Industry 4.0: Work design and the role of management choice

(1)

Industry 4.0: Work design and the role of management choice

By

Justus van Berkum

University of Groningen

Faculty of economics and business

-

Newcastle University Business School

Double Degree Master in Operations Management

December 2018

j.w.a.van.berkum@student.rug.nl

S3248526 – B7067153

Groningen supervisor: dr. J.A.C. Bokhorst Newcastle supervisor: dr. A. Small

(2)

1

(3)

2

Preface

This thesis is the final assignment of finishing my Double Degree Master in Operations Management. The project has been a rewarding and sometimes challenging process.

My interests have always been in technological innovations and process improvement projects. However, during previous internships & research projects I also encountered the indispensable role of the human factor in those, often technology driven, projects. While noticing that there are a lot of tools, methods and technologies available, I experienced that the human stake has always been a decisive part in enabling improvements to work. Since the masters programme mostly focuses on the technological aspects, I wanted to acquire more knowledge about the soft (human) aspects of management studies. This thesis allowed me to do this by studying novel technologies from a work design perspective.

Without the help of other individuals and organizations I would not have been able to write this thesis. Therefore, I want to thank my supervisor, who has always provided me with useful feedback and guidance. The current shape of this thesis is because of his input. In addition, I want to thank my Newcastle supervisor for the several skype meetings, in which I gained many new insights. Secondly, I want to thank the Hogeschool of Arnhem Nijmegen for providing the contacts to several interesting companies. It have been these companies which allowed me to gain the empirical insights to build my thesis. I want to express them my sincere gratitude for their willingness to collaborate and help me during this process. Lastly, I want to thank my family and friends who always supported and encouraged me during the process of writing my thesis.

(4)

3

Abstract

Nowadays, manufacturing industries have to cope with an increasingly complex competitive landscape; with changing demographics, globalization, resource constraints, mass customization and new and dynamic technologies. Industry 4.0, or synonymously smart manufacturing, is Germany’s answer to this new industrial revolution. Herewith aiming at technological innovation leadership, creating new values and constructing new business models. This new era of advancements does not only change business models and technologies, it also has profound implications for jobs. It is forecasted industry 4.0 will either augment (upskilling) or replace (deskilling) current workers and their skills. In conceptual research, it is expected that work design implications depend on the way the technology is designed and work around is organized. However, literature still lacks sufficient empirical insights, and is mainly relying on predictions. Therefore, this study identified the influences of industry 4.0 technologies and management decisions on work design, by means of an explorative case study. Eleven interviews were held at a single company, wherefrom the results have been validated at a second company with a single interview. Empirical evidence is found for both the upskilling and deskilling perspectives, which appeared to be related to the technologies utilized. In addition, it is found that management has influence over work design outcomes, in both technological design and task allocation. However, the choice management has in these phases is limited by the investment objective.

(5)

4

4.3.1 Task characteristics ... 31 4.3.2 Knowledge characteristics ... 35 4.4 Management choice ... 40 4.4.1 Pre-investment ... 40 4.4.2 Post-investment ... 43 5. Validation check ... 45 5.1 Technological difference ... 45 5.2 Work design ... 46 5.3 Management choice ... 46 5.4 Conclusion ... 47 6. Discussion ... 48 6.1 Work design ... 48 6.2 Management choice ... 50 7. Conclusion ... 52

7.1 Answering the research question ... 52

7.2 Limitations and implications for further research ... 53

References ... 54

(6)

5

(7)

6

1. Introduction

Nowadays, manufacturing industries have to cope with an increasingly complex competitive landscape; with changing demographics, globalization, resource constraints, mass customization and new and dynamic technologies (Prause and Weigand, 2016). Industry 4.0, or synonymously smart manufacturing, is Germany’s answer to the altered demands in this new competitive landscape, leading to a new industrial revolution. Thereby aiming at technological innovation leadership, creating new values and constructing new business models (Hermann, Pentek and Otto, 2016; Kang et al., 2016; Müller, Kiel and Voigt, 2018). Within this new industrial revolution, much research has been done to the leveraging of novel production strategies (such as Agile Manufacturing and Mass Customization) (Brettel et al., 2014; Prause and Weigand, 2016). This fourth industrial revolution is characterized by sensor technology, interconnectivity and data analysis allowing mass customization, increased flexibility, integration of value chains and greater efficiencies (Davies, 2015).

Besides changing business models and technologies, this new era of advancements has profound implications for jobs (Autor, Levy and Murnane, 2003; Brynjolfsson and McAfee, 2012; Gorecky et al., 2014; Autor, 2015; Waschull, Bokhorst and Wortmann, 2017). Gorecky et al. (2014) stated that the development of Industry 4.0 will change tasks and demands from humans. Computer capital has the potential to substitute for workers performing cognitive and manual, easy to automate tasks (Autor, Levy and Murnane, 2003). Simple repetitive work, related to the collection of data, will get increasingly automated. Thereby creating a shift towards assigning the human worker tasks emphasizing control, coordination and improvement of production based on improved transparency. This implies that some aspects of jobs might become increasingly simplified while other aspects may be enhanced (Waschull, Bokhorst and Wortmann, 2017). For example, the function of machine-operator that previously consisted of multiple physical tasks, will now contain more monitoring tasks, with technology substituting the physical set-up and data collection activities. In turn, this alters the composition of tasks together with the required skills and knowledge to perform those tasks.

(8)

7

role in Industry 4.0 are still rare. Many contributions to theory describing the effect of Industry 4.0 technologies on work design remain conceptual, therefore still relying on predictions (Waschull, Bokhorst and Wortmann, 2017). In addition, many of these studies rely on job design literature as basis for analysis. However, job design literature has been criticized for lacking a number of dimensions to represent today’s more broad work design environment (Morgeson and Humphrey, 2006; Oldham and Richard Hackman, 2010; Oldham and Fried, 2016). This study distinguishes itself by adopting an integrative and contemporary perspective to work design for an empirical qualitative case study, with a focus on the motivational characteristics.

Additionally, previous studies argue that work design implications depend upon the way the technology is designed and work around it is organized (Waschull, Bokhorst and Wortmann, 2017). It is expected that management has choice over (1) technological design, where prior to an investment is decided which tasks to automate, and (2) allocating tasks amongst workers and/or to workers, after the investments are made. Therefore, it can be argued that it is paramount to take human or job aspects into account when design decisions regarding Industry 4.0 technologies are being made (Nelles et al., 2016; Longo, Nicoletti and Padovano, 2017; Pacaux-Lemoine et al., 2017). This research aims to contribute to literature by empirically studying management’s methodological approach and considerations in Industry 4.0 technological design, in relation to work design characteristics. In addition, the effects of the management decisions, vision and goals on work design outcomes are described.

Research question

This previous described research aim has led to the following research question:

What are the influences of Industry 4.0 technologies and management decisions on work designs?

(9)

8

2. Theoretical Background

This section contains the theoretical background on the concepts of Industry 4.0, work design and management choice. Two stages of management choice are recognized: (1) pre-investment decisions during the technological design and (2) post-investment decision: choices to allocate task amongst workers and/or to workers. All concepts will be reflected on from a literature perspective, which results in a conceptual model. The conceptual model is presented in the fourth paragraph, presenting the summary and conclusion of this chapter.

2.1 The concept of Industry 4.0

Industry 4.0 is the term used to describe the latest industrial revolution, as initially proposed for developing Germany’s economy in 2011 (Lu, 2017). The first industrial revolution started late

18th century and was characterized by water- and steam powered mechanical manufacturing.

Next, the second industrial revolution began, characterized by electric-powered mass production and division of labour (think of the example of the Ford production line). From the 1970s onwards the third industrial revolution begun, an era of automation began, with electronics and information technologies advancing. Although the third industrial revolution is still on going, a new and fourth industrial revolution is taking place.

According to Prause and Weigand (2016), Industry 4.0 is about combining cyber-physical systems with industrial automation, creating context aware factories, which align people and machines real-time, eventually becoming self-regulatory and digitally integrated with all business functions. The development is an integrated process, consisting of complexity and agility among humans and machines (Lu, 2017).

(10)

9

Although the enabling technologies and principles on Industry 4.0 appear quite univocal, no commonly agreed upon definition exists (Lu, 2017). Many authors refer to Industry 4.0 in terms of cyber-physical systems (CPS) (Brettel et al., 2014; Lee, Bagheri and Kao, 2014; Lee, Bagheri and Jin, 2016; Prause and Weigand, 2016). Therefore, CPS will be used as theoretical lens describing the level of adoption of Industry 4.0 technologies.

Lee et al. (2016) argue that cyber manufacturing is at the core of CPS. “Cyber manufacturing is a transformative system that translates data from interconnected systems into predictive and prescriptive operations to achieve resilient performance. It intertwines industrial big data and smart analytics to discover and comprehend invisible issues for decision-making. In addition, a cyber-physical interface (CPI) plays a key role in cyber security for connected machines. (Lee et al., 2016, p.14)”

In cyber manufacturing, cyber-physical systems provide the integration between computational models and physical components while offering interoperability and resilience through the 5C architecture displayed in figure 2.1 (Lee, Bagheri and Jin, 2016). This 5-level architecture provides a guideline for developing and deploying CPS for manufacturing applications, by providing 5 steps (Lee, Bagheri and Kao, 2014). Where Lee, Bagheri and Kao (2014) build their five level architecture based on the functions of CPSs, Qin, Liu and Grosvenor (2016) developed a framework in which automation & intelligence levels were combined to determine the level of ‘Industry 4.0 readiness’. However, the constructs described in the framework described by Qin, Liu and Grosvenor (2016) lack clear operationalization, making it hard to measure. Therefore the 5C architecture is used. Providing a framework to indicate the level of CPS adoption the companies involved in this study have achieved. The sequential steps of the 5C architecture encompass the following:

1. Smart Connection: the acquiring of accurate and reliable data from machines and components.

2. Data-to-information Conversion: the tools and methodologies to transfer data into meaningful information.

3. Cyber: an information hub that is being used to form machine networks providing machines with self-comparison abilities and predict future behaviour.

4. Cognition: at this level info-graphics are used to allow experts users to make decision based upon the knowledge of the monitored system(s).

(11)

10

Figure 2.1 The 5C architecture adopted from Lee, Bagheri and Kao (2014)

(12)

11

2.2 Work design and Industry 4.0

Industry 4.0 technologies, allowing new methods of production, will cause drastic changes in the design of jobs (Waschull, Bokhorst and Wortmann, 2017; Müller, Kiel and Voigt, 2018). As mentioned earlier, two scenarios are discussed in literature: a scenario in which technologies augment the worker and its skills, and on the other hand a scenario where Industry 4.0 technologies replace them (Waschull, Bokhorst and Wortmann, 2017).

These scenarios are related to the deskilling and upskilling perspectives. The deskilling perspectives refers to simplified jobs and reduction in the level of skills required. This perspective can be related to capitalist management, which focuses on achieving greater control over workers, causing simplified and routinized processes. Opposed, the upskilling perspective refers to the scenario in which technologies augment human skills. In this case simple and routine jobs are automated, resulting in more complex jobs in terms of tasks and cognitive demands, requiring high level skilled workers (Waschull, Bokhorst and Wortmann, 2017). A lot of authors under scribe this upskilling perspective, arguing mainly simple manufacturing processes will be replaced (Gorecky et al., 2014; Hecklau et al., 2016; Müller, Kiel and Voigt, 2018).

In addition, Müller, Kiel and Voigt (2018) and Hecklau et al. (2016) extend the argument of the far reaching consequences of Industry 4.0 technologies to employee qualifications and acceptance. They argue that workers need to be qualified to approach the new technologies, requiring; willingness to learn, creative problem solving in social and practical settings as well as understanding network technologies, data analysis and processing. A second aspect is the acceptance of those new technologies. Concerns related to data transparency, dependency on technologies and workplace safety are considered hurdles in adopting Industry 4.0 technologies (Müller, Kiel and Voigt, 2018).

(13)

12

In identifying human factors in job design, the job characteristics model (JCM) as proposed by Hackman and Oldham (1976) is still frequently used. The JCM divides a job into five characteristics: skill variety, task significance, task identity, job autonomy and feedback. It is argued that these characteristics influence three physiological states regarding work: experienced meaningfulness, experienced responsibility for outcomes and knowledge of the actual results of the activities. These physiological states in turn affect work outcomes, namely: internal work motivation, growth satisfaction, overall job satisfaction, work effectiveness, and absenteeism (Hackman and Oldham, 1976; Fried and Ferris, 1987).

However, although the job characteristics model is frequently used, it has also been criticized. Fried and Ferris (1987) supported the multidimensionality of the model but did not find agreement on the exact number of dimensions. Moreover, extending the latter argument, Oldham and Fried (2016), Morgeson and Humphrey (2006) and Parker, Morgeson and Johns (2017) argued the existing measures to be incomplete. In addition, Morgeson and Humphrey (2006) argued other equivalents to the JCM lacked detail or key characteristics. Therefore,

Morgeson and Humphrey (2006) focused on work design instead of job design. Whereas job

design is mostly referred to as an aggregation of tasks (Parker, Morgeson and Johns, 2017), work design refers to “the study, creation, and modification of the composition, content, structure, and environment within which jobs and roles are enacted” (Morgeson and Humphrey, 2008). Therefore, this study will research the effect of Industry 4.0 on work design.

(14)

13

The WDQ consists of three categories of characteristics. The first, most researched, category consists of motivational characteristics, which are divided in task and knowledge characteristics. The category task characteristics is mostly concerned with how work is accomplished and tasks associated with a job. The knowledge characteristic reflects the kind of knowledge, skill and ability demands necessary to perform a function. By distinguishing task and knowledge characteristics, Morgeson and Humphrey (2006) acknowledge that jobs can be (re)designed to increase the task demands, knowledge demands or both. The basic principles of motivational characteristics state that jobs that possess many of these characteristics, are experienced as more motivating and satisfying. The second category includes social characteristics, recognizing that work is performed in a broader environment. The third and last category describes the work context, including physical and environmental contexts (Morgeson and Humphrey, 2006).

This study focuses mainly on the motivational characteristics. Since these characteristics can be used for (re)design of functions, which are most affected by Industry 4.0 technologies. In addition, these motivational characteristics allow best for comparison with earlier conceptual contributions describing the effects of Industry 4.0 technologies on work design, in terms of upskilling and deskilling perspectives. However, in contradiction to other researchers who used the JCM for their research, this study utilizes the more comprehensive contemporary view on work design because of the above mentioned reasons.

(15)

14

2.3 Management choice in Industry 4.0 system design and task allocation

Another way to assess the impact of work design theory is to study the extent to which it affects management thinking (Parker, Morgeson and Johns, 2017). “Ultimately, the implications for the human worker depend on how the technology is designed and how work is organized around it. As such, management has significant influence over the technology selection, design and implementation process. The role of the human worker in CPS hence is not determined by the technology itself, but a variety of other factors” (Waschull, Bokhorst and Wortmann, 2017). This point is also supported by Fantini et al. (2018), who state that placing the operator in the centre of design can direct the integration of people with technology in a specific, desired, direction. This would indeed imply management choice during design does influence work design outcomes.

However, there is a lack of information about the effects of management decisions in an Industry 4.0 setting. In recent literature (Borges and Tan, 2017; Fantini, Pinzone and Taisch, 2018), much attention is given to models which account for ‘the human factor’ during the investment process. However, the extent to which such models are actually used in practice and the effect they had is largely unknown. Therefore, this research aims to study to what extent management considers implications on work design during the technological design and allocation of tasks.

As pointed out by Fantini, Pinzone and Taisch (2018), who propose a model in which the operator is the centre of design, it is key to consider the employee, to direct the integration of people with CPS into the desired direction. Traditional approaches in enhancing human-machine interaction, are often based on the Fitts List (D ’addona et al., 2018), which describes if either a human or machine could better perform a certain function. However, as D’addona et al. (2018) state, a more effective approach assigns functions according to tasks. Therefore, Fantini et al. (2018) proposed a methodological path (figure 2.2) for engineers, operations managers and factory or work designers to follow when designing new jobs, assuming novel manufacturing systems are introduced or added to an existing setting.

(16)

15

Figure 2.2 Methodological path towards job design, adopted from Fantini, Pinzone and Taisch (2018)

In literature (Kumar, Murphy and Loo, 1996; Chan et al., 2001; Borges and Tan, 2017) various models, frameworks and methodologies for investment appraisal and decision-making are described, for the investment in advanced manufacturing technologies (AMTs). Or, as adopted by more recent studies (Borges & Tan, 2017), the investments in advanced automated manufacturing technologies (AAMTs).

According to Lo Storto (2018) AMTs refer to a family of manufacturing technologies, including flexible manufacturing, CAD, CNC, cellular manufacturing, computer-integrated manufacturing, additive manufacturing, automated material handling systems and robotics. Looking at these ‘families of manufacturing technologies’, it is seen AMTs comprise technologies related to the previous industrial revolution focused on automation (Qin, Liu and Grosvenor, 2016). However, AMT investment appraisal models do not solely relate to explicit technologies, their focus is foremost on the intangible benefits and justification methodologies associated with increasingly complex technologies (Kumar, Murphy and Loo, 1996; Chan et

al., 2001). Hence, this study argues these investment decision and appraisal models might be

(17)

16

At the basis of any investment lies a decision process. According to Kumar, Murphy and Loo (1996), who researched decision-making patterns in the light of AMTs investments decisions, a decision process consists of five main stages:

(i) Stimulus, the recognition of a need or opportunity.

(ii) Solution identification, this stage is threefold. Starting by the recognition of a problem, opportunity or crisis, followed by a diagnosis by finding a solution. Finally, AMTs might be recognized as possible solution.

(iii) Detailing, this includes preliminary authorization and AMT search activities. (iv) Evaluation, this stage is concerned with screening and justification of the AMT

technology.

(v) Authorization, the presentation of previous results for upper management approval.

(18)

17

(19)

18

Moreover, this study focuses on the strategic planning, justification and training & installation phases with regard to Industry 4.0 technologies and incorporation & training of human factors. In these stages (several) technologies are considered before reaching a decision regarding which technology, or whether, to adopt. In their 1999 paper, Chan, Chan and Mak developed three different justification methodologies for the investment appraisal for AMTs: strategic, economic and analytic justification. The strategic justification emphasizes technical importance, business objectives, research & development and the gaining of a competitive advantage. Next, the economic justification refers to more traditional, financial, measures and ratios as break-even analysis and ROI. The economic justification measures are often used in combination with strategic justification techniques. Last, the analytical justification emphasizes value analysis, portfolio analysis and risk analysis. When analysing the conceptual framework depicting the four major steps towards adopting AMT, as proposed by Chan et al. (2001) (figure 2.3), it becomes clear that the type of system (stand-alone, intermediate or integrated) determines which justification methods can be used. For example: a stand-alone system is justified using economic methods, where an integrated system can be justified using either of the three methods or a combination of these. As Industry 4.0 mainly relates to integrated systems technologies it is argued all justification methods or a combination of those might be applicable.

Although original publication stems from the previous century (Chan et al., 1999), AMTs investment justification methodologies are still being used (Borges and Tan, 2017; lo Storto, 2018). However, the importance of the incorporation of human factors in these justification methodologies is increasingly emphasized (Borges and Tan, 2017; Fantini, Pinzone and Taisch, 2018). Human factors, such as morale and workers skills, greatly influence the success of AMTs technologies. The lack of attention to human factors and other intangible factors is often mentioned as cause of failure in AMTs justification (lo Storto, 2018).

(20)

19

2.4 Conclusion & conceptual model

In conclusion, the aim of this research is twofold. First, the work design literature has been studied and related to the Industry 4.0 context. Two perspectives were found, upskilling and deskilling, which are commonly used in describing the predicted impacts of these technologies. This research will use the integrative work design questionnaire, with a focus on the motivational characteristics, as proposed by Morgeson and Humphrey (2006) to measure the impact of Industry 4.0 technologies. Therefore this research will provide novel empirical insights on work design within an Industry 4.0 context.

Second, the influence of management choice is studied as part of work design within companies that utilize Industry 4.0 technologies. It is found that there is little attention given to management choice, especially in relation to work design. However, there are methodological models which emphasize the importance of the inclusion of human factors during technological design and task allocation. This research acknowledges two points of management choice; during technological design and task allocation. It is expected that within those two steps, management can influence work design outcomes by the choices that are made.

Figure 2.4 provides a graphical representation, to illustrate the focus of this research. This conceptual model relates to the following research question:

What are the influences of Industry 4.0 technologies and management decisions on work designs?

Management choice

Industry 4.0

technologies

Tasks

(to allocate)

Work design

outcomes

(21)

20

3. Methodology

This chapter describes the methodology by which this study is performed, building up on the knowledgepresented in the introduction and theoretical background. First a description of the research design is provided. Second, data collection is discussed, as last issues regarding data analysis and validity are described.

3.1 Research design

This research aims to study the influences of Industry 4.0 technologies and management choice on work design, focussing on the shop floor level. The research distinguishes itself from conceptual research by collecting data empirically. This is done by an explorative case study, at a company that has implemented Industry 4.0 technologies, instead of mainly relying on predictions. The explorative design of this study enables the collection of novel and unexpected insights that cannot fully be anticipated. The research is both deductive as inductive, testing existing theories from conceptual research but also aiming to build novel theories based on empirical insights. As the research question and goal is twofold, we first focus on insights regarding work design and then extend this to management choice. The remainder of this study is presented in this order. The same methods apply to both parts, unless mentioned otherwise. Despite being time consuming and much care is needed in drawing generalizable conclusions, case research can have a very high impact. “Unconstrained by the rigid limits of questionnaires and models, case studies can lead to new and creative insights and development of new theory, and have high validity with practitioners – the ultimate user of research. (Karlsson, 2016, p. 166)”. Where survey research and experiments address a specific and focused problem, qualitative case study research allows to take a broader view on a problem (Blumberg, Cooper and Schindler, 2014). Since current literature lacks sufficient (empirical) insights, and the research area initially put forward by companies in Industry was broad, a broad view on the research is required. Therefore, a case study is best suited for this research. In addition, case study research creates the ability of addressing the problems faced by the companies involved in this research, the practitioners as referred to by Karlsson (2016), providing those with novel solutions while contributing to literature. In order to test if the same phenomenon occurs under similar conditions (e.g. companies operating I4.0 technologies) a validation check has been performed at a second company.

(22)

21

company, people have been identified and selected that are knowledgeable and have experience with the several research facets. Therefore, an analysis of problems encountered during and after implementation can be made. In addition, it becomes possible to obtain insight into the current and past situation and analyse differences in the company specific context.

3.2 Data collection

In order to collect data within this explorative case study, face-to-face semi-structured interviews were performed, in combination with a short questionnaire on the work design characteristics. Therefore, a mixed methods approached is used, with an emphasis on the qualitative interviews. Afterwards, a validation interview at a second company was held. Combining the findings of the theoretical framework with those of the qualitative interviews, quantitative questionnaire and validation check, triangulation is achieved, enlarging this research’ validity.

Since the duration of this research spanned only a few months it is logical to choose for a cross-sectional duration of research, taking snapshots from certain points in time; since many Industry 4.0 technology implementations will take significantly more time from stimulus until implementation. The interviews were conducted with employees who have different job titles and responsibilities. Hence, it became possible to obtain insights in specific areas and draw conclusion for different roles. At the case company, managers, engineers, team leaders, education managers, technicians (e.g. white collar-workers) and operators (blue-collar or shop floor workers) have been interviewed. The focus of the interview depended on the role of the interviewee. An list of interviewees, with anonymized functions, can be found in table 3.1.

Table 3.1 Interviewees according to function and organisation

(23)

22

As mentioned, to acquire information from the participants, a mixed method approach was used. Before the interviews were conducted, every interviewee was asked to fill out a short questionnaire. Case studies using qualitative interviews are often used to provide meaning to a phenomenon based on the respondent’s reaction (Blumberg, Cooper and Schindler, 2014), as is the aim for this explorative research. Additionally, interviews allow for a more in-depth dialogue in which new insights can be put forward. Therefore, semi-structured interviews were used. The interview topics and questions (appendix I) were derived on the basis of the theoretical background of this study and tailored to each function. The topics and questions were used as a guideline rather than a checklist, to ensure all topics were discussed. Characteristics identified in literature were discussed, while keeping an open dialogue to allow interviewees to talk freely and allow for new out-of-the-box insights, as is the aim of this explorative study. All interviews were held confidential to prevent biased answers. Furthermore, all interviews were recorded, transcribed and used for analysis. The duration of the interviews differed between 51 and 142 minutes. All interviews took place on the companies premises, a familiar and comfortable place for all interviewees.

For each category of employees (e.g. blue-collar workers and white-collar workers) a separate topic list (appendix I) was used to guide the interviews. For example, the management employees were interviewed about their view on the impact of technology on work, as well as the effects of certain management choices. During each interviewee the main points were summarized by the interviewer to guarantee a correct interpretation. Ending the interview, the interviewee was asked for other additional relevant insights. In addition, every interviewee was asked for allowance to be contacted for any additional questions or member checking at a later point in time. Since all interviewed blue-collar workers were operators, these terms are used interchangeably.

(24)

23

within the Industry 4.0 paradigm), it was expected that the same relationship was demonstrated, ensuring the studies internal validity. In order to test external validity, the results of the interview drawn from the case study at company A, are compared to the findings of the validity interview at company B. Following the theoretical replication logic, it is expected that the validation check might produce contrary results, in relation to the upskilling and deskilling perspectives, but for predictable reasons (e.g. the different setting). The extent to which findings drawn from the first case study apply to the second case illustrates the degree of external validity (Karlsson, 2016).

3.3 Data analysis and validity

The first step in data analysis was set by providing a description for the case. Although a write-up is only a description, it is central to the generation of Insights (Eisenhardt, 1989). The case write-up for the case company, based on expert interviews, can be found in appendix III. For company B, no write-up is provided due to confidentiality issues, a context is depicted in the validation check chapter using a technology classification framework. The second step of data analysis consisted of the transcription of all recorded interviews. To ensure confidentiality no company or interviewee names were mentioned. Only the blue- and white-collar workers were specifically separated. Based on the transcripts the coding process was started. In the coding process a combination of structural and data driven codes was used. The structural codes are those codes emerged from the research goals and questions, whereas the data driven codes grew from the interview data (DeCuir-Gunby, Marshall and McCulloch, 2011).

For coding, the program ATLAS.ti was used, following Corbin and Strauss's (1990) grounded theory for structuring the open and axial coding process. After transcription the interviews were open coded, where segments of text were coded with a label (appendix IV). An example of such label is ‘increased complexity’, labelling the following quote: ”Before, you had separated

operations. Now, there is one process flow. You cannot focus on a single operation anymore, an operator has to know the entire process. The amount of knowledge required increased”.

(25)

24

Although data on both work design and management is collected within the same interviews, the coding and pattern search was executed in two various, sequential, stages. First, all insights regarding work design were processed, analysed and described. Based on the insights gathered during this process, the relation to management choice & decision were analysed in a second step. This order makes sense, since first (un)expected outcomes have to be known before the possibility exists to include these insights into (future) investment decision models.

To ensure reliability on the interpretation of interview outcomes, member checking was executed during or after the interviews. Afterwards, a validation interview was held at a second company, which successfully implemented industry 4.0 technologies. The outcomes of the validation interview have been analysed using the same method as for the other interviews. Moreover, the outcomes of several studies were used to verify this studies results in the discussion chapter. To prevent premature or false conclusions a search for patterns will be executed by looking at the data in many divergent ways.

(26)

25

Table 3.2 Summary of methods used per researched topic

Work design Management choice - pre-investment

Management choice - post-investment Type of research Mixed methods Qualitative Qualitative

Data collection instrument Interview and

questionnaire

Semi-structured interviews Semi-structured interviews

Target group Blue- and white-collar workers

White-collar workers White-collar workers

Data analysis method Transcription, coding, analysis of

questionnaire

Transcription & coding Transcription & coding

Validity Member checking & validity interview

Member checking & validity interview

Reliability & triangulation Triangulation by mixed

methods, use of validated questionnaire, validation interview

(27)

26

4. Results

In this chapter the results of the interviews and surveys are presented. All results are described and analysed according to the traditional approach as described by Burnard et al. (2008). Meaning that the results will be based on the data collected as described in the methodology section. Comparison with existing literature will be described in the discussion chapter. The data is visualized and structured consulting the work of Miles, Huberman and Saldaña (2013). The first paragraph presents an analysis of the survey data. Second, based on the interview data, the results with regard to work design in the context of the overall workforce (blue and white-collar workers) are discussed in paragraph two. The results relating to work design in the context of shop floor workers (or blue collar workers), the main focus of this research, are described in paragraph three. Ending this chapter, the outcomes with regard to the aspects of management choice are discussed in paragraph four.

4.1 Survey outcomes

Prior to the interview, all interviewees were asked to fill out the Dutch translation of the work design questionnaire answering each question twice, for both the old and new situation. Eight out of ten interviewees filled out the questionnaire, of which 7 for both the old and new situation. Therefore, the sample of the survey has been relatively small, with an n of 7. For statistical analysis, a minimum of approximately 30 respondents is ought to be required in main-stream scientific literature. Therefore, the data will not be considered ‘statistically significant’. However, the results still provides very relevant insights and is used accordingly, as addition to the main data source of this study: the interviews.

(28)

27

about the coherency of answers. For example, the answer with regard to work scheduling autonomy between operators largely differs, whereas the answers on skill variety largely concur with one another. It is therefore likely that the overall trends indicated are more relevant for the characteristic skill variety, in which the answers are coherent.

It is noteworthy to mention that whereas blue-collar workers tend to experience differences between the old and new situation, the group of white-collar workers mention almost no differences.

(29)

Table 4.1 Survey outcomes. Colours used as descriptive statistic, see legend Si tu at io n Function Wo rk s ch ed u lin g au ton o my D ec is io n -ma ki n g au ton o my W o rk me th o d s au ton o my Ta sk v ar ie ty Ta sk s ig n if ic an ce Ta sk id en ti ty Fe ed b ac k fr o m th e jo b Jo b c o mp le xi ty In fo rma ti o n p ro ce ss in g P ro b le m so lv in g Sk ill v ar ie ty Sp ec ia liz ati o n So ci al s u p p o rt In iti ate d in te rd ep en d en e R ec ei ve d in te rd ep en d en ce In te ra cti o n o u ts id e th e o rg an iz ati o n Fe ed f ro m o th er s Er go n o mi cs P h ys ic al d ema n d s W o rk c o n d iti o n s Eq u ip me n t u se Process operator (1) 3,00 4,00 3,00 4,25 3,75 3,50 4,00 3,00 3,25 3,25 4,00 3,25 3,50 3,67 3,67 2,00 3,00 3,67 4,00 2,40 3,67 Process operator (2) 2,00 3,67 2,33 4,00 3,25 4,00 5,00 3,75 4,50 4,75 4,00 4,00 4,00 2,67 3,67 2,25 3,67 3,33 2,33 2,40 4,00 Process operator (3) 4,67 4,67 4,67 4,75 3,25 3,25 3,00 5,00 5,00 5,00 5,00 5,00 3,83 5,00 5,00 2,75 2,67 2,67 3,33 1,60 4,33 Mean new 3,22 4,11 3,33 4,33 3,42 3,58 4,00 3,92 4,25 4,33 4,33 4,08 3,78 3,78 4,11 2,33 3,11 3,22 3,22 2,13 4,00 Process operator (1) 4,67 4,33 4,33 4,50 3,75 3,25 4,00 3,50 3,25 3,50 4,00 3,25 4,00 3,67 3,67 2,00 4,00 3,67 4,00 2,40 3,67 Process operator (2) 2,00 2,33 2,33 3,00 3,25 3,00 3,00 2,75 3,00 3,00 3,50 3,00 4,00 2,67 3,67 2,25 3,67 2,67 3,33 2,40 3,33 Process operator (3) 4,33 3,67 4,67 3,50 2,25 4,00 3,67 4,00 3,50 4,00 3,75 3,50 3,83 4,33 4,33 1,50 3,00 2,67 4,00 1,80 4,00 Mean old 3,67 3,44 3,78 3,67 3,08 3,42 3,56 3,42 3,25 3,50 3,75 3,25 3,94 3,56 3,89 1,92 3,56 3,00 3,78 2,20 3,67 Difference in mean -0,44 0,67 -0,44 0,67 0,33 0,17 0,44 0,50 1,00 0,83 0,58 0,83 -0,17 0,22 0,22 0,42 -0,44 0,22 -0,56 -0,07 0,33 White-collar worker (1) 4,00 4,00 3,67 5,00 3,50 3,00 3,00 4,00 5,00 5,00 5,00 4,75 4,83 4,00 4,00 4,00 4,00 3,33 3,00 2,60 4,67 White-collar worker (2) 2,67 4,33 4,33 5,00 3,00 2,75 3,67 4,00 4,00 4,25 4,00 2,75 3,50 3,33 2,33 3,25 3,00 2,67 2,00 2,60 3,33 White-collar worker (3) 5,00 5,00 5,00 5,00 3,25 4,00 3,33 5,00 5,00 5,00 5,00 4,50 3,50 4,67 4,00 3,50 4,00 2,67 1,00 1,80 4,00 White-collar worker (4) 4,33 5,00 4,33 4,75 2,00 2,50 1,00 4,25 5,00 4,25 3,75 2,00 4,33 Mean new 4,00 4,58 4,33 4,94 2,94 3,06 2,75 4,31 4,75 4,63 4,44 3,50 4,04 4,00 3,44 3,58 3,67 2,89 2,00 2,33 4,00 White-collar worker (1) 4,00 4,00 3,67 5,00 2,50 3,00 3,00 3,25 4,00 4,75 4,75 4,25 4,33 4,00 4,00 3,50 4,00 3,33 3,00 2,40 4,33 White-collar worker (2) 2,67 4,33 4,33 5,00 3,00 2,75 3,67 4,00 4,00 4,25 4,00 2,75 3,50 3,33 2,33 3,25 3,00 2,67 2,00 2,60 3,33 White-collar worker (3) 5,00 5,00 5,00 5,00 3,25 4,00 3,33 5,00 5,00 5,00 5,00 4,50 3,50 4,67 4,00 3,50 4,00 2,67 1,00 1,80 4,00 White-collar worker (4) 4,33 5,00 4,33 4,75 2,00 2,50 1,00 4,75 5,00 4,25 3,75 2,00 4,33 Mean old 4,00 4,58 4,33 4,94 2,69 3,06 2,75 4,25 4,50 4,56 4,38 3,38 3,92 4,00 3,44 3,42 3,67 2,89 2,00 2,27 3,89 Difference in mean 0,00 0,00 0,00 0,00 0,25 0,00 0,00 0,06 0,25 0,06 0,06 0,13 0,13 0,00 0,00 0,17 0,00 0,00 0,00 0,07 0,11 Legend:

Motivational characteristics - research focus

Task characteristics Knowledge characteristics Social Characteristics Contextual characteristics

(30)

29

Table 4.2 Interview outcomes - descriptive statistics based on codes in appendix IV

C o m p an y Function Ph ys ic al Te amw o rk P att er n re co gn iti o n Su p er vi si o n In iti ta iv e A Process operator - - - = + + - + + + + + - + + A Process operator + = - + = + + + - + + + + + + + - + A Process operator + + = + + + + + + + - + + + + + + = + - + + A White-collar worker =- + + + - + = - + = -A White-collar worker = - + +* + + - + + = + - + -A White-collar worker + +* - = + + + = + A White-collar worker - - = + + + - + + = + + A White-collar worker + - = = + + +* + + - + + + + = + + -A White-collar worker + + = - + + + -A White-collar worker - - + - + + + -A White-collar worker - - + + + + - - + + + + = + -B Validation check - - = - = = + - = = - + + = = = = = + - + - - = +

Legend: * difference per type/level of operator

Ta sk s ig n if ic an ce Ta sk v ar ie ty W o rk me th o d s au ton o my D ec is io n -ma ki n g au ton o my

Social Characteristics Contextual characteristics

In te ra cti o n o u ts id e th e o rg an iz ati o n Sp ec ia liz ati o n So ci al s u p p o rt In iti ate d in te rd ep en d en e R ec ei ve d in te rd ep en d en ce Fe ed b ac k fr o m o th er s Er go n o mi cs P h ys ic al d ema n d s W o rk c o n d iti o n s Eq u ip me n t u se

Motivational characteristics - research focus

Increase Neutral Decrease

Skill variety P ro b le m so lv in g In fo rma ti o n p ro ce ss in g Jo b c o mp le xi ty Fe ed b ac k fr o m th e jo b Ta sk id en ti ty W o rk s ch ed u lin g au ton o my

(31)

4.2 Work design – implications on the overall workforce level

Although the main focus of the work design part of this research is on the shop floor level, a number of interesting findings on the overall workforce level (blue- and white-collar) were found. This will shortly be presented in this paragraph.

Besides the changes in work design for shop floor workers, it was mentioned frequently that the composition of the entire workforce has changed. The number of white-collar workers grew faster compared to the number of blue-collar workers. This can be explained by the increased technological complexity that was incurred due to the new technological design and by the choice to insource work previously executed by third parties and contractors. Many new functions had to be created in different levels of the organization, to support these technologies. An example can be found in a new layer of maintenance employees, called ‘control technicians’, who track the root causes of failures by studying a Programmable Logic Controller (PLC). Another example is found in the growth of the process automation department, which facilitates automation and is responsible for the programming, required for new products or changes in the production line.

Due to the increased complexity, and the fact that the process line has to be configured by humans, it was seen that the share of white-collar workers grew, relative to the share of blue-collar workers. Illustrating that although some simple tasks might be replaced, new, more complex tasks arose. Therefore creating the need for new functions, creating some increased specialization within and between the supporting departments.

4.3 Work design – implications on shop floor level

In this paragraph, each motivational (task and knowledge) characteristic is discussed separately. First, a short analysis on the survey results (table 4.1) will be provided. Second, the interviewee results (table 4.2) are described using quotes, codes and observations made, before ending with a conclusion on each characteristic.

(32)

31

4.3.1 Task characteristics

Autonomy

Work scheduling autonomy

On the characteristic of work scheduling autonomy, the survey provided different results. With differences in the height of scores and the trend indicated.

During the interviews, a criticism that was given in answering this characteristic was the perspective from which to answer. This provides an explanation for the incoherent answers in the survey. For example, a process operator answered: “I acquired a lot more responsibility

[due to growing to first level operator]. But all in all, I think that we [operators] now have a lot of freedom in how we organize our work”. The same operator also argued that on a higher

level, there is a standardized planning, which determines the sequence in which orders are produced.

A similar reasoning was provided by a white-collar and blue-collar worker, due to the fact that higher level operators acquire more steering tasks, it is likely to assume they have more freedom in work scheduling autonomy. For the remaining operators, no real changes were found.

Decision-making autonomy

For the characteristic decision-making autonomy, in the survey, two out of three operator indicated an increase. However, during the interviews the majority of respondents indicated a decreasing trend. Whereas the white-collar workers agreed rather unanimously to a decrease, the operators were more dispersed in their answers. As an operator argued “In the old situation

we had lots of freedom when setting up a machine. We did this together [with a process engineer], but at a certain point you had enough experience to know what to do and what not to. This became more difficult in the new situation”. This quote illustrates a decrease in

decision-making autonomy. A manager made a similar comment, but from a different point of view: “In the old situation we encountered operators to be somewhat stubborn, at the beginning

of their shift, many operators tended to change machine settings to their liking. That is something we did not want here obviously, as it should be the job of the process engineer”.

This increased emphasis on the separation of certain tasks can be explained by the increased complexity of the new process and the enlarged consequences of making mistakes. Which is considered a consequence of the new technologies. A manager provided an example: “Before

(33)

32

The last mentioned quote presents an interesting issue indicated by many operators, who argued the ‘degree of decision-making autonomy’ largely depends upon the perception of the concept. Since there was an increase in procedures and regulatory boundaries, a decrease in decision-making autonomy can be argued. Opposed, the same operator also argued that there are other tasks, such as coordinating his shift, in which he has increased decision-making autonomy. However, only the highest (first) level operators have such coordinating tasks. Therefore, it can be argued that the level of decision-making autonomy is dependent on the qualification of the operator. This could be an explanatory mechanism to the differences in answers. However, assuming that assigning an operator with coordinating tasks is an organisational choice, rather than a consequence of technology, it is argued that due to Industry 4.0 technology decision-making autonomy has decreased.

Work methods autonomy

Work methods are comprised of the freedom and independence determining how work is done. In the survey, all but one operator indicated this characteristic to remain unaffected. During the interviews, answers tended more towards a (small) decrease. As a white-collar employee described, the work instructions remained more or less the same, not affecting the degree of work methods autonomy.

However, most white-collar workers described the impact of making mistakes has severely increased. Whereas in the previous situation operators sometimes diverted from working instructions, based on experience, the current process line is described as too complex to do so. The impact of making mistakes is too large. Although the amount of working instructions might not have been increased, it is emphasized that it is of increased importance that operators stick to these instructions. In this perspective, the work methods autonomy of operator’s decreases, as working according to working instructions is increasingly emphasized. Although the trend might be decreasing, the degree to which it decreases is only little. This coincides with the survey outcomes and reactions during interviews.

Task variety

Task variety is a characteristic rewarded with one of the highest scores in the survey, mainly the new situation. Similarly for the characteristic skill variety. Since one needs certain skills to perform tasks, and tasks require certain skills, the two characteristics are different although interrelated. What became apparent is that the number of physical tasks decreased. “Before,

(34)

33

automated.” This trend is also reflected in the contextual characteristic ‘physical demands’ for

which all respondents mentioned a decrease.

Opposed, tasks requiring pattern recognition, analytical, initiative and teamwork skills increased. However, a minor difference can be made between the various levels of operators. Since a low level operator has different tasks, this operator mostly performs manual labour. The higher level operators are imposed with more analytical, supervision and coordinating tasks.

“Interviewer: do you have an increased variety of tasks now? Operator: Yes, for me perhaps, but that is because I grew to first level operator and are therefore responsible for managing the shift.” A white-collar worker mentioned the same: “The first level operator also has some steering tasks, managing people within his shift.” This might be an explanatory mechanism for

the difference indicated in the survey.

The difference between levels of operators and perceptions is also what made task variety a hard to judge variable. For certain levels of operators it increases, whereas for other operators it might decrease or remain the same. This aspect is reflected by answers of both operators as white-collar workers, as illustrated by the following two quotes:

“We [operators] have an increased number of tasks. At the old site, you worked at a certain machine. Here we have this [process] line. Everything we had in the old site, is here combined in one process.”

“Before, you worked everywhere, someday at machine x, someday at machine Y. That made the job more fun and exciting. I don’t want to argue the work is not exciting here, but it is more of a controlling function.”

In summary, mainly physical tasks have been replaced by technology. However, some other tasks were born to live, sometimes controlling in nature, sometimes cognitively challenging. Whether this leads to a change in task variety, depends on one’s job and situation. Two white-collar workers summarized:

(35)

34

Task significance

In the survey, one operator reported a small increase in task significance whereas his two colleagues filled-out the characteristic to remain the same. During the interviewees most respondents reported an increase.

When relating to task significance the primary focus has been on the significance one’s job has within the organization. Many interviewees indicated an increase in teamwork. For example between operators themselves or between the operator and process engineer. In both cases the employees rely on each other’s input to perform their functions. Therefore it can be argued the work of the operator has an increased significance. An example was provided by an operator

“The tasks have shifted to the early signalling of distortions and recording these in the database, this is enabled by the increase availability of process data”.

Task identity

Task identity is the ability to finish a piece of work from beginning to end. In the survey, two operators indicated an increase, whereas the third operator indicated a decrease. The reactions during the interviews predominantly described increases.

Before the product has undergone all operations, it past several work stations. The results of the finished product in that sense is a team effort. As a manager described “I can imagine this

decreases, since in the old plant one could say ‘I drilled 200 holes today’, whereas in the current situation one would say ‘today we made 200 springs as a team’. I can therefore imagine the feeling of task identity to decline, but do not hope this is the case”. An operator stated; “The work allows us to finish a complete product, so yes the work allows us to finish a complete piece of work. Additionally, product tracing allows us to retrieve data from each specific product, another big change.” Whereas the manager feared for a decrease in task identity, due to the

(36)

35

Feedback from the job

In the survey operators provided quite differing, incoherent answers. During the interviewees few useful answers were provided. However, the answers that were given indicated a clear increase. As an operator argued “feedback from the job in relation to results and effectiveness

is pretty big, an example can be found in the OEE. These kinds of insights are new, and a big improvement in my opinion”. Overall, the new technologies provide more (real-time)

information on task performance. Therefore, it can be argued feedback from the job has increased.

4.3.2 Knowledge characteristics

Job complexity

In the survey, the majority of operators indicated an increase in job complexity. During the interviews, an increase of job complexity was described on the shop floor level. In addition, respondents also argued an overall increase (for the entire workforce) in job complexity. On the shop floor level, interviewees indicated that the low-level skills, required for physical tasks and handling, have been largely substituted by technologies. For instance by the use of robots and Automated Guided Vehicles (AGV’s). “In the old location, every machine required

an operator, sometimes two if the machines were quite large. For every machine, products had to be loaded, unloaded and transported. Here [new location] we have got this line [process lay-out]. This means the process has to be loaded and unloaded only once, and those tasks we replaced, since we now use an AGV.”

Furthermore, it is argued that the degree of complexity (1) varies between workstations and (2) is dependent on the level of the operator. “The degree of complexity differs per workstation.

For example, when operating the [3d] rolling process, it can be quite hard to get the product on specification. Opposed, when operating the press, there is little you can do. This is more of a controlling function, which is not that exciting.” In addition, the level of the operator is related

to the tasks an operator executes, influencing the degree of job complexity. “The first level

(37)

36

Besides affecting the tasks and skills required from operators, the technological changes impose different requirements to the white-collar workers. In order to automate, robotize and digitalize certain processes, an increased technological understanding is required. Moreover, due to the process lay-out and ‘connecting technologies’ the technological understanding exceeds the machine level. Understanding the way in which the different machine interact with each other, using different technologies, is of utmost importance. Therefore, it is argued the new technologies caused an increase in complexity among different functions at most levels in the organization.

In conclusion, many respondents described an overall increase due to increased knowledge requirements, educational requirements and increase of high-level skills. Furthermore, the operators described that their jobs imposed increased cognitive demands. Although not required at all times, these skills and abilities are found to be required to perform ones job. An engineer provided an example reflecting the above: “Yes, I think we are heading towards a type of

employee that can switch more easily, is able to use his head and process information quickly.”

Information processing

Information processing, an aspect in which both blue-collar and white-collar workers describe an increase unanimously, both during interviews as in the survey. In the survey, the process operators even indicated a (mean) increase of one (out of five), the highest difference in the entire survey.

During the interviews it was indicated that due to the increased number of IoT technologies, much more information is gathered than was done before. In the current situation, the process line is controlled and monitored by using Supervisory Control and Data Acquisition (SCADA) software. This system is used to present (real-time) data to operators for control and monitoring purposes. Thereby, the amount of information that has to be processed increased severely. Where in the old situation the operator mainly used his senses and craftsmanship, the operators now also need to monitor the data that is provided to them. As a respondent indicated, the information provided by SCADA is not meant to replace the information gathered by human senses and craftsmanship, but used as complement.

(38)

37

Problem solving

In both the survey and interviewee, the majority of respondent indicated an increase in the degree of problem solving.

Since the process line and technologies are new, the problem solving often involves diagnosing and solving non-routine problems. Moreover, the problem solving often requires increased analytical and pattern recognition skill: “The further in the process a problem arises, the more

impact it has on the extent to which the previous process steps have to be investigated”.

Opposed, as some respondents described, due to the increased complexity, there is an increased number of problems that an operator is unable to solve alone. “Where before one could try to

change a switch or fuse, that’s not the case anymore. The impact of changing something is way higher. We have our ‘out of control plans’, but if those do not work, the technical service department has to be contacted”. However, these technical workers do rely on the descriptions

provided by the operator for trouble-shooting. “As operator you have to provide the technical

service department with a good description of the problem. You have to describe what has gone wrong and what happens before and after.”

Although the operators are required to provide a description of failures on a more abstract level, the actual problem solving tasks are more often transferred to specialized employees. This reasoning also provides an explanation for the two respondents that argue problem solving decreased for operators. However, another respondent argued that the real problem solving has always been done by the technical service department which has not changed. “I’m not sure if

that [the solving of failures] has ever been part of the operators tasks.” These respondents

linked problem solving to the solving of failures or malfunctions. Taking another perspective to problem solving, an operator provided an example for a problem relating to the set-up:

“There is an increased cognitive demand if you have a deviation in the product. There are many parameters, let’s say buttons, which one can adjust [to compensate for the deviation in the product]. It’s always a quest of how to solve this adequately, for instance by a deviation of standard size”.

(39)

38

Skill variety

Both the survey and interview results demonstrate an increase in skill variety. As shown in table 4.2, a trend from low-level skills to the utilization of high-level skills can be identified. During the interviews, questions were asked about five orthogonal dimensions of skill:

Physical skill: besides physical demands being reduced, the physical skill required to perform

those tasks has decreased as well. Due to the robotization and digitalization, the setting up of machines has become less physical. For example, in the old situation a centre station was used, now parameters are set digitally. “Before, you had lots of things, centre tables and some other

mechanical things that were set-up by the operator. Here [in the new situation] those tasks are not here anymore, they all have been automated. A robot picks up the spring and places it in the correct places. Change-over to a new product meant making some physical alterations to a machine in the old situation, now a new set of [product-related] parameters is downloaded and the robot puts the product in the correct place. The tasks of setting-up, those are no more.” Teamwork skill: an operator indicated that in the new situation, self-directing teams were

introduced. Since the line operates as a single process, the process has to be run as a team. If one operation fails, the entire line comes to a hold. “Yes [team effort has increased], the

operators in a shift rely on each other and the way they interact.”

Not only does the teamwork between operators increase, the teamwork between different types of functions increases as well. “If there is something [a failure or deviation], or something can

be made easier [for instance the user interface], operators involve the process engineers”.

Concluding, it can be argued teamwork skill has increased.

Analytical skill: all respondents indicated an increase in the analytical skill. Due to the increased

digitalization, complexity, information processing and dependency between operations it was generally agreed that operators increasingly need analytical skills. An operator provided an example: “Interviewer: is there more information to process? Operator response: Yes, you

have to analyse, what if I do x, what would be the consequence? If the spring would be too wide at a certain machine, this might have consequences for machine further down the line”. Another

example was provided by a manager: “Especially the demands on the analytical abilities have

(40)

39

Supervision skill: the respondents that described this dimension of skill agreed to an increase in

supervision skill required. For example, first level operators, who acquired managing tasks, required increased coaching skills. Furthermore, because of the increased complexity of the overall workflow, coordination skill increased. Examples were related to the increased complexity of keeping overview over the entire process.

Pattern recognition skill: An operator stated “Recognizing patterns is increasingly important in controlling, and detecting malfunctions in the process”. As described earlier, due to the

complexity of the overall process, operators have to be able to identified patterns among processes. As argued before, it is important an operator recognizes what consequences a change early in the process would have on following steps. Another example of pattern recognition is provided by a maintenance engineer: “Once, we had this situation where the operators started

the line. Normally, when the line would be running for a while, the temperature of a certain process step would climb from 176 to 180 degrees Celsius, an operator noticed this did not happen. After checking, it appeared one of the sensor was malfunctioning.” In conclusion, the

degree of pattern recognition skill has increased. Specialization

Industry 4.0: Work design and the role of management choice