Master Thesis for Double Degree MSc Technology and Operations Management
MANUFACTURING
IN THE REALMS OF INDUSTRY 4.0
A Case Study on Knowledge Management and
Quality Assurance in the Context of Digitalisation
by
NIEK BARTELDS
1
MANUFACTURING IN THE REALMS OF INDUSTRY 4.0
A Case Study on Knowledge Management and Quality Assurance in the Context of Digitalisation
Date:
10 December 2018Word count:
14402Master Thesis Double Degree TOM
EBM028A30
MSc Technology and Operations Management University of Groningen
Supervisor: prof. dr. ir. J.C. Wortmann NBS8399
MSc Operations and Supply Chain Management Newcastle University
Supervisor: dr. G. Heron
Company supervisors:
i
ABSTRACT
In recent years, Industry 4.0 has been introduced as a paradigmatic shift based on digital transformation of the manufacturing industry. Because it so far remains uncertain how digitalisation will affect the position of manufacturing organisations towards the management of knowledge and quality, conceptual and empirical understanding of the interrelations between digitalisation, KM and quality is required. This study responds to this need, by thoroughly exploring how knowledge processes – creation, transfer and application of tacit and explicit knowledge – might be leveraged by digitalisation to enhance the assurance of quality. For this research, a single case study methodology is adopted. The case involves a composite manufacturer in the aerospace industry, that is currently in the middle of a digitalisation effort to progress towards Industry 4.0. A hybrid approach of inductive and deductive thematic analysis was applied to data from direct observations and in-depth interviews, to gain understanding in the way digitalisation, KM, and quality are interrelated in this empirical context. The results suggest that the beneficial effect of KM to assuring quality will be amplified with the implementation of fully digitised manufacturing systems and the emergence of advanced digital technologies. As conceptualised, findings confirm that education and work instructions are effective instruments for facilitating knowledge flows and are apt to be digitally-enabled, and thus are at the core of quality assurance in a digitalised context. It was found that the adherence to standard operating procedures plays a central role in linking KM and quality. Moreover, novel insights are generated as to the role of organisational culture and promoting cross-functional communication. Therefore, this paper suggests adopting a socio-technical view to further the state-of-the-art in KM, especially in the context of Industry 4.0.
Keywords: digitalisation; knowledge management; quality assurance; tacit and explicit knowledge;
ii
TABLE OF CONTENTS
ABSTRACT ... i LIST OF TABLES... iv LIST OF FIGURES... v I. INTRODUCTION ... 1II. THEORETICAL BACKGROUND ... 3
Digitalisation in the Context of Industry 4.0 ... 3
A Closer Look at Digitalisation ... 4
Manufacturing Execution System ... 4
Managing Knowledge in Theory and Practice ... 5
Terminology of Knowledge-Related Concepts ... 6
Zooming in on Knowledge Management ... 7
Linking KM to Quality ... 9
Conceptualisation of Quality ... 9
Connection with KM ... 10
Linking Knowledge, Quality, and Digitalisation in Practice ... 11
III. METHODOLOGY ... 13
Research Design ... 13
Considerations for Data Collection ... 13
Data Analysis Approach ... 15
Ensuring Quality in Case Study ... 16
IV. RESEARCH SETTING ... 17
Case Background ... 17
Digitalisation in the Aerospace Industry ... 17
Description of Pilot4.0 ... 18
Case Findings on Digitalisation... 18
V. FINDINGS AND ANALYSIS ... 20
Combination ... 20
Digitalised Combination Process ... 20
Implications for Quality ... 20
Internalisation ... 21
Effect of Digitalisation ... 21
Implications for Quality ... 22
Externalisation ... 22
Digitalised Externalisation ... 22
iii
Socialisation ... 23
Socialisation Facilitated by Digitalisation ... 23
Implications for Quality ... 24
VI. DISCUSSION ... 26
Knowledge Management ... 26
Digitalised SECI Processes ... 26
Work Instructions and Education ... 27
Concluding Remarks on KM ... 28
Quality Assurance ... 28
Considerations for Instruction Quality ... 29
Process Quality Assurance ... 29
Concluding Remarks on Quality ... 30
Influencing Factors ... 31
Research Implications ... 32
Theoretical Contributions ... 32
Practical Contributions (and Managerial Recommendations) ... 33
VII. CONCLUSION ... 35
REFERENCES ... 37
APPENDIX A. MES Functions and Technologies ... 43
APPENDIX B. Case Study Protocol ... 46
APPENDIX C. Abbreviations and Acronyms ... 54
APPENDIX D. Further Theoretical Background on KM ... 55
iv
LIST OF TABLES
TABLE 1 Interview Details ... 14
TABLE 2 Methodological Elaboration of Research Variables and Elements ... 47
TABLE 3 Code Book for Quality Variable ... 52
v
LIST OF FIGURES
FIGURE 1 Model Map of MES Functionalities and Technologies ... 5
FIGURE 2 Knowledge Conversion Modes in SECI Framework ... 7
FIGURE 3 Visualised Conceptual Framing of KM Elements ... 8
FIGURE 4 Visualised Conceptual Framing of Quality Elements ... 10
FIGURE 5 Basic Conceptual Framework ... 11
FIGURE 6 Expected Interrelations between Variable Elements... 12
FIGURE 7 Data Analysis Process ... 15
FIGURE 8 Research Contributions to Conceptual Model ... 31
FIGURE 9 Interview Outline and Questions ... 50
1
I.
INTRODUCTION
The world of manufacturing industry is on the verge of a revolution, comparable to the past three waves of industrial innovation driven by mechanisation, electrification and automation respectively. This fourth industrial revolution, termed Industry 4.0, entails a radical transformation of the manufacturing environment enabled by the proliferation of disruptive digital technologies (Kagermann, 2015). This digitalisation changes the way manufacturing firms generate, collect, and process their manufacturing operations data. Therefore, firms need to make sense of huge amounts of real-time data and information regarding their production processes. As Naisbitt (1982: 24) famously epitomises this situation: “we are drowning in information, but starved for knowledge”. Successfully governing all those data and information to generate knowledge can create value for organisations, which is the central premise of knowledge management (North, Maier, & Haas, 2018). Such value for instance lies in improving quality, for which digitalisation – through Big Data or cloud computing for example – introduces new possibilities to the manufacturing landscape (e.g. more sophisticated data analytics). Additional to this technological perspective, making sure people in the organisation are knowledgeable aids in pursuing quality as well (Wiig, 1997).
2
Although the separate constructs of digitalisation, KM, and quality have been studied previously, the interrelations between these phenomena have been largely overlooked. Whether digitalisation changes the role of KM in pursuing quality thus remains unclear, which might lead to serious impediments for companies in their quest to Industry 4.0. Besides, conceptual understanding regarding the management fields of knowledge and quality thus is limited in the Industry 4.0 literature. This paper seeks to narrow these shortcomings by examining how the three constructs are related, and whether they can be used to support and foster each other in the progression towards Industry 4.0. Therefore, these practical and theoretical gaps are aimed to be bridged by addressing the following research question:
How can the digitalisation of manufacturing processes in line with Industry 4.0, support the pursuit of quality through facilitating the management of (digital) knowledge?
This question is rather generic to give a comprehensive answer to, as it includes elements touching upon the broad disciplines of Industry 4.0, KM and quality management. To be able to answer the central question, this calls for an in-depth investigation in a real-world context reflecting the key phenomena central to this research. Hence, a single case study was performed at a Dutch manufacturing firm. The relevance of the case study methodology is further explained in Chapter III, as well as the choice for the case company.
3
II.
THEORETICAL BACKGROUND
For years, the literature has recognised knowledge and quality as crucial ingredients in organisational success. Yet, there is dispute in terms of the exact link between the management approaches of both aspects. However, a recent literature review from Marchiori & Mendes (2018) identifies one stream of research that considers KM an enabler of quality management. It builds on the view of Waddell & Stewart (2008), who suggest that KM can be regarded as a vehicle for achieving quality objectives. As implied in the research question presented in the previous chapter, this paper also takes that stance. It seeks to contribute to the understanding how this relationship between KM and quality takes shape in a digitalised context, as digitalisation is expected to give rise to a different approach to managing both knowledge and quality. Comparably, the third industrial revolution also caused such a transformation in knowledge and quality management. Through the emergence of automation and information technology (IT), organisations began to understand that intangible resources such as information and knowledge should be regarded as dominant resources for manufacturing (North et al., 2018). Also, it sparked a “whole new quality philosophy”, in which quality was assured more preventively and quality management functions needed to be automated as well (Davis, 1985: 75). Accordingly, these developments changed the way knowledge and quality were managed. This parallel with the third industrial revolution adds to the relevance of studying KM and quality in the context of digitalisation: it deals with how current and near-future digital technologies associated with Industry 4.0 will shape the management of knowledge and quality, in a similar vein automation and IT did before. In this specific context, it is underexplored how manufacturing digitalisation facilitates the creation, transfer, and application of knowledge, and how this may aid in assuring and enhancing quality. This requires an insight in the academic understanding pertaining to the link between management of knowledge and quality, by providing a theoretical framework on the concepts of KM and quality. However, before that, the following section discusses the literature on Industry 4.0 digitalisation.
Digitalisation in the Context of Industry 4.0
4
A Closer Look at Digitalisation
In the context of digital transformation in the manufacturing industry, both academics and practitioners adopt the terms digitalisation and digitisation. Although there are semantic differences between those terms, they are often used interchangeably. Therefore, a terminological disambiguation is needed in this work. The distinction between the two is that digitisation entails translating various sortsof artefacts – e.g. documents, archives, guidelines – from analogue into a digital format, whereas digitalisation refers to how such digitised artefacts and relationships between them alter the socio-technical structure in the organisation (Nylén, Holmström, & Lyytinen, 2014). As such, manufacturing digitalisation is a more rigorous transformation than mere digitisation of existing processes: it fundamentally changes business models, processes, and IT infrastructures (i.e. technical structures) and organisational roles and values, decision-making, and ways of working (i.e. social structures) as well (Parviainen, Tihinen, Kääriäinen, & Teppola, 2017). In a similar way, Kagermann, Wahlster, & Helbig (2013) characterise the role of digitalisation in Industry 4.0 as merging the digital and real manufacturing domains. This cyber-physical convergence involves the digital integration and networking of machines, people, and production facilities and systems, with ICT (e.g. the Internet). Corresponding to Wokke (2018), three trends from the cyber-physical convergence can be identified: data analytics (Bartodziej, 2017; Roblek et al., 2016), real-time human-machine interaction (Bartodziej, 2017), and changing employee skillset (Hirsch-Kreinsen, 2016; Waschull, Bokhorst, & Wortmann, 2017). For organisations progressing towards Industry 4.0, these trends imply a drastic transformation of again the socio-technical structure of the manufacturing environment, driven by digital technologies. In other words, the cyber-physical aspect of Industry 4.0 is an essential component in discussing digitalisation in this context.
Summarising the above, Industry 4.0 is about fundamentally rethinking the production organisation, by means of employing a digitalised system of people, machines, tools, and technologies. Especially, the connectivity and interaction among these entities – enabled by digital technologies – is key, which characterises the importance of digitalisation in Industry 4.0 (Sjøbakk, 2018). Hence, cyber-physical convergence through digitalisation also transforms the systems architecture of the organisation. Digital manufacturing systems can be achieved by implementing a Manufacturing Execution System (Qiu & Zhou, 2004), as Iarovyi, Mohammed, Lobov, Ferrer, & Lastra (2016) argue that MESs are suitable for adopting a cyber-physical character. Also, Arica & Powell (2017) regard MES as an enabler for realising Industry 4.0 opportunities, since the core functionalities of a MES serve as a platform for implementing various (future) digital technologies associated with Industry 4.0.
Manufacturing Execution System
5
Hence, the usefulness of MESs in manufacturing digitalisation in the Industry 4.0 context, can be illustrated by the key terms of integration and convergence. These buzzwords imply the tendency of different systems, processes and data flows to evolve towards synchronised interoperable manufacturing entities that are interacting synergistically (Oesterreich & Teuteberg, 2016; Olawuyi & Friday, 2012). Moreover, Almada-Lobo (2016) suggests that the features of a MES make it possible to gradually move towards a cyber-physical design of manufacturing execution. This is engendered by significant advances in network computing and software technologies (Saenz de Ugarte, Artiba, & Pellerin, 2009). Based on these developments in the MES domain, Saenz de Ugarte et al. (2009) have identified several improvements that might lead towards a MES suited to Industry 4.0. Regarding this study’s context, it is unclear which MES functions and technologies contribute to KM. Additionally, the question arises how the various technological enhancements as set out above might affect the management of quality through a MES. Figure 1 shows a diagram of MES functionalities (MESA, 1997), together with technology enhancements that show how MESs have developed and how they might develop further over the coming years. This paper will not go into the specifics of the functionalities and technological trends associated with MESs, however they are described in Appendix A for reference purposes.
FIGURE 1
Model Map of MES Functionalities and Technologies
Managing Knowledge in Theory and Practice
6
concept of KM. In a business context, most of these agree on KM being a systematic and organized approach (e.g. Ragab & Arisha, 2013; Tomas & Hult, 2003; Wiig, 1997), that ultimately is aimed for creating value and achieve competitive advantage (Mårtensson, 2000). As O’Dell, Grayson, & Essaides (1998: 6) simply but comprehensively put it, KM is “a conscious strategy of getting the right knowledge to the right people at the right time” so they can take action and create value.KM can hence be regarded a managerial effort to pursue organisational success through its knowledge processes.
Terminology of Knowledge-Related Concepts
As knowledge is an abstract concept, it is useful to provide some terminological background on various related concepts before diving deeper into the conceptualisation of KM. Besides the brief discussion in this subsection, further reading on the theoretical foundations of knowledge and KM is presented in Appendix D. Historically, the distinction between tacit and explicit knowledge is key in KM research. The difference between the two types has been widely discussed in literature. Nonaka (1991) poses that explicit knowledge has a formal and systematic nature, and is therefore easily expressed and shared. It is articulated, generalised knowledge (Alavi & Leidner, 2001), which is mainly communicated in codified forms such as electronic databases, or written specifications, manuals, and protocols (Ragab & Arisha, 2013). Tacit knowledge on the other hand, is hard to formalise, express and communicate, because it is highly personal and contextually-bounded. It has a technical and cognitive dimension, in which the former covers “informal, hard-to-pin-down skills” often referred to as craftsmanship or know-how, and the latter entails taken-for-granted mental models that shape the way one perceives the world (Nonaka, 1991: 98). It is often argued that knowledge is more valuable in tacit than in explicit form. However, as one of very few scholars, Bohn (1994) contends that explicit knowledge is most desirable for organisations. Based on this viewpoint, Alavi & Leidner (2001) stress the potential of technology-enabled KM processes, i.e. (digital) technologies being used in explicating tacit knowledge.
7
Zooming in on Knowledge Management
Fundamental to the concept of KM are knowledge processes. These are the broad approaches through which knowledge is managed. They are specifically focused on the flow of knowledge through the organisation, with an emphasis on how knowledge is created, shared, and distributed (Alavi & Leidner, 2001). More specifically, the interaction between tacit and explicit knowledge is embodied in knowledge processes by knowledge conversion modes (Nonaka, 1994). These are patterns of knowledge creation, known under the acronym SECI: socialisation, externalisation, combination, and internalisation. This influential framework crucially revolves around the dynamic interchange between tacit and explicit knowledge, defining the mode of knowledge conversion. A specifically important notion is that the model suggests that each of the four modes is essentially based on social interaction between individuals within groups in the organisation. This is depicted in Figure 2, which is adapted from Nonaka & Toyama (2003). Nonaka (1994) describes the processes as follows. Socialisation revolves around an individual gaining shared experience from other individuals through direct practice or observation. Externalisation concerns the articulation and communication of personal perspectives triggered by meaningful dialogue. Social forms of interaction, such as meetings and conversations, are also at the basis of the combination mode, to exchange and reconfigure existing information. Finally, internalisation involves the embodiment of formal knowledge through learning-by-doing aided by interacting with other individuals in the same position. Hence, social processes are at the core of knowledge conversion, as they are embedded in all four modes. However, with the digitalisation trend in mind, the question arises how this will be affected by pervasion of digital technologies.
FIGURE 2
8
As knowledge and knowledge processes are intangible, they need to be materialised by KM tools (Sabherwal & Becerra-Fernandez, 2003). Such tools can be regarded as vehicles for knowledge processes, concrete practices and activities that are utilised to effectively manage knowledge. Examples of tools are team collaboration mechanisms, databases, cooperative projects across departments, and training (Sabherwal & Becerra-Fernandez, 2003). This paper identifies two crucial KM tools, being education and work instructions. First, training and education for shop floor workers are vital instruments in general knowledge processes: it provides the organisation with the opportunity to share and disseminate critical knowledge amongst its employees, and enables employees to create new or update existing knowledge. As such, it has the potential to embody all SECI processes. Internalisation can be represented by classroom learning (Asif, De Vries, & Ahmad, 2013), socialisation through on-the-job training (OJT) and mentoring (Nonaka, 1994). Training staff or mentors articulating their expert knowledge are the basis for externalisation (Sabherwal & Becerra-Fernandez, 2003), and combination can be rooted in the development of educational material. Education is thus an ideal channel that organisations can use for practicing KM. Second, manufacturing documents, such as manuals and operating instructions, possess similar potentials for sharing and using knowledge (Singh & Soltani, 2010). Critical knowledge can be externalised in and combined through work instructions (Asif et al., 2013). Such instructions also are instruments for internalisation as workers repetitively execute tasks according to those (Asif et al., 2013). Socialisation can be attained when operators explain or discuss instructions amongst each other. The importance of education and work instructions as concrete tools for enabling knowledge processes is thus explained by their potential for materialising SECI modes. This encapsulates this paper’s theoretical view on KM. The conceptual linkages as discussed in this section are presented in visual form in Figure 3. The question then rises how education and instructions facilitate KM to subsequently affect quality, especially when supported by digitalisation.
FIGURE 3
9
Linking KM to Quality
Quality is a very broad concept: a myriad of theories, perspectives, and definitions over time emerged from literature. To establish focus in this quality jungle, the aspects of quality are being conceptualised in light of this paper’s focus. This provides a guiding framework based on which the link with KM can be explained.
Conceptualisation of Quality
Quality is subject to diverse definitions. This paper adopts the view that quality is defined as conforming to set standards, specifications, and expectations (ISO, 2015a; Yong & Wilkinson, 2002): in other words it entails compliance to requirements. However, regardless of what definition is adopted, “the concept of quality rests on its management” (Ruževičius, 2006: 31). Quality management contains four main dimensions: planning, control, assurance, and improvement (ISO, 2015a). Central to that is the evolution of quality management (Yong & Wilkinson, 2002), that provides a framework on which the role of KM can be mapped. Martín-Castilla & Rodríguez-Ruiz (2008) illustrate this quality management evolution as ‘layers’ that expand the previous stage of the quality concept. In its most basic form it comprises inspection, from which the quality control layer follows subsequently. Expanding this view by emphasising prevention and embedding quality in products and processes, leads to the concept of quality assurance. Total Quality Management (TQM) is the next layer, in which continuous improvement became key. Finally, in its ultimate form, quality management constitutes excellence (EFQM, 2003). This model is shown and briefly elaborated in Appendix E.
10
are less efficient and make more errors. Instructional information thus needs to be of adequate quality (i.e. complete, correct, timely) to embed quality in manufacturing (Haug, 2015). This paper thus presumes that process and instruction quality are embodying the assurance and reproducibility of quality in manufacturing. Similar to the previous section, the conceptualisation of these quality elements is illustrated in Figure 4.The next section discusses the link with KM concepts.
FIGURE 4
Visualised Conceptual Framing of Quality Elements
Connection with KM
11
Linking Knowledge, Quality, and Digitalisation in Practice
The foregoing literature review clarified this study’s theoretical framing of elements of digitalisation, KM, and quality. Based on the advanced insights from this, the general research question proposed in the introduction is redeveloped to make a closer fit with the narrowed-down and conceptualised research topics as well as current academic gaps:
How can manufacturing organisations use digitalisation to effectively leverage knowledge processes, with the aim to enhance reproducibility through quality assurance?
The conceptual foundation of the research is twofold. At the core, it is about the theoretical relationship between KM and quality, as set out in this chapter. The novelty, however, lies in how digitalisation might impact this link. Figure 5illustrates these links in a basic conceptual model.
FIGURE 5
Basic Conceptual Model
12
▪ What role do digital systems and technologies play in the knowledge (conversion) processes? ▪ How can digitalisation contribute to balancing tacit and explicit knowledge in work instructions
and education?
▪ How do digitally-enabled knowledge processes drive the quality of processes and instructions? ▪ Which aspects of process and instruction quality promote reproducibility by assuring quality?
FIGURE 6
13
III.
METHODOLOGY
This chapter deals with explaining the adopted research approach and methodology. The following sections address two methodological aspects of studying the case described in chapter III. Firstly, it deals with the design of the study and proposed research methods. Secondly, measures for safeguarding quality in this research are discussed.
Research Design
As indicated before, a single case study was performed based on the case as described in the previous section. The choice of such a design for this study can be explained along the case research characteristics identified by Benbasat, Goldstein, & Mead (1987). First, it is most appropriate for early stages in building understanding on the topic, so exploration is key. It follows from the literature review that theoretical comprehension of the constructs and relationships pertaining to KM and quality assurance is still in its infancy when placing it in the perspective of Industry 4.0 and digitalisation. As such, explorative research is desired, as affirmed by the Industry 4.0 research agenda proposed by Oesterreich & Teuteberg (2016). Second, case research is valuable in answering “why” and “how” questions (Yin, 2013). Looking back to the research question posed, the current study aims for investigating how manufacturing organisations, through digitalisation, can effectively make use of knowledge in quality assurance. As clarified in the previous chapter, the case company is eager to understand how to deal with this in practice. This epitomises the third characteristic: the ability to study the phenomenon of interest in its natural setting. These three aforementioned characteristics are argued to be particular strengths of case research (Meredith, 1998), which made case research justifiable, as all three are present in the briefly introduced case company. Mainly due to the first point – the importance of exploration – a single case was chosen over multiple cases, to obtain greater depth and understanding in exploring the research topics. The case setting will be further clarified in the next chapter, which also elaborates on the specific unit of analysis.
Considerations for Data Collection
14
company were closely observed to obtain in-depth understanding of the case context, to be able to draw meaningful research conclusions from the empirical setting. Additionally, observations were made while attending and participating in workshops, corporate meetings, project team discussions, and other forms of daily communication. Furthermore, interviewing relevant stakeholders (e.g. operators, team members, and/or managers from respective units of analysis), aimed for collecting qualitative data on the multitude of perceptions, insights, beliefs, and knowledge of the people concerned with the Industry 4.0 project at the case company. This aided in interpreting the observations and to place those insights into perspective. Table 1 shows further details of the interviews, such as duration and interviewee information.In addition to interviews and observations, the research setting also provided the possibility to collect data from the three remaining sources as identified by Yin (2013): company documents, archival records, and physical artefacts. Where required, these were used to complement data obtained by interviews and observations, as to increase understanding of the case.
TABLE 1
Interview Details
INTERVIEW DURATION INTERVIEWEE FUNCTION / TITLE UNIT OF ANALYSIS
A 57 minutes Senior Production Process Specialist Manufacturing Engineering
B 20 minutes Production Engineer Manufacturing Engineering
C 16 minutes Pre-cutting Operator Shop floor operations
D 33 minutes Lay-up Operator Shop floor operations
E 38 minutes Debagging Operator Shop floor operations
F 22 minutes Machining & Benchwork Operator Shop floor operations
G 51 minutes Quality Lead Operations Quality department
H 50 minutes Quality Control Officer Quality department I 61 minutes Manager Learning & Development Training Centre J 42 minutes Project Coordinator Learning & Development Training Centre
15
Data Analysis Approach
Partly parallel to the raw data collection, the qualitative interview and observation data were prepared for analysis by careful documentation, i.e. transcribing the recorded interviews and adopting a structured approach for taking field notes. Analysing such qualitative data can be performed in various ways. This study adopted the following approach and procedures, which is comprehensively illustrated in Figure 7. A hybrid form of inductive and deductive thematic analysis was deployed (Fereday & Muir-Cochrane, 2006). Based on the first impression of the case as developed in the initial phases of data collection, preliminary themes were observed and identified through a priori coding (Miles & Huberman, 1994). As such, a deductive approach was adopted, as these codes were ‘fitted’ to the data by labelling relevant fragments and quotations that reflected the themes associated to the a priori codes. During the analysis, however, notes and transcripts were re-read and recoded: a priori codes were modified, merged, or deleted based on important concepts and phenomena that emerged from the data, or were complemented with newly emerging codes. In doing so, the deductive a priori approach is combined with this inductive technique (Gioia, Corley, & Hamilton, 2012). The resulting final set of codes was used to fully analyse the data, to find and link the key themes and patterns.
FIGURE 7
Data Analysis Process
16
to certain codes or quotations – that emerged during coding and analysing the data were logged by means of adding comments. Memos served for documenting some broader themes or theories that were relevant but not directly attributable through codes. Subsequently, all these data were analysed by looking for certain interrelations, patterns, and overarching themes. Then, structuring these data in a code book aimed for providing deeper understanding of themes and patterns in the data. This contributed to an improved comprehension of the relationships between the relevant constructs of KM, quality assurance, and digitalisation in the empirical setting. Hence, it enables the interrelations as proposed earlier in the detailed conceptual model map in Figure 4 to be either confirmed or refuted, or maybe even extended.
Ensuring Quality in Case Study
With respect to the data collection and analysis, it is of utmost importance that it is conducted properly to be able to generate solid results and draw useful conclusions. Voss et al. (2016) provide an overview of quality issues in case research, involving validity and reliability. These issues will be explained in brief, clarifying how they will be addressed in the case context.
17
IV.
RESEARCH SETTING
This chapter provides further background on the case company. Moreover, its specific digital context is explained, in which the research concepts have been examined. Hence, an initial analysis of the findings regarding digitalisation is performed, laying the groundwork for presenting and analysing the case study results in the following chapter, specifically regarding its effect on the link between KM and QA.
Case Background
The case company is an industry-leading organisation specialised in composite manufacturing. It is producing components for both commercial and military aircrafts. The case study is performed in one of its composite factories in the Netherlands. To remain global leader within the industry, the company gradually aims for creating a digital factory of the future by running an Industry 4.0 pilot (hereinafter: Pilot4.0). However, as the aerospace business is fairly conservative, the smart and high-tech innovations that Industry 4.0 promises, are rare in production processes of such firms. Therefore, the company initially concentrates on digitising their manufacturing environment. It is aimed for to make the complete workflow paperless, which paves the way for extending Pilot4.0 with more advanced technologies in the future. Pilot4.0 focuses on the production engineering and dispatching processes. Currently, within these processes three documents (BV, TPD, and PI; see Appendix C) are leading. The BV is a paper document based on a product order and is comprised of product specifications and required process steps for that product. For each of those steps a barcode links to the TPD, an electronic document in which detailed instructions on how to carry out corresponding tasks can be found. Some of the tasks are critical, the so-called quality items (Q-items) which are essentially checkpoints. The PI provides separate process instructions on these items, for which completion must be approved by an authorised operator. The TPD and PI are composed of mostly product-specific instructions, created by the production engineers. Together, the TPD and PI documents serve as inputs for registration purposes. The completion of process tasks and Q-items is currently manifested by physically signing or stamping the accompanying lines on the BV document. Because of the paper-driven production documentation, this way of working is far from responsive to adjustments in work instructions or product specifications. In the current situation, it often takes several days before such changes are being processed and actualised.
Digitalisation in the Aerospace Industry
18
Description of Pilot4.0
Although the use of digital technologies in manufacturing engineering and dispatching is not taken for granted in this industry, the case company has realised the current approach to the creation, registration, and documentation of production orders is far from optimal given the industry developments. To that end, Pilot4.0 merges and replaces the BV, TPD, and PI documents into a centralised digital knowledge base, the Manufacturing Process Designer (MPD). The pilot is initially rolled out on a particular part of the manufacturing environment, with a limited number of products. For this, the Apache program has been chosen due to its maturity and workable scale. This program’s relatively well-developed production documents as used and developed by Manufacturing Engineering (ME), are converted into digital form in the MPD. Next to the MPD, three additional applications are incorporated into the digitised environment. These involve the Job Manager (JM), Shop Floor Viewer (SFV), and Registration Manager (RM). The JM enables identification of activities and tasks for planned production orders, and dispatches these jobs to the operators. Contrary to the analogue situation, the JM makes the digitised manufacturing organisation task-focused. The SFV then facilitates the execution of jobs by the shop floor workers. The need for handwriting and stamps is eliminated due to the introduction of digital authorisation checks for Q-items and sign-offs. These provide the system with important hard controls in operator self-verification (OSV) of the executed tasks. Moreover, the SFV differentiates between worker competences to provide optimal support. Experienced operators receive instructions on the broader activity level, instead of per specific task. One of the fundamental features of the SFV is visualisation: the job instructions and bill-of-material requisites are complemented with illustrations or pictures, to most effectively support the operators. The RM finally collects all the digital data on the completed jobs. Additionally, this function creates comprehensive as-built reports, which would (in the future) also be suitable for advanced data analytics.
Case Findings on Digitalisation
19
standard elements enable modularisation, which facilitates the dynamic creation of work instructions. This is done through the MPD, with which appropriate operating procedures can be determined, based on the standard sets of process elements. These elements are still based on applicable specifications, customer requirements, specialised product and process knowledge, and operating procedures, but can now be more dynamically managed as they are not product-specific.
The digitisation of the manufacturing documents signifies the first step in moving towards a cyber-physical manufacturing system. The SFV within the Pilot4.0 system is the application that presents the digital work instructions to the operators. As such, it is a platform that links the digital world of ME with the shop floor reality, and thus could be regarded as the basis of a human-machine interface (HMI). Operators perform their registrations directly in the SFV, instead of manual stamps or sign-offs on paper registration sheets. As a result, this generates digital process data, which is logged in the RM application in as-built reports. Also, registered data on work-in-progress can be viewed in the SFV during the process. Through these functions, Pilot4.0 allows for direct collection of manufacturing data, as well as the potential to instantaneously analyse it. Although this cannot yet be classified as real-time data acquisition for cyber-physical manufacturing execution, the current digital foundation enables the future adoption of associated technologies. The real-time ability in Pilot4.0 can be regarded in two ways. First, it is indicated that the instantaneous instruction change process is an important aspect in the real-time aspect of the Pilot4.0 architecture. If a change is made in as-planned production data, this needs to be alerted to the operator in order for him to perform his tasks adequately. Another manifestation of real-timeness is in the production logistics. This is about generating immediate information on the product and process status, indicating where exactly the product is in the process at any given time. Contrary to the real-time change process, the logistic real-timeness is not yet fully supported in the Pilot4.0 systems. When integrations can be made with systems, machines, or data from the shop floor, this aspect can be realised.
20
V.
FINDINGS AND ANALYSIS
In this chapter, the findings of the case study are presented, regarding the observed phenomena about the effect of KM on quality within the digitalised context described above. Analysing these findings leads to the ultimate results that help answering the research questions. The presentation and analysis of the findings is organised as follows. The observations regarding SECI processes manifested in Pilot4.0 are presented, and their subsequent impact on instruction or process quality is briefly described. That way, insight is provided in how the case company capitalises on digitalisation to effectively leverage knowledge processes, and what the implications are for QA. This shows the empirical substantiation for answering the research sub-questions by interpreting the findings in the following chapter, to draw inferences about the effects of KM on quality and reproducibility.
Combination
Digitalised Combination Process
Primarily the creation and consolidation of explicit knowledge through combination thrives in the digitalised ME environment, with a key role for the MPD application. The process of converting the existing manufacturing knowledge to digital form, entailed a change in mindset towards document control in that the existing explicit knowledge within heterogeneous documents needs to be scrutinised. Specifications and procedures were critically reviewed, to determine whether they were still relevant and up-to-date regarding current shop floor practices. During this, social interaction remained essential for the combination process, as relevant insights and expertise from various parts of the organisation are shared during meetings of multifunctional and ME teams, uncovering areas of improvement. For instance, situations can occur in which certain tasks in practice were performed differently from what was described in instructions. This discrepancy then could for example be explained by individual best practices developed by operators based on their experience, or by the fact that new equipment is being used and ME did not update instructions accordingly. As such, comparing prescribed instructions to shop floor practice uncovered where specialised operator knowledge could be used to enrich the formal procedures. In other words, tacit operator knowledge is made explicit by finding inconsistencies with specifications and instructions, after which it is consolidated with pre-existing explicit knowledge by making use of production engineers’ expertise. As a result, the digital explicit knowledge base hence is dynamically improved and optimised.
Implications for Quality
21
The presented information is basically the same, but I think it is even better now than before. That is because by looking through a magnifying glass they examine what information they are issuing right now, which of that is still important and correct, and what is still to be added.
In addition to instructional quality in terms of correctness, the increased responsiveness of digitalised ME also improves timeliness and relevance of work instructions, because changes in specifications or customer requirements can almost instantly be processed and actualised in documentation. Also, in case of flawed instructions, the impact is reduced since issues can be resolved quicker. Altogether, this helps preventing situations in which operators work according to outdated or faulty instructions, since it is ensured that they are presented with information from currently applicable manufacturing documents. When operators adhere to these instructions, this also generates process quality benefits. That is, the execution of process steps is performed according to instructions based on optimised explicit knowledge that is known to lead to compliant products.
Internalisation
Effect of Digitalisation
Work instructions have been found to be an important mechanism through which operators internalise explicit knowledge on manufacturing procedures. As instructions are used repeatedly over time, the prescribed tasks and activities become familiar. The visualised SFV instructions ease the internalisation process, as it removes the need to read and comprehend pieces of text. Instead, operators unanimously indicate that picture-based instructions are easier to understand and use. Hence, digital work instructions are not only valuable in knowledge internalisation for apprentices, but also for experienced operators since they can see at a glance what is expected from them. They know the sequence of steps and required tasks in a job by heart, as epitomised by a statement of an operator during observation. While scanning through an activity’s instruction he already expressed which activity comes next, to prepare in advance:
I did not take a look yet, but I just know what to do based on experience.
Although experience generally is a valuable attribute, it also creates the risky situation that internalised knowledge might have become outdated or irrelevant over time without the operator being aware of that. For that, the vital role of training and education was stressed. Skills and expertise is taught in operator training, ensuring operators acquire and hold required knowledge relevant to their function. However, it is found that there is currently some serious misalignment between operator training and the shop floor, which is thus detrimental for operations relying on the firm’s tacit knowledge base. The misalignment is both in terms of gaps in prescribed information (i.e. neither covered by training nor by instructions), and in terms of taught techniques or procedures that are rendered obsolete on the shop floor. The observed reason is that either team leads do not clearly communicate the educational needs for their operators to the Training Centre, or they are not adequately aware the responsibility of composing the training packages is with themselves instead of with the Training Centre staff.
22
Implications for Quality
The Training Centre made a start with modernising the education, for instance by setting up an e-learning project. This development can be elevated by taking the digital explicit knowledge libraries as a basis. E-learnings, but also traditional training packages, can be assembled based on the standard explicit knowledge elements and be tailored according to the shop floor units’ needs. Yet, the absence of integration between the Learning Management System (LMS) and the digitised knowledge repository was frequently mentioned as hindrance to fully benefit from this. Currently, ME needs to request a list of operators that will be working with the F4.0 system, to manually indicate their OSV authorisation based on completed training. Hence, automating this process by connecting the LMS and MPD will both generate massive efficiency improvements and minimise the probability of human error in the manual process. For education, such digital connectivity ensures training and e-learning material is always kept up-to-date. Hence, critical operating procedures can be internalised more effectively. Together with the fact that error-prone physical stamps are replaced by requiring operator confirmation of within-tolerance work through digital signatures, the system thus can ensure tasks are performed by authorised personnel and only when previous steps have been executed according to specifications. As such, it can assure quality in the manufacturing process execution upfront.
Externalisation
Digitalised Externalisation
In contrast to the importance of combination, externalisation appears to be less crucial than anticipated in terms of its role in a digitalised environment. The SFV allows for dynamic and visualised work instructions, making certain product knowledge more externalisable. Certain skills and know-how of manufacturing techniques and activities can be codified by visualising it in the digital instructions. Currently, photos and figures serve this purpose by conveying the instructional message more efficiently than pieces of text. Also, including instruction videos is being regarded as highly desirable, as illustrated by this claim of Interviewee G:
I am strongly convinced that videos are more effective [than text or pictures]. If you need to know something at home you also barely read instructions, but instead look it up on YouTube.
This would bring the SFV to a next level towards HMI. However, the potential for tacit knowledge externalisation through visualisation by videos is limited. Product knowledge in terms of relatively basic manual skills – such as paint spraying, bagging material application, or equipment cleaning – seems to be most appropriate for video visualisation, but some activities (e.g. composite ply lay-up or visual surface inspections) simply require a lot of hands-on experience and complex fine motor skills which cannot be conveyed by video instructions.
23
Implications for Quality
As clarified in the first section, correctness and timeliness of the digital instructions is improved since they are based on optimally combined knowledge. However, the unambiguousness and completeness of instructions is evaluated less positively. In general, balancing informational completeness is critical as neither overburdening operators nor providing too little information brings quality in the production process. However, many operators expressed their concerns regarding the instructional completeness in the SFV. In that regard, deficiencies such as missing process steps or inspection points, involve distinct explicit knowledge gaps that can easily be overcome through the combination process in the MPD. The specificity of instructions (i.e. unambiguousness) however is operator-dependent, as not every worker would need the same degree of elaboration in instructional information. The SFV anticipates on this issue with the functionality to customise instructions’ level of detail to the experience of operators. In addition to this individual diversity, the differences between shop floor units are subject to criticism, as this quote from Interviewee H signifies:
Ideally, you want to guarantee that certain activities are executed. This is now lacking in the autoclave unit, whereas during lay-up every triviality needs to be documented.
These contrasts are largely caused by the required type of work and knowledge. The autoclave is a machine-dependent process, in which operators require process optimisation and decision-making capacities. The lay-up process, on the contrary, requires more operator-driven work and is thus highly reliant on manual skills and product knowledge of workers. Regarding this contrast, it is important to ensure operators hold and apply the appropriate type of skills and knowledge to achieve a good-quality process. In that respect, the findings indicate that instructions are the most suitable for conveying product knowledge, as this is better externalisable or already codified. Improved SFV instructions therefore ensure that operators are provided with required product knowledge, which reduces the probability of human error in the product work. Process knowledge, however, requires noticing and assessing a specific situation, and applying suitable decision-making and problem-solving capabilities based on experience. Such ‘deeper’ tacit knowledge is not (fully) transferable through codification in SFV instructions. Hence, these should be gained through experience, endorsed by other means such as social processes or training and education.
Socialisation
Socialisation Facilitated by Digitalisation
24
officer (QCO), and planner – and operator training programmes are at the core of the organisation’s approach to create and transfer knowledge. Currently, these tacit-to-tacit knowledge processes are presenting themselves through traditional and social infrastructures, involving mainly face-to-face or telephone contact in daily operations and classroom or on-the-job training (OJT), respectively. These flows of information and knowledge through interdepartmental lines of communication received some critical notes from interviewees. The interaction between ME and operations is generally of a unidirectional nature, in which information is being ‘pushed’ downwards to the operator. This causes problematic situations, illustrated by Interviewee E:
We encounter problems with production engineers who change equipment items, but do not update the associated images or even the text.
However, the Pilot4.0 systems have the potential to overcome such issues by supporting or modernising these mechanisms for cross-functional knowledge sharing through digital technologies. The SFV includes the so-called disruption functionality, in which operators might put the process on hold in case of any emerging issues – being a product defect, machine breakdown, quality deviation, or even an error in the instruction. Creating a disruption enables the operator to provide ME with the necessary information to solve the problem. Relevant feedback based on operators’ specialised knowledge and expertise can be shared with ME and QCOs, which might accelerate problem-solving and if required the documentation change process. However, the actual benefits of this are still largely hypothetical. In practice, the facilitation of knowledge transfer through this SFV function is limited: as the disruption functionality is only used in case of process issues, the system does not support general communication. Hence, when operators face common uncertainties, the need for sharing experiences and best-practices, or have any other queries, it is not possible to use the SFV for these issues. However, most interviewees indicated that they expected a chat function would be included in Pilot4.0. As it not yet within the scope of the pilot, people are slightly disappointed with the current improvements for this matter. This setback is part of more general disappointment with the low level of technological sophistication of Pilot4.0, as expressed by Interviewee H:
An issue about which many people are let down by Pilot4.0 is the amount of artificial intelligence and such, which I think is about zero.
The majority would encourage a digital messaging system in the SFV. From the perspective of a cyber-physical approach to knowledge sharing and conversion, an upgrade of the disruption functionality and the implementation of a chat function thus would be desirable.
Implications for Quality
Social processes, either supported by digitalisation or not, have been found to be critical in articulating and sharing experiences, values, and insights. Those partly define the culture of the firm, which is built on its shared values. The findings suggest these cultural and behavioural aspects are important in
“
25
pursuing quality through (social) knowledge processes. For example, negligence in registration is common for operators, for which it would be beneficial to keep drawing attention to the importance of quality commitment. This is exemplified by Interviewee G:
By communicating the importance of changing their mindset you hope they recognise why it is so essential. […] Now they might think, “as long as my product is fine, what is the point of that paperwork: it will soon disappear into someone’s drawer and we will never look at it again”. But that is not the case. Shop floor workers, but also production engineers, need to be aware of the importance of quality. Employees also should create a quality culture and mindset, and be encouraged to show behaviour that is in line with pursuing compliance. As this interviewee added, the value of technical hard controls can be boosted by simultaneously focussing on such cultural soft controls, which both contribute to quality in processes as operations are performed according to desired standards.
26
VI.
DISCUSSION
The case findings as presented in the previous chapter show how digitalisation in the case of Pilot4.0 has realised an improved interplay between KM and QA. This provides the empirical basis for answering the research questions. In this chapter, the essence of the findings is inferred and discussed. By seeking theoretical underpinning, it is shown how the case study results support or challenge the conceptual context as set out in Chapter II. In that regard, it is illustrated how this study added to the theoretical framing of digitalisation, KM, and quality. Finally, the academic and practical contributions are shown.
Knowledge Management
In the Industry 4.0 era, it is expected that the role of traditional or IT-based KM systems will be largely replaced by fully digitalised manufacturing systems (North et al., 2018). The case results showed insight in how this digitalisation will affect knowledge processes and conversion modes. In this section, these will be discussed in relation to theory, to answer the first two sub-questions.
Digitalised SECI Processes
The first question that is answered in this subsection, is what role digitalisation plays in knowledge conversion processes. In that regard, the case findings clearly showed that Nonaka's (1994) SECI modes can be effectively supported by digital technology. Combination is the most fundamentally affected by digitalisation, which is in line with earlier findings from Lee & Choi (2003) on the effect of IT. However, it was demonstrated that social coordination through for example functional team meetings and face-to-face discussions still play a role by uncovering where the explicit knowledge base can be improved or expanded. This is in accordance to Nonaka (1994), who posits all SECI modes are rooted in social processes. Hence, when both social and digital aspects are used together, an iteration of knowledge combination is engendered to optimise heterogeneous knowledge. This iteration is manifested by firstly creating and consolidating standardised explicit knowledge into a digital repository, and secondly combining these modular knowledge elements again into KM tools (Sabherwal & Becerra-Fernandez, 2003), such as educational material and work instructions.
27
repeated task execution according to instructions, as suggested by Nonaka & Toyama (2003). Through these channels, appropriate manufacturing procedures are ingrained in the operators’ working methods. In contrast, Haug (2015) raises his concerns regarding this learning- or knowledge sharing function of instructions, as he stresses it is not the purpose of instructions to aid in acquiring or internalising (new) knowledge or skills. This paper however contends Haug’s (2015) view: although it might not be the main purpose, clear beneficial side effects of instructions regarding KM have been found.
Externalisation of product knowledge, in terms of codification in visualised instructions, is facilitated by digitalisation. Findings suggest that video-based instructions are extremely valuable for translating tacit knowledge in codified form, consequently engendering internalisation. This is consistent with the notion that especially animated visuals improve task execution and subsequent learning (Ayres, Marcus, Chan, & Qian, 2009). Process knowledge, however, cannot be communicated in digital form, but ‘social’ externalisation processes (e.g. interaction in team meetings) enable sharing of such knowledge through articulation of views, experiences and perspectives (Nonaka, 1994). This may be supported by digital communication platforms, such as chat function (Sabherwal & Becerra-Fernandez, 2003; Qiu & Zhou, 2004). However, findings demonstrated that tacit knowledge can be made explicit by uncovering discrepancies with codified knowledge in instructions, based on social interaction in cross-functional teams. This suggests that creating and sharing tacit knowledge through socialisation is more effective than through externalisation (Haldin‐Herrgard, 2000), which calls for an emphasis on social interaction rather than on digital codification.
Despite the evident key role of social interaction, knowledge socialisation is suitable to be supported by digitalisation. In terms of (interdepartmental) communication, a chat messaging function would be useful to share individual experiences (Sabherwal & Becerra-Fernandez, 2003). Also, education can benefit from digitalisation. This is especially important in OJT, which is essentially a social form of on-the-spot education. Gaining hands-on experience through practice is a human-driven process, but the effectivity can hence be improved by enabling technologies. For instance, OJT requires less guidance of experienced mentors when virtual assistance devices such as AR/VR glasses (Richert et al., 2016) provide guidance based on standard procedures from the digital knowledge base.
Work Instructions and Education
28
codification) was found less crucial than anticipated, which is in accordance with Dooley's (2000) claims that tacit knowledge is increasingly critical to organisations as it is not easily imitated, because it is implicitly embodied in certain key individuals. Tacit knowledge on the other hand, can also be created and shared through education and instructions, as they aid the internalisation and socialisation processes.
Concluding Remarks on KM
The case findings in this study demonstrated that despite digitalisation offers new possibilities for managing knowledge, this does not necessarily mean that socially-embedded tacit knowledge loses its value if it cannot be codified in digital format. In fact, it was shown that social interaction remained decisive in digitalised knowledge processes. In other words, KM entails both a technical as well as a human side. This corresponds with the central paradigms of KM: a technologic, computational view based on empirically validated facts on the one hand, and an organic, cultural view covering tacit and explicit knowledge in human and social dynamics on the other (Argote, 2005). Given the context of Industry 4.0 and digitalisation, the technical paradigm is becoming increasingly relevant, as underlined by the growing importance of the role of IT in managing knowledge (Easterby-Smith & Lyles, 2011; Ragab & Arisha, 2013). However, the knowledge processes central to KM – in particular the SECI conversion modes as defined by Nonaka (1994) – are traditionally rooted in social interaction. This study’s findings reflect the critical role of social processes, but additionally suggest that digitalisation supports these by enhancing the efficiency and effectiveness of social mechanisms. Hence, instead of replacing the role human knowledge with technology through for instance Artificial Intelligence (Fast-Berglund et al., 2018), digitalisation is expected to establish the synergistical coexistence of both paradigms within Industry 4.0. Revisiting the central principles of the fourth industrial revolution, it is valuable to integrate the social and technical aspects of knowledge in a similar fashion as the virtual-physical convergence and human-technology interaction elements of Industry 4.0. In other words, it constitutes a more comprehensive socio-technical view of knowledge, comparable to the work of Prieto & Easterby-Smith (2006).
Quality Assurance
29
QA by embedding quality in the manufacturing process (Martín-Castilla & Rodríguez-Ruiz, 2008). In addition to the digital hard controls for these quality aspects (enabled by combination and externalisation), internalisation and socialisation are also crucial for providing soft controls. By engendering tacit knowledge and appropriate mindset, these soft controls reinforce hard controls by improving unambiguousness of instructions and encouraging compliance to manufacturing procedures. Finally, by discussing these findings, the last sub-question is answered about which aspects of process and instruction quality promote reproducibility through QA.
Considerations for Instruction Quality
From the findings, it can be concluded that correctness of work instructions is imperative to assuring quality. Also, considering the improved change process for manufacturing documents, timeliness is important as well. Through consolidating optimised explicit knowledge in digital instructions, these are based on correct and timely information. Hence, it can be better guaranteed that tasks are executed adequately through ingraining applicable SOPs. In addition, unambiguousness is a more complex factor when digitising the creation of work instructions. The issue of ambiguousness entails the person-dependent required level of detail in explaining how to execute an activity (Haug, 2015). Hence, thought should be put into finding the optimal balance between comprehensiveness and specificity, to effectively provide tailored operator support based on critical product and process knowledge. Consistent to Asklund & Eriksson (2018), it was found that dynamic digital instructions contribute to that since they prevent overburdening operators with information overload, while also avoiding presentation of too little information that impedes optimal work execution. To that end, involving operators in cross-functional teams would reveal instructional issues and thus provide useful input for determining the optimal level of instruction detail. Academics studying the link between KM and quality management support this view, as many proposed employee involvement as an important factor (Marchiori & Mendes, 2018). Work execution according to SOPs will be of higher quality and reproducibility, partially due to increased acceptance and motivation from involved and empowered operators (De Treville, Antonakis, & Edelson, 2005; Olson & Villeius, 2011). Moreover, independent of operator involvement, tailored education can bridge the gap of potentially incomplete instructions. Experienced operators might perform their tasks according to obsolete work methods, so especially refreshment trainings will ensure that even the most experienced workers internalised applicable SOP-based knowledge for adequate task execution (Olson & Villeius, 2011). Findings suggest this minimises registration issues, reduces quality deviations, and enables faster deficiency handling and -solving. On these grounds, instruction quality thus also affects the quality of the manufacturing process.
Process Quality Assurance
30
procedures. This emphasises the importance of SOPs, since these can improve the output consistency and learning rate of a given process, leading to variability reduction (De Treville et al., 2005). Moreover, Othman & Abdullah (2012) argue that adhering to SOPs together with tacit knowledge sharing improves the effectiveness of work. As such SOPs play an integral role in quality management (De Treville et al., 2005). Besides this positive influence of instruction quality on the reproducibility of production processes, process quality can also be assured by systematically applying quality tools such as SPC, monitoring scrap and rework of key processes, and applying error-proofing practices (Ahire & Dreyfus, 2000). The findings of this study suggest that digital production systems, such as MES, contribute to QA by providing the digital foundations for these process quality measures. Comprehensive and accurate manufacturing data that is collected in real-time or near-real-time, allows in-depth analyses to monitor quality deviations and to statistically control and improve the process (Saenz de Ugarte et al., 2009). However, especially in the aerospace industry manufacturing firms are reliant on analogue ways of working (Shan, Zao, & Hua, 2013), so full digitalisation needs to have priority before precipitating high-tech solutions. Thus, before the full benefits can be reaped, systems integration is required for the digitalised manufacturing system to be networked and interoperable with relevant legacy systems of the firm (Sjøbakk, 2018). Connecting educational information systems (i.e. LMS) with the digital knowledge base, allows optimisation of operator training packages. Only then, the benefits of digital hard controls can be obtained completely, as these soft controls encourage operators to actually comply with the instructed SOPs. Thus, a steady and good-quality process can be assured repeatedly.
Concluding Remarks on Quality
31
Influencing Factors
The results of this case study and the above theoretical underpinning have uncovered three important factors that were not specifically accounted for in the conceptual framing of this work. This is depicted in Figure 8, showing the insights this research has added to the conceptual model. First, as was most concretely discussed in this chapter, the significance of SOPs. These can help reducing variability (De Treville et al., 2005) for structurally assuring quality in processes (i.e. reproducibility). In non-digitalised manufacturing environments this is likely to be impeded by a lack of digital hard control and different ways of working between individuals. Such reproducible quality can be “built-in” by improved SOPs, where work instructions help spreading the standard explicit knowledge elements and education facilitates tacit knowledge sharing (Othman & Abdullah, 2012), hence encouraging SOP adherence. Good-quality instructions can assure process quality by making SOPs clearer and thus improving the adherence to them. Through this approach, repeatable compliance to requirements thus can be assured.
FIGURE 8
Research Contributions to Conceptual Model
