α¯=Σj=1nωjαj, ω=(ω1, ω2, …,ωn)T\bar \alpha = \sum\nolimits_{j = 1}^n {{\omega _j}{\alpha _j},\,\omega = {{\left( {{\omega _1},\,{\omega _2},\, \ldots ,{\omega _n}} \right)}^T}} is the weight vector of the s_αj (j = 1,2, ..., n) and ωj∈[0,1], Σj=1nωj=1{\omega _j} \in [0,1],\,\sum\nolimits_{j = 1}^n {{\omega _j} = 1} , sαj∈S¯{s_{{\alpha _j}}} \in \bar S , then LWAA is referred to as the linguistic weighted arithmetic averaging (LWAA) operator. In the case of this paper, the weight vector of the s_αj is assumed to be equal, thus, the average of s_αj is given by (4), (4)LWAAω(sα1,sα2,…,sαn)=Σj=1nsα1,sα2,…,sαnn{LWAA_\omega }\left( {{s_{\alpha 1}},{s_{\alpha 2}}, \ldots ,{s_{\alpha n}}} \right) = {{\sum\nolimits_{j = 1}^n {{s_{\alpha 1}},{s_{\alpha 2}}, \ldots ,{s_{\alpha n}}} } \over n}

It can be seen that the first question stated previously underlies the pre-adoption stage. The pre-adoption stage involves all other conditions needed for the adoption of innovation before the evaluation of the decision is made to adopt or reject the innovation (Miranda et al., 2016). As Rogers (1995) puts it, the pre-adoption stage involves the previous conditions for adoption, knowledge or awareness of the innovation, and the perception of the potential adopter by acquiring more profound knowledge on the innovation. Inherently, adoption decisions usually come from information acquisition periods, which is implicit in technological innovations (Dimara & Skuras, 2003). In the relevant literature, several scholars infer that the lack of awareness of innovation may explain the reason why its widespread adoption does not occur (Dimara & Skuras, 2003).

The degree of awareness of I4.0 can be acquired depending on the attitude of stakeholders towards the innovation. One of the reasons why attitude plays a vital role in determining the degree of awareness is because it dictates the optimism or pessimism of potential adopters (Kerschner & Ehlers, 2016). One of the most straightforward indicators that show the awareness of I4.0 is the perception of the term “Industry 4.0” (Priyadarshinee et al., 2017). Although the literature offers no strong support to the way the term used to describe an innovation affects its adoption, several papers consider the name significant in creating different perceptions or self-concepts (Garwood et al., 1980). On the other hand, perceived usefulness, perceived ease of use, relative advantage, trialability, observability, compatibility, and complexity, are indicators strongly supported in the literature and related to the perception of potential adopters (Rogers, 1995).

For instance, the perceived relative advantage is considered a sine qua non or necessary for adoption (Greenhalgh et al., 2004). If users do not perceive a relative advantage of innovation, it is generally not adopted (Rogers, 1995). However, although considered to be critically important, it does not guarantee widespread adoption, thus, suggesting the need to look into other factors (Greenhalgh et al., 2004). Trialability — the extent, at which an innovation can be tried on a limited basis — is also strongly supported by many scholars in the literature (Miranda et al., 2016). It is strongly argued that although being able to test the innovation on a smaller sample space does not guarantee success when applied at a larger scale, it increases the confidence of adopters in the innovation (Plsek, 2003). Rogers (1995) argues that it is positively linked to the adoption of an innovation.

Aside from directly testing the innovation, observing the innovation already adopted by others may also affect the perception of potential adopters (Miranda et al., 2016). Potential adopters use a risk-reduction strategy of seeking information from others who have already adopted the innovation of interest because adopters are usually faced with uncertainty about the consequences of their decisions, which contributes to their perceived risk (Mehrad & Mohammadi, 2017). It must be noted that trialability and observability are different. Trialability involves direct testing of innovation, and observability involves indirect testing of the innovation through others who have adopted the innovation.

Similarly, another important indicator of innovation adoption is compatibility or the degree of how accustomed an innovation is to existing standards, norms, and values of potential adopters (Greenhalgh et al., 2004). Many scholars argue that compatibility may significantly predict whether innovation will be accepted or not (Greenhalgh et al., 2004). For instance, if a government agency intends to make their citizens use services online, they must provide information and services in a manner that is consistent with other ways citizens have dealt with the government, e.g., online forms should resemble paper forms that citizens are familiar with (Rogers, 1995; Joia et al., 2016). If innovation results in actions that are very different from the existing practice, potential adopters perceive it as risky, thus, possibly rejecting the innovation (Joia et al., 2016). Likewise, if innovation is perceived to be highly sophisticated, potential adopters will likely reject it (Agarwal & Prasad, 1998; Greenhalgh et al., 2004). A complex system is one that cannot be broken down into manageable parts. Several scholars claim that most organisations that opt for innovation operate in such a manner if not appropriately managed; thus, they may cause a negative perception for potential adopters (Szczerbicki, 2008). To this end, complexity plays a crucial role in indicating the perception of potential I4.0 adopters.

Several scholars also point out the importance of communication channels used in spreading information about the innovation, which contributes to the perception of potential adopters (Agarwal & Prasad, 1998). This result was found to be significant by scholars in relevant fields (Adegbola & Gardebroek, 2007). Mainly, Adegbola & Gardebroek (2007) found that when the information about the innovation was spread through external sources (e.g., knowledgeable external sources) adopters tended to have a more favourable perception of the innovation than when it was spread through internal sources (e.g., adopters who were still in the process of adoption). Hence, the perceptions of potential I4.0 adopters can be affected by the flow experience of I4.0, the emergence of global distribution networks, and some information sources.

The pre-adoption stage is concerned mainly with perceptions of potential adopters, which result from acquired awareness and more profound knowledge of I4.0. Technological innovativeness, unlike the indicators mentioned above, is more of an inherent characteristic of potential adopters rather than one generated as a result of acquired awareness. Moreover, it plays a vital role in connecting the pre-adoption (perception) indicators to the adoption (persuasion) indicators of I4.0 (Agarwal & Prasad, 1998).

The adoption stage encompasses the period, in which the decision unit is engaged in activities that lead to the choice to adopt or reject innovation, otherwise known as the decision stage (Miranda et al., 2016). In contrast to the pre-adoption stage, the adoption stage involves the persuasion phase of the organisation to decide if innovation must be adopted or rejected (Miranda et al., 2016). In other words, it comprises activities (e.g., financial, technical, and strategic) that evaluate the readiness of systems to implement I4.0 in an organisation (Hameed et al., 2012). This section discusses the adoption stage indicators.

Financial evaluation activities are a straightforward indicator of the readiness for innovation (Quevedo et al., 2017). It is common practice for managers to evaluate the risks of innovation projects, mainly financial risks, since they usually may cause the failure of some innovation projects (Pellegrino & Savona, 2017). Also, risks, costs, and uncertainties are weight against benefits and incentives that would be gained by the organisation that implements innovation (Chor et al., 2014). Such activities are practical ways used by organisations to analyse the desirability of innovation (Prest & Turvey, 1965).

Several scholars point out the importance of leadership and support in the decision to adopt innovations (Greenhalgh et al., 2004). In particular, this mostly relates to CEO advocacy (Chor et al., 2014). The alignment between innovation and prior organisational goals makes the adoption more likely (Greenhalgh et al., 2004). Some scholars maintain that innovation adoption is more probable when key individuals (e.g., CEO) are willing to support innovation in their social networks (Greenhalgh et al., 2004). Likewise, the organisation must also have the technical capability to evaluate innovation (Boh et al., 2014). To this end, the presence of technical support and expertise, as well as IS infrastructures, are important indicators of an organisation evaluating its readiness for I4.0 (Hameed et al., 2014).

As already mentioned, an organisation must be capable of the successful adoption of innovation. However, not only it needs technical capability but also the expertise to deal with its market environment (Zhang & Hartley, 2018). As such, both the level of competitive pressure and the level of customer interaction are useful indicators for the adoption decision (Priyadarshinee et al., 2017). Several scholars maintain that the level of competitive pressure is an implicit consequence of the accelerated competitive environment, primarily due to the desire to create new products and processes in an improvised manner (Zhang & Hartley, 2018). Subsequently, scholars claim that the level of customer interaction is a critical determinant of the organisational performance; thus, also important for the adoption decision (Manral, 2010).

The geographic location of an organisation also plays a vital role in indicating the status of the adoption decision. Several scholars claim that geographical proximity is a necessary condition in the diffusion of knowledge or innovation (Martinez-Noya & Garcia-Canal, 2017). It is, thus, widely accepted that determining the geographic location of an organisation is a first-order strategic decision of stakeholders (Escuer et al., 2014). On the other hand, the reputation of suppliers is an important indicator at the decision stage. Since I4.0 promises new products and process innovations, potential adopters ensure that during the implementation stage, they would not face the risk of knowledge leakage. Knowledge leakage becomes a potential problem in at least two ways, i.e. (i) when suppliers serve a competitor who puts the organisation at the risk of knowledge spill-over in favour of the competitor, and (ii) when the supplier becomes a potential competitor due to the knowledge spill-over (Martinez-Noya & Garcia-Canal, 2017). Moreover, scholars argue that such risk is higher at locations with weak intellectual property (IP) protection.

The post-adoption stage occurs when a decision unit puts the technology in use (Rogers, 1995). That is, a decision unit, as in an organisation, finally implements the technology and correspondingly evaluates the advantages and disadvantages of technology adoption, which in turn, guides organisations in their decision of whether such adoption should be continued or not. Such action boosts the efficiency of an organisation, given its successful application of new technology in the local context as well as its capability to compete according to strategies and action plans (Oyemomi et al., 2019). This adoption decision, however, is subject to a certain degree of integration depending on the available resources and risk aversion of an organisation (Hameed, 2012). When an implemented technology is perceived to be riskier, an organisation's willingness to continue the adoption may correspondingly diminish.

In the case of the I4.0 adoption, it is a basic necessity to ensure open access to critical technologies, such as IoT, CPS, smart manufacturing, and cloud computing (Priyadarshinee et al., 2017). When organisations have a sense of ownership stake in one or more critical I4.0 technologies, the eventual post-adoption of such technology becomes more attractive. Otherwise, the acquisition of these technologies may potentially delay the I4.0 implementation on the grounds of economic risk barriers, financial leverage, and functionality of service quality offered by the new technology amidst the competition in the supply chain network (Joachim et al., 2018).

Also, the amount of operational cost reduced in the implementation of I4.0 must be deemed reasonable for organisations to continue using the technology (Lee et al., 2015). When representations of cost reduction as in pay-per-usage and reduced facilities are substantial enough to warrant a continued adoption of technology, organisations can be further driven to do so. The same inference can also be drawn for economies of scale (Priyadarshinee et al., 2017) and work simplification (Jeyaraj, 2006).

An organisation that implemented I4.0 is expected to have a system that is adaptive by plug-and-work mechanism (Monostori et al., 2016), autonomous (Letia & Kilyen, 2018), decentralised (Terziyan et al., 2018), dependable (Alguliyev et al., 2018), interoperable (Priyadarshinee et al., 2017), capable of real-time operations (Terziyan et al., 2018), remote monitoring and control (Sung, 2018), resilient (Chang et al., 2016), robust (Alguliyev et al., 2018), capable of handling full information (Sabherwal & King, 1991), flexible (Browne et al., 1984), capable of maintaining data lifecycle (Tao et al., 2018), usable (Priyadarshinee et al., 2017), and with intelligent lots (Li et al., 2012). Given that such capabilities are made possible by critical I4.0 technologies, an organisation is more likely to confirm the continued adoption of such technology, based on trials when its performance meets prior expectations. That is, a positive assessment should be observed with the benefits outweighing the issues arising from the adoption of I4.0 (Miranda et al., 2016). Such issues may include maintainability (Blanchard et al., 1995), multilingualism (Jeyaraj et al., 2006), and exposure to operational risk (Hoffman, 2002). In such a case, the new technology proceeds to be institutionalised and part of the daily operations of the adopting organisation (Rogers, 1995). Otherwise, a probably discontinued adoption, if not withdrawal, of the technology may be decided when an organisation perceives more inhibited changes in the transformation process (Lienert, 2015).

Other strategic activities, such as the formalisation of systems development (Mathiassen & Munk-Madsen, 2007), knowledge transformation (Chor et al., 2014), a culture of change (Lienert, 2015), customer co-creation (Sung, 2018), intense research and development (Chor et al., 2014), and partnership establishment (Priyadarshinee et al., 2017), are also considered as indicators of this post-adoption phase. While the I4.0 adoption continues, it is imperative for organisations to regularly align their goals to a prescribed implementation on the following: (1) required brand and methodology specifications, (2) ideal product and project developments, (3) perceived change in the demands of core tasks, (4), open innovation among organisations and customers, (5) substantial efforts of generating innovative ideas, and (6) strong ties within a network of suppliers. As a result, straightforward transparency among customers and just the involvement of humans (i.e., operators or workers) in the loop may be upheld.

In summary, sets of indicators for each stage are positioned according to relevance as stipulated by a robust guideline. This guideline is streamlined like a conventional decision-making process where actions are preceded by the process of assessment and selection. In the context of the I4.0 implementation, this process is broken down into three segments, that is, perception, decision, and implementation, represented by the three stages of adoption. The pre-adoption stage encompasses indicators relevant to the perception of stakeholders as to the level or status of an organisation's I4.0 technology adoption. This apprehension is influenced by the knowledge and understanding of the involved personage about technologies that fall under the concept of Industry 4.0. This stage also includes the evaluation of the advantages and disadvantages of potentially implementing or rejecting I4.0 technology. The adoption stage includes indicators that discuss the decision-making process undergone by stakeholders in their conjecture for the potential to adopt or reject the technologies highlighted at the previous (pre-adoption) stage. The suitability shapes the final decision regarding the technology in terms of the fit for the organisation's needs and goals. Going from one stage to the next, the scope of the decision-making process becomes more complex. Moreover, at this stage, indicator sets provide a detailed outlook on being able to initially distinguish the capabilities of an innovation that tailor to the organisational needs. The post-adoption stage mainly concerns indicators that aided in confirming the initial evaluation set at previous stages and, thus, provide insights on the likelihood to continue the implemented innovation, or otherwise, end its use when proven to be depreciatory.

In summary, there are 11 indicators at the pre-adoption stage, 14 indicators at the adoption stage, and 37 indicators at the post-adoption stage. Every stage of adoption has a different number of indicators, which signifies its outright position in terms of stage suitability. In terms of a stage, suitability means that indicator sets are assigned to the most appropriate stage where a thorough management dashboard is deemed most necessary. It is also important to emphasise that this paper aims to present a set of indicators for every stage of adoption concerning its function and contextual representation rather than evaluate indicators at each stage quantitively. Some indicators need to be firmly established at a particular stage, so that its implicit representation in succeeding stages may already be covered. As an illustration, take, for example, the following indicators of the pre-adoption stage: the perception of the term Industry 4.0, perceived ease of use, and observability. It is canonical for business stakeholders to be able to initially distinguish the capabilities of innovation before making the adoption decision (i.e., adoption stage) and its eventual implementation (i.e., post-adoption stage). Following this principle, indicators — such as the perception of the term Industry 4.0, perceived ease of use, and observability — are believed to be most suitable for the pre-adoption stage since they represent the knowledge or awareness of innovation as well as the perception of stakeholders. This compels stakeholders to create an effective management dashboard based on such indicators and other strategic data that comes with it at this stage. Furthermore, it is necessary to understand that as stages of adoption progress, the scope of decision-making attributed to stakeholders becomes vast and even more complex due to the implied transfer and continuous management of tasks embedded in each indicator. That is, despite a unique set of indicators in succeeding (i.e., adoption and post-adoption) stages, it is nevertheless suggestive of the continued attention to the indicators of the previous stage.

The first case study was performed at a premier designer and manufacturer of electronic components for mobile communications and consumer electronics, producing microphones, speakers, and medical hearing devices. The company has been introducing disruptive technologies for over 65 years and is one of the industry leaders at present. Its strong drive for continuous improvement has pushed the firm to acquire and integrate innovative concepts and new forms of technology in its manufacturing processes. For the past two years, it has adopted the concept of IoT in its products to further enhance user experience. Also, it has ongoing plans for the implementation of smart manufacturing in incoming brands. Consequently, it can be argued that the firm has had a proper understanding of I4.0 and its implementation. To demonstrate the application of the proposed innovation stages and actual industry implications, this section presents the results of the evaluation of company A.

For company A, a sample research questionnaire used at the pre-adoption stage is presented in Table 3, further showing the linguistic ratings given by respondents for each indicator on the list. Subsequently, based on Xu (2006), the respondent answers were coded into five extended discrete linguistic labels expressed in equation (2). Then, to come up with a single score aggregate index for each stage, equation (3) was used. For the pre-adoption stage, using the proposed procedure presented in Section 3.3, company A attained an overall score of 1.364, signifying a good performance. Meanwhile, the adoption and post-adoption stages achieved good and fair scores of 1.643 and 0.784, respectively. These ratings imply that initiatives performed by the firm are aligned with the performance of each indicator concerning the implementation of I4.0 technologies. Since company A has a fair adoption performance, it needs to devise a holistic strategy to enhance its performance. Table 3 also provides insights into the performance of company A in terms of each indicator. Of the 11 indicators, the indicator of perceived risk has the lowest performance value, which implies that company A has limited risk assessment efforts in the adoption of I4.0 technologies. Thus, a thorough assessment of risks in various sources must be implemented to improve its pre-adoption status.

Furthermore, the evaluation provides insights into areas of improvement for the company A. By dichotomising each indicator, company A may adopt insights into its decision-making process, resource allocation decisions, and strategy formulation. However, these results must be considered with caution as some limitations in the evaluation process exist. First, just for the sake of demonstrating the applicability of the proposed indicator system, the case study only focuses on a small group of decision-makers who performed the linguistic evaluation process. For a more rigid application, an evaluation must be carried out holistically and involve a well-represented group of decision-makers from among all company stakeholders. Second, the evaluation is rough, as more quantifiable metrics were not determined for each indicator. Metrics could provide more meaningful and realistic measurements of indicators, which would provide a clear understanding of performance in terms of the I4.0 adoption. Third, coming up with a single-valued index, weights of indicators are considered equal. It is straightforward to note that each indicator has a varying degree of importance to a stage of the I4.0 adoption. Thus, having indicators of equal weights is just an oversimplification of complex real-life decision-making. Stakeholders may collectively assign a weight for each indicator based on its importance for an I4.0 adoption stage.

The second case study focused on a leading supplier of automotive seating solutions and electrical distribution systems and architectures. Products of this company have consistently delivered an elevated automotive experience for the end-users. As a global business that has been in the market for more than a century, it has continually achieved excellence by rigorously adapting to new technologies. To date, it has adopted the concept of IoT and smart manufacturing in its production processes to achieve a smoother flow of materials and workforce from the dock to dock. The extensive experience of the company in acquiring and implementing innovative strategies makes it another suitable source for the verification of the practical relevance of indicator sets.

Ratings issued by company B are presented in Table 4. Using the same procedure as with company A, the respondent's answers were coded as in equation (2). For the computation of the overall performance, equation (3) was used. As for the post-adoption stage, the overall index had a value of 1.541, which denoted a good implementation performance. Thus, indicators of this stage of the I4.0 implementation are well-evident at the firm. However, similar implications and precautions of the results, as discussed in the case study 1, are also applicable in this case. For brevity, these discussions are not presented.