Impacts of the fourth industrial revolution on production networks of the Dutch manufacturing industry : a study into the Dutch Smart Industry program as it tries to facilitate the new industrial revolution and maintain

(1)

Impacts of the fourth industrial revolution on production networks of

the Dutch manufacturing industry

A study into the Dutch Smart Industry program as it tries to facilitate the new industrial revolution and maintain the international competitiveness of the Dutch manufacturing industry

Author: Friso Prins (10323643)

August 9

th

_{, 2019}

Supervisor: Niels Beerepoot

Master Human Geography – Track economic geography

University of Amsterdam

(2)

2

Abstract

This research examines the expected changes in production networks in the Dutch manufacturing industry as a result of developments in the fourth industrial revolution. The aim of this research is to provide real life context to a still very academic debate. This revolution is characterised by very rapid developments in ICT and production technologies, meaning that production is becoming increasingly digitised and automated. It is defined as a new industrial revolution as the velocity, scope and disruptiveness of these developments are unprecedented. The Netherlands, and so do many other countries, recognises the necessity to keep up with these developments to maintain economically competitive and to address major societal issues such as climate change. In order to coordinate the transition to Industry 4.0 in the Netherlands, a platform called Smart Industry was founded about five years ago. The platform raises awareness of the necessity of the transition within the Dutch industry, and stimulates the development and implementation of new technologies through its 39 field labs. These field labs are generally a collaboration of the private- and public sector and education institutions, and are aimed at developing and implementing new technologies ranging from 3d printing to safe data exchange. The Smart Industry platform was key in this paper, as my analysis was based on Smart Industry policy documents and interviewing field lab representatives. From an economic geographical point of view this research is focused on the changes in the organisation of production networks. It is also investigated how much 3D printing is influential in these changes, now and in the future. The main findings are that due to an increasing digitisation of production, the production stages within a network are likely to be much better connected. Meaning that production processes are becoming much more flexible, and producers and supplier can quickly adapt to changing customer demands. Although 3D printing does fit perfectly in a flexible production process, it currently is too limited to be a significant impact. But as the 3D printing keeps on

developing, it might change production paradigms in the future. Furthermore, an integration of services and industry is occurring. Due to increasing data flows, producers are more capable of providing (after-sale) services. Additionally, new roles that mostly provide services to producers emerge within production networks. Some things remain unclear however, different opinions and expectations are expressed with regards to fractionalisation of production networks and the effects on the labour market as a result of Industry 4.0.

(3)

3

Acknowledgements

I would like to thank all my respondents who took the time and effort to share their knowledge and experiences with me. Furthermore, I want to thank my supervisor Niels Beerepoot who was very supportive and was always willing to provide great feedback to my work throughout my research.

(4)

4

We are on the brink of a fourth industrial revolution. This revolution, just like the previous three, will fundamentally change industry and impact economies and societies as a whole (Lasi et al., 2014; Li et al., 2017; Schwab, 2016; Ślusarczyk, 2018; Tjahjono, 2017). As everything is so widely connected nowadays, it will do so from a regional to a global scale (Schwab, 2016). This revolution can also be described as a digital revolution and is increasingly integrating the human world with the digital world (Schwab, 2016). Although the existence of the fourth industrial revolution is now widely accepted, it is still unclear what its impacts are. There are a lot of challenges and opportunities that arise from these technological developments. A primary example is the digitisation of services. From ordering a taxi, listening to music, or doing your groceries, it can all be done online and remotely. The hard part is to keep up with all current technological developments as they follow each other up at a very rapid pace (Brynjolfsson & McAfee, 2014; Schwab, 2016).

Due to the fact that the producer and consumer are increasingly connected, the consumer demand is becoming ever more influential. New technologies allow mass production to be evolved into mass customisation meaning that products can be customised to meet personal demands, yet still be mass produced (Hannibal & Knight, 2018; Smart Industry, 2014a; Shrouf et al., 2014; Janssen et al., 2014). This revolution is changing the relationship between the consumer and the producer. Moreover, a huge challenge will be the distribution of wealth, from a regional to a global scale. It is widely

accepted that a lot of (middle class) jobs will disappear as a result of new technological development, and that the income distribution between owners of labour and owners of capital is changing. This is leading to a new division of labour (Brynjolfsson & McAfee, 2014). We can already see this trend occurring in primarily the developed world, and it is likely that this will continue at a rapid pace in the near and distant future. Or as Brynjolfsson & McAfee (2014: p.11) put it: ‘Technological progress is

going to leave behind some people, perhaps even a lot of people, as it races ahead.’

There is a need for a wide body of research in order to “map” these developments, so that

governments, companies and knowledge institutions can create policies to exploit opportunities, but also to address challenges that arise from the fourth industrial revolution (Xu et al., 2018). A lot of studies discussing the fourth industrial revolution the geographical impact tends to be overlooked. The spatial consequences for supply chains is often underestimated or ignored. In Brynjolfsson & McAfee’s much discussed but also widely accepted book ‘The Second Machine Age’ (2014) for example, the focus is on the impacts of technological advancements on labour and a (global) wage gap, but it barely mentions anything on spatial changes in supply chains and its consequences. The role of geography is often being neglected by researchers. This research will therefore focus on supply chain changes as a result of new technological developments. The connection between technological advancements and supply chains changes has so far been undervalued in academic literature (Ancarani & Di Mauro, 2018; Hannibal & Knight, 2018).

Most of the Industrial powers in the world have developed programs to develop and implement the fourth industrial revolution technologies (Santos et al.,. 2017). The most common term is industry 4.0 which was introduced in 2011 by German representatives of business, government and science as the way to future competitiveness for German industry (Lasi et al., 2014; Li et al., 2017; Ślusarczyk, 2018). A few countries already had such a program in place, but many followed the industry 4.0 concept. Including the Netherlands which founded a Smart Industry program, this is a collaboration of public- and private actors and knowledge institutions (Smart Industry, 2014a). Smart Industry highlights that the high rate at which information and communication technologies are being

(6)

6 developed is leading to widespread impacts on the Dutch economy and society. The Dutch

manufacturing industry is the third fastest growing manufacturing industry in Europe behind Germany and Austria. Although this sector is still relatively small compared to other European countries, it highlights the importance of the manufacturing industry for the Dutch economy (Smart Industry, 2014a). Facing the challenges and harvesting on the opportunities that arise from the new industrial revolution is crucial for the international competitiveness of the Dutch manufacturing industry and economy as a whole (Smart Industry, 2014a; 2014b). For an overview of the European programs see figure 1.

Although this research will be focused on smart industry in the Netherlands, it has to be realised that the effects of industry 4.0 are never isolated on a local scale. If something changes at one end of the supply chain, it will lead to changes at the other end as well. Industry 4.0 will lead to fundamental changes in supply chain structures, and the Dutch government is fully expecting that as well. It expects two major changes in supply chains. An increasing upstream supply chain integration and the reshoring or backshoring of production (Smart Industry, 2014a).

This research will be looking into the current ongoing changes in supply chains, and the expected changes that will occur in the (near) future. As industry 4.0 impacts so many sectors in the economy on different scales, a specific manifestation of industry 4.0 is highlighted in this research, namely 3d-printing, also referred to as additive manufacturing. Its development is still in its early stages, yet it is predicted that this will cause a shockwave through global supply chains (Baldwin, 2012; 2016; Hannibal & Knight, 2018). A TNO report provides a detailed definition of 3d-printing: ‘3-D Printing is

the “process of joining materials to make objects from 3-D model data, usually layer upon layer.”(…) The 3-D printers deposits microscopically thin layers of the raw material, and the print gradually materialises as the layers are built up step by step (Janssen, 2014: p5). In short, products are digitised

and are then printed out. This means that every product that is printed out can be customised to a customer’s needs, and through the digital world these orders can be placed at any time, from

anywhere in the world. Especially in the world of manufacturing it might have a significant impact on global value chains. As Smart Industry is heavily invested in developing the Dutch manufacturing industry, including 3d printing technologies and applications, it allows me to gain a good insight of the current and future developments and impacts of this specific manifestation into the fourth industrial revolution.

(7)

7 This thesis is structured as followed. First, an extensive literature review on supply chains and the fourth industrial revolution is presented. Next, the research methodology, the research questions and the operationalisation of the main concepts are presented. Further, we discuss the results of the research. The final section concludes the research and will include a reflection on the study.

2. Theoretical Framework

To develop a broader understanding of industry 4.0, an extended literature review will be presented. It is important to clarify what defines the fourth industrial revolution, and why it is revolutionary. Firstly, I will elaborate on the context in which industry 4.0 occurs by explaining previous industrial revolutions and the 2nd_{unbundling. Secondly, I discuss the definition of the fourth industrial}

revolution and why it can be qualified as a revolution. Additionally, the possible impacts on

globalisation and supply chain structures and possible industry 4.0 challenges are discussed. Finally, I will go into the role of 3d-printing as a manifestation of Industry 4.0 and supply chain networks.

2.1. Global value chains / Global production networks

The most obvious economic geographic perspective into the fourth industrial revolution is to look at developments in global value chains (GVC). A good starting point is to quickly define a supply chain and a global value chain. A supply chain is quite a general term to capture in- and outputs to create and sell a product, reaching from raw materials to marketing and sales. A global value chain is very similar to a supply chain, but tends to focus more on the relative value added to the product in every stage of the production process (Gereffi et al., 2001).

In this chapter I will start of by briefly describing the previous three industrial revolutions to highlight what qualifies an industrial revolution. Additionally I will explain GVC trends from the past decades to provide some context in which the fourth industrial revolution is occurring. These trends are defined by what Baldwin (2011; 2012) refers to as the 2nd_{unbundling. I will elaborate on the 2}nd_unbundling

in chapter 2.1.2.

2.1.1. The first, second and third industrial revolution

In order to understand why we are currently facing a new industrial revolution, it is important to look at what qualifies an industrial revolution. Therefore I will briefly provide an overview of the previous three industrial revolutions, followed by an explanation of why the fourth revolution can be qualified as such. The first revolution is characterised by the mechanisation of production trough the use of wind and steam power (Blinder, 2006; Schwab, 2016; Ślusarczyk, 2018). The first revolution led to fundamental changes in the economy and society. As large factories started to arise, there was a huge shift in demand from agricultural labour to industrial labour. This caused unprecedented urban growth in many Western countries. It also changed education, government policies and the

organisation of businesses (Blinder, 2006). Before the first revolution production and consumption had to be close to each other due to a lack of sufficient long-distance transport technology. This changed through the invention of especially the steam locomotive and steamboat. The world was transforming due to the separation of production and consumption. This is what Baldwin (2011; 2012) refers to as the 1st_{global unbundling. It led to globalisation, emergence of economies of scale}

and clustering.

The second industrial revolution is characterised by the invention of a lot of industrial technologies, most important being electricity. These technologies lead to mass production and standardisation of the production process which lead to a new division of labour, where every worker did just a part of

(8)

8 the entire production process, rather than building the entire product by himself. This caused a massive increase in the production efficiency and labour productivity. (Atkeson & Kehoe, 2001; Blinder, 2006; Schwab, 2016; Ślusarczyk, 2018). The key feature of the third industrial revolution were the developments in communication technology in the second half of the 20th_{century (Schwab,}

2016; Ślusarczyk, 2018). This revolution is therefore also called the digital or the ICT revolution (Baldwin, 2012). Exchanging information became much easier, cheaper and quicker through the invention of primarily the computer and later on the internet. Coordination costs limited the separation of production and consumption before, but now that the coordination of a production process or a supply chain could be done from anywhere in the world, it was much less of a barrier to separate production from consumption anymore. This meant a huge shift in which manufacturing increasingly moved from industrialised countries to low-wage developing countries (Strange & Zuchella, 2017). This is a process called offshoring and what Baldwin (2011; 2012) calls the 2nd_global

unbundling. Or as he puts it: ‘2nd_{Unbundling: ICT made it possible, wage differences made it}

profitable’ (2012: p4). This will be elaborated in the next section. So these three industrial revolutions

completely altered the nature of production, and therefore fundamentally changed economies and societies across the globe. A brief overview of the revolutions, including the fourth, are given in figure 2.

2.1.2 The 2nd unbundling

A reduction in trade costs characterised the 1st_{global unbundling, but this remained more or less the}

same during the 2nd_{unbundling. The key feature of the 2}nd_{unbundling was the reduction of}

transmission costs through the ICT revolution (Baldwin, 2011; 2012). In other words, it became easier to coordinate production networks over a global scale, which allowed for a more specialised and more geographically dispersed network (Gereffi, 2014). It is important to note however, that international offshoring also means the transfer of knowledge to more links in supply chain. ICT developments lead to the ability to control and therefore reduce the risk of unwanted knowledge spill over to keep a competitive advantage (Baldwin, 2011; 2012).

The process of the 2nd_{unbundling can be divided into two separate broad steps. The first step is}

fractionalisation, which means a specialisation of the supply chain, and leads to more links in the chain. The other step is offshoring / international dispersion of supply chain links. Although often both go hand in hand, the decision to either specialise or offshore is made separately (Baldwin, 2016). These decisions are largely based on costs, and these separation costs are based on moving

(9)

9 people, ideas and goods across the world. It is likely that the cost of moving goods will not change much the coming years. However, the costs of moving people and ideas are increasingly reduced by the current rapid developments of ICT technologies. Due to the digitisation of machines the need for human interaction is decreasing, so fewer people have to be moved around. Additionally, Baldwin expresses the importance of the development of video and audio technology, so that people don’t have to be in the same room with each other in order to have a face to face meeting.

The second unbundling is also characterised by ‘smile curve economics’, which means that the value added of each production stage is becoming less at offshored stages (Baldwin, 2012). On the one hand this is basic economics, production costs drop at offshored stages and therefore its added value drops. On the other hand it is because in today’s economy, design, sales and (after-sales) services add more value than the assembly of a product. These stages are often not offshored, and therefore remain primarily in developed countries. This means that the value added at start and end of a value chain is usually higher than the assembly stages in the middle. This is illustrated by the smile curve in figure 3.

Due to the continuous decline in costs of moving people and ideas around, Baldwin (2012; 2016) expects globalisation and unbundling to continue. And even though the wage gap with previous ‘offshore destinations’ is decreasing, there are still a lot of low-wage countries in the world where supply chain stages can be moved to. Therefore offshoring is likely to continue. This prediction is strengthened by the fact that a lot of countries became wealthier by joining supply chains after the ICT revolution (Gereffi, 2014), which lead them to become promising consumer markets as well (Baldwin, 2016). This results in a deepening of the smile curve, especially as consumers are increasingly influential in the design stages and companies are increasing their portfolio by adding services to remain competitive. Industry 4.0 does not just facilitate these trends, it is a main driver behind them (Smart Industry 2014a). Furthermore, through the increased connection between supply chain stages, it is expected that companies within a chain are more concerned about the added value of the entire chain, rather than their own added value (Smart Industry, 2014a).

(10)

10

2.2 Defining Industry 4.0

According to Schwab (2016) we are currently at the start of the fourth industrial revolution. Current technological developments are not just continuous ICT developments from the ICT revolution, they are causing new fundamental alterations to industry. Schwab (2016) names three reasons for this. The first reason is velocity. The technological advancements follow each other up at a rate that the world has never seen before. Whereas the previous industrial revolution developed at a linear pace, the current one is developing at an exponential rate. The second reason is scope. Nowadays there are billions of people connected to the digital world, mostly through their phones. Any ICT

developments are therefore likely to spread quickly. Furthermore, it is expected that this revolution will basically disrupt every industry everywhere around the world. The last reason he mentions is

systems impact. The new technologies of this revolution are so radical that it will change how

production systems are managed, and how governments design their policies regarding production systems. In short, the structure of production networks will go through significant changes (Schwab 2016;Xu et al., 2018).

Now it is established why we are at the start of a new industrial revolution, it is important to define the key characteristics of Industry 4.0. The first thing that is prevalent about most industry 4.0 explanations, is the application of Cyber Physical Systems (CPS). This means the increasing

integration of humans and machines in a production process. This also includes that machines are increasingly communicating and sharing information with each other, and take action according to it without human intervention (Li et al., 2017; Ślusarczyk, 2018; Shrouf et al., 2014; Thahjono et al., 2017). The integration and development of digital and physical technologies are two key drivers of the fourth industrial revolution. Digital developments include for example the Internet of Things (IoT) through which machines and humans communicate with each other, Artificial Intelligence (AI) and machine learning and the use of big data. Physical developments are for instance autonomous cars and 3D-printing (Brynjolfsson & McAfee, 2014; Li et al., 2017; Ślusarczyk, 2018). Li et al (2017) also mentions biological developments such as genetic engineering as a key feature of this revolution, but this is not relevant for this research so it will not be discussed.

As industry 4.0 encompasses so many things and technologies it is also quite broadly defined, but according to Ślusarczyk (2018) there are in general three main characteristics: Firstly, The increasing digitisation and enabling of constant communication between and amongst humans and machines. Secondly, The application of disruptive innovations, which are developed at an increasing rapid pace. The most prevalent example is 3D printing, which is discussed later in this paper (see section 2.3). This means that the operation of socio-economic systems are becoming ever more effective and efficient (Baldwin, 2016; Brynjolfsson & McAfee, 2014; Schwab, 2016) Within this second characteristic lies the main focus of my research as I will investigate the expected impacts on production networks of improved ICT and production technologies.The third objective is the increasing implementation of AI and machine learning in order to create a largely autonomous production process. (Ślusarczyk, 2018).

2.2.1 Impacts of industry 4.0 on globalisation and global supply chains

Industry 4.0 will build on, and contribute to this its global impact. Through the rapid development of the Internet of Things (IoT) the management of production chains will fundamentally change. Because of this development the coordination of product- and information flows within supply chains doesn’t have to be synchronised anymore. This process is becoming increasingly automated. Machines within and between production stages are linked to critical information of products, such as their use and destination, within the production network. This will lead to an improved

(11)

11 production- and distribution efficiency of international supply chains. It is therefore expected that supply chains will increasingly fractionalise (Baldwin, 2011; Strange & Zuchella, 2017). Despite the prediction that global value chains will continue to globalise, a contrary trend is being noticed and researched. This trend is reshoring (sometimes referred to as backshoring). Reshoring means that manufacturers are bringing production stages that they offshored, back or close to their home country (Ancarani & Di Mauro, 2018; Stentoft et al., 2016). Research that was conducted in the past ten years shows that there are a numerous amount of drivers for reshoring. Ancarani & Di Mauro (2018) found that the most important reshoring driver is changing costs such as increasing labour costs in offshoring locations and rising coordination and transaction costs. Other important reasons for reshoring are issues with the quality of production, the time it takes to produce through several production stages, a lack of flexibility of the value chain and the limited access to knowledge and skilled personnel. Additionally, the value of local branding (Made in Holland rather than Made in China for example) and the threat of losing knowledge are also mentioned as drivers for reshoring (Stentoft et al., 2016). However, offshoring is still very likely to continue in near future. Reshoring is not yet a disruptive shift into a different direction (Ancarani & Di Mauro, 2018). But a lot of the above mentioned issues that arise with offshoring production can be addressed by industry 4.0

technologies. Automatization and robotization reduce the need for low-skilled labour, and

digitisation leads to an increase in flexibility and a decline in coordination costs. Therefore it is likely that reshoring will not just gain momentum, it will also lead manufacturers to decide to not offshore at all (Baldwin, 2016; Brynjolfsson & McAfee, 2014).

Another expected trend of industry 4.0 is accelerated value chain integration, which means that two or more links in the chain will be combined to one. This trend has been going on for some while, but new technologies are likely to accelerate this (Santos et al., 2017; Smart Industry, 2014a). A simple example to highlight value chain integration is that in recent years Dutch supermarkets not only sold their products in their stores, they started delivering them too. Due to the integration of the human and physical world through digitisation, the customer has a much more direct connection to a production process. This leads to more consumer demands, but also quicker responses from the value chain to these demands. This concept is known as mass customisation (Hannibal & Knight, 2018; Janssen et al., 2014; Santos et al., 2017; Shrouf et al., 2014; Smart Industry, 2014a). Industry 4.0 facilitates this because new technologies will be able to mass produce “small batches” to facilitate increasing consumer demands. An example of such a technology is 3D printing or additive manufacturing (Ancarani & Di Mauro, 2018; Baldwin, 2016; Brynjolfsson & McAfee, 2014; Strange & Zuchella, 2017). I will elaborate on 3d-printing in section 2.3

2.2.2. Industry 4.0 challenges

As the fourth industrial revolution is still in its early stages, there is still a lot of uncertainty about its impacts. Although there is a lot of excitement about the new opportunities, there are also some possible drawbacks and uncertainties. Probably the biggest challenge is the loss of jobs due to automation. This trend has been steadily going on for a while now, but it is expected to intensify in the industry 4.0 era. The digitisation of human tasks has been growing at a rapid rate in the past few years, and will continue to do so in the (near) future. This digitised tasks are then performed by robots as robotics technology developments have also picked up the pace (Autor, 2015; Brynjolfsson & McAfee, 2014). The tasks that are the most easily replaced are repetitive tasks and routine work. In combination with the fact that state-of-the-art technology is becoming ever more important in a production process, the need for low skilled workers is diminishing, whilst the demand for high skilled workers is increasing (Autor, 2015; Brynjolfsson & McAfee, 2014). This is also an important trend in the light of reshoring. Reshoring decisions are influenced by the degree of available

(12)

12 resources, labour and suppliers for the production process (Ancarani & Di Mauro, 2018). This

emphasises the need to educate current and future workers, in order to successfully transition towards industry 4.0 (Santos et al., 2017). A lot of technological innovations never reach their full potential because there is not enough attention spend to educate workers to work with these technologies. Technological innovation also requires social innovation (Buhr, 2015; ECBI, 2018). However, not every lower skilled task is being automated. Important distinctions have to be made whether a task is cognitive or manual, and if it is a routine or a nonroutine job. Any routine tasks are very susceptible to computerisation and therefore automation. These are mostly middle-class jobs and the demand for these workers is therefore declining. Any nonroutine jobs, ranging from a financial analyst to a janitor, are much harder to digitise. This leads to a hollowing out of the middle class (Autor, 2015; Brynjolfsson & McAfee, 2014). Autor agrees with Brynjolfsson & McAfee on the argument about the polarisation of the work floor, however he states that new technologies do not necessarily lead to a loss of jobs altogether. He argues that the loss of routine jobs, leads to a simultaneous growth of high skilled jobs at one end, and low skilled jobs at the other (Autor, 2015). Additionally to the loss of middle-class jobs, Industry 4.0 will also lead to a fundamentally different allocation of wealth. Industry 4.0 technologies are mostly developed at a relative low cost and create a huge increase in wealth, but this wealth is unevenly distributed. People that are competing with robots and computers either lose their job, or have to do the job for a low income to remain competitive. Meanwhile people who have more human and/or physical capital will receive most of the new wealth created. This causes an ever increasing gap between the rich and the poor.

Brynjolfsson & McAfee (2014) highlight that America has gotten much wealthier in the period of 1983 till 2009, yet this wealth increase was only for the top 20% of incomes in the US. The other 80% actually witnessed a decline in their wealth. This trend is not only seen in the US , but also in many other Western countries including The Netherlands (Autor, 2015). Another Industry 4.0 challenge next to job polarisation and the increasing wealth gap, is the massive amounts of data that become available due to digitisation and rapid developments in the IoT. The availability of big data allows international businesses to track emerging trends and highlight opportunities in markets across the world. This comes with two potential drawbacks. Even though there is a lot of data available, firms need to have the capabilities to analyse and utilise this data to reap the benefits. The second

drawback is the threat to the privacy of individuals. The more valuable big data becomes, the further firms are willing to go to obtain it. It is necessary to regulate and protect the privacy of individuals by sufficient (international) governance (Strange & Zuchella, 2017). This also stresses the need for increasing cybersecurity capabilities. After all, new technologies can not be utilised if the

infrastructure on which these technologies are build are not protected (Santos et al., 2017; Smart Industry, 2014b).

2.3 The emergence of 3D printing

A technology shock is a technological development or invention that has the possibility of disrupting existing markets and value chains. Through the creation of (new) products and services that are significantly better and cheaper, these technologies create either new markets, or disrupt existing ones. This might even lead to a changing production paradigm (Hannibal & Knight, 2018). Such a technology might be Computer Integrated Manufacturing (CIM), which is likely to have a large impact on global supply chains (Baldwin, 2012; 2016). Developments in this technology means that the work relationship between humans and machines is changing. Machines used to help humans to make things, but it is turning around where humans are helping machines. This means that an increasing amount of tasks are being automated and computerised (Baldwin, 2012; Brynjolfsson & McAfee, 2014).

(13)

13 These do not just include manufacturing tasks, but for instance also include design and engineering tasks (Baldwin, 2012). The five key features of CIM are:

1) Developing new products will be much more efficient time and money wise; 2) A shift from mass production to mass customisation;

3) Coordinating (international) value chains will be much easier, it is therefore likely that value chains fractionalise and specialise even more. Contrary there is also a belief that international dispersion may decline as a result of 3D printing, as it involves so many automated tasks. Not only can several chains within a supply chain integrate, the labour demand will also decline. So there is much less of a desire to offshore as location decisions are more based on customer proximity rather than production costs (Brynjolfsson & McAfee, 2014; Hannibal & Knight, 2018; Santos et al., 2017; Strange & Zuchella, 2017; Janssen et al., 2014);

4) Several tasks that require varying levels of skill that were previously performed by workers, are now undertaken by computers and machines.

5) A more polarised work floor (Autor, 2015; Baldwin, 2012; See chapter 2.2.2).

The prime example of CIM is 3D printing. For now it still has a limited impact due to the fact that it is still very time consuming and not suited for mass production. Additionally, only a limited amount of materials can be used for 3D printing and it is still lacking engineering precision (Strange & Zuchella, 2017; Janssen et al., 2014). Currently it is mostly used in niche markets or for building protypes. Yet it is predicted that 3D printing will be the prime driver for mass customisation and will therefore change the ways of global production (Baldwin, 2012; 2016; Gress & Kalafsky, 2015; Hannibal & Knight, 2018). A research conducted by ING Bank (2017) estimates that by 2060, or possibly even as soon as 2040, half of all products are made by 3d printers. 3D printing might cause a new production paradigm as it digitises physical goods. This means that physical goods can be “downloaded” all across the world and can be printed out. However, the broad access to 3d designs is also a reason for concern. It serves as an opportunity now, but might become a challenge later on due to the fact that it is very hard to copyright 3d designs. Copyrights are easily avoided by making minor adaptations to a design. This is an opportunity now as it results in open-source innovation where a lot of people can contribute to the development of this technology. However the 3d printing sector will likely face the same issues as the movie and music industry, namely the illegal downloading of designs (Laplume et al., 2016). Still, the digitisation of physical goods will have huge consequences for manufacturing, logistics and supply chain management of MNEs (Hannibal & Knight, 2018). It brings manufacturing much closer to the consumer. Hannibal & Knight even go as far to call this development a possible

‘disruption to manufacturing industry equalled only by the industrial revolution’ (2018: p.1118). It is

generally expected that manufacturing will continue to fractionalise and globalise as a result of the fourth industrial revolution (see chapter 2.2.1). But 3D printing has the potential to reverse these trends. It will reduce the need for offshoring and fractionalisation of the value chain. It may even lead to a “de-globalisation” of global manufacturing, although this may vary between Industries and different types of products (Hannibal & Knight, 2018; Laplume et al., 2016). To different degrees some industries are already impacted by 3d printing, and some industries are predicted to be impacted in the future. The impacts on global value chains is still unclear at this point, so more research is required to understand the spatial consequences for manufacturing supply chains.

(14)

14

3. Methodology, research and data

In order to execute this research it is important to determine the best research methods for this study. Qualitative research methods are chosen to research the impacts of industry 4.0 on

production networks. This chapter provides an overview of the research questions, a justification of the selected research methods, how to realise a sufficient level of validity and reliability, and finally ethical considerations are discussed.

3.1 Research Questions

The following research question will guide my research to study the changing supply chain structures as a result of the fourth industrial revolution.

‘How are production networks within the Dutch manufacturing industry expected to change as a result of the fourth industrial revolution?’

To address the main research question, I employ the following sub questions:

1. How is the interaction amongst companies, and also between companies and consumers within a production network expected to change as a result of automation and digitisation of the Dutch manufacturing industry?

2. To what extent is 3d-printing currently a disruptive technology in the Netherlands, and what are the short- and long-term impacts on production networks?

3. What main challenges are expected to arise in the transition to the fourth industrial revolution, and in what limitations do these challenges result?

With these sub questions I will look into the changing relationships between companies in a production network, as a lot of their interaction is expected to be standardised, digitised and automated. I want to describe the degree of this change at this moment and the expectation for the future. This also includes the expected changing relationship between consumer and producer. Furthermore, from the literature review it becomes clear that 3D printing is likely to have a huge impact on these developments, so it is relevant to find out how 3D printing is manifested in the Netherlands. Finally I will provide a more critical note on these development by looking into challenges and limitations of the fourth industrial revolution.

3.2 Qualitative research methods

In order to address the research questions, the research methods need to correspond with these questions and the aims of this study. As Smart Industry is still a relatively new phenomenon, the main goal is to add relevant research in a current academic debate. This is mostly an exploratory study to map the debate on Smart Industry development in the Netherlands, therefore this research is qualitative and mostly inductive in nature (Boeije et al., 2009; Bryman, 2012). It is mostly inductive because although this research is built on a theoretical framework, the theories discussed do not predict the same outcome. Therefore, the approach is not to test a theory, but to understand an aspect of the fourth industrial revolution. Consequently, a case-study design is employed (Bryman, 2012). A quantitative approach would not be feasible as it is very hard to determine the limits and barriers of a supply chain and of a new industrial revolution, especially when the research is limited

(15)

15 to a single country. Additionally, qualitative research allows for the possibility to uncover unexpected findings which could add to the understanding of industry 4.0 (Boeije, 2010). This research is

conducted in the Netherlands mostly for practical reasons concerning time, money and language constraints. In the introduction of this research it became clear that most Industrialised and Western countries have programs in place to develop and facilitate the fourth industrial revolution in their respective countries, therefore the results of this study might be applicable to other countries as well. Although It must be mentioned that not every country’s industry is similar and their ‘Smart

Industry platforms’ might therefore have a different focus which could lead to different outcomes.

3.3 Data collection

To increase the confidence in the results of this research, it is important to collect data through more than one research method, a process called triangulation (Boeije, 2010; Bryman, 2012). The several methods can add to, and verify the results of the different methods used. Triangulation is also used as a tool to increase the validity of a study. I will elaborate on the research validity in section 3.3. To collect the data I combined two research methods, namely direct content analysis and qualitative stakeholder interviewing. Both are among the most common qualitative research methods (Bryman, 2012).

3.3.1 Direct content analysis

Direct content analysis was used to analyse policy documents from the Smart Industry program to highlight the main concepts, and what is expected to be the main manifestation of industry 4.0 in the Netherlands. These documents can be accessed on the Smart Industry website. The literature review provides a foothold for this analysis. The documents that were analysed are:

- Dutch industry fit for the future (2014a). This is the original document in which the

importance of developing smart industry is highlighted.

- Actieagenda Smart Industry – Dutch Industry fit for the future (2014B). Building on the first

document, this document discusses the agenda to actually realise the smart industry developments.

- Actieagenda Standaardisatie – Dutch Industry fit for the future (2017). Building on the

experience of the first few years of the Smart Industry platform, this document discusses the importance of digital standardization within the Dutch industry to realise a more efficient and automated way of sharing information and data.

- Implementatieagenda 2018-2021 – Dutch Industry fit for the future (2018). The first actieagenda had a broader focus on raising awareness within the Dutch industry about the

importance of the developments in digitization and automation. This ‘implementatieagenda’ is meant to actually implement the new technologies.

The literature review provided a foothold for the analysis. The analysis was guided by the main concepts found in the literature. Through the use of Atlas.ti these concepts were highlighted within the Smart Industry documents. This allows for an overview to be created of which concepts are discussed the most within the Smart Industry platform, what objectives for the Netherlands are set, what obstacles may lie ahead, and how these obstacles are addressed to achieve the set goals. Through the use of coding, sections of text were highlighted and were then appointed to one or several categories that correspond with the main concepts. All the codes together construct a codebook which can be seen in Appendix A. This codebook was gradually built as the Smart Industry documents were analysed one by one to maintain an open mindset. This made sure that the analysis

(16)

16 was not too skewed towards the concepts found in the literature, which allowed me to identify unexpected concepts that might provide a deeper understanding of the fourth industrial revolution in the Netherlands. This analysis does not necessarily provide a deeper understanding of the phenomenon (Bryman, 2012), but it clarifies the context in which it occurs. As a result it serves as a great stepping stone for the next phase of data collection (Hseih & Shannon, 2005).

3.3.2. Qualitative expert interviews

The next step of data collection was in depth interviewing with representatives from the Smart Industry network. Interviewing is a great method to gather inside knowledge from participants that have a direct relation to the phenomenon (Cohen et al., 2011). The interviews were conducted with the aim to provide real life experiences and expectations to mostly theoretical assumptions. I selected several Smart Industry field labs that were working on production network developments, changes and solutions, and also 3D printing. In total I interviewed six representatives from six different field labs. Their roles differed from field lab and/or education coordinator, project

(17)

17 managers and researchers in innovations programs, and growth accelerator helping companies transitioning to a more digitised production process. Every field labs has its contact information on the Smart Industry website, so it was easy to reach out. The interviews were semi-structured to maintain a similar focus on the same topics with all my interviewees. The interview guide was based on the literature review and the direct content analysis so that the key concepts could be discussed. The main topics included defining the changes within production networks, mass customisation and 3D printing. The interview guide can be found in Appendix B. To make sure that the key concepts were interpreted the same by me and the interviewees a clear operationalisation of the key concepts is necessary (Boeije et al., 2009). An overview of the main concepts and their definitions are given in table 1 on the previous page. However, as the interviewees were from different field labs, their fields of expertise also differed. Therefore the focus of each interview skewed a bit towards a different topic than the next interview. Afterwards the interviews were transcribed in order for me to analyse them in detail with the use of the Atlas.ti codebook that was constructed in the direct content analysis. That way I could highlight overlapping themes between the literature review, the direct content analysis and the interviews.

3.3 Conceptual Model

In this chapter I present a conceptual model to visualise the causality of this thesis. In this model the outcomes of the fourth industrial revolution are not very detailed due to the fact that the

implications of the fourth industrial revolutions are very broad, and also because the outcomes are still hard to precisely predict. Furthermore there is also reciprocal connection between causes and predicted outcomes, as the outcomes tend to strengthen the necessity of continuing developments.

This means that with increasing (international) competition, companies have to innovate to maintain their customer base. To remain competitive companies have to appeal to the wishes of the

customers (mass customisation & servitisation). This in turn leads to innovation that can facilitate that need. The labour market shifts mean that the industrial labour pool is changing to more high skilled workers. This is necessary to innovate and implement new technologies and ways of doing business.

(18)

18

3.4 Validity, replicability and ethical considerations

The validity of a research project is one of the most important criteria to ensure the integrity and reliability of the results of the study (Boeije et al., 2009; Bryman, 2012). For qualitative research this is usually divided into internal, external and ecological validity (Bryman, 2012). Internal validity measures the causality of the results which means if the research questions are really answered through investigating the independent variables. External validity is determined by the extent to which the results of the study can be generalised. Ecological validity determines if the research can be applied to an everyday social setting. If a scientist intervenes in a social setting, for instance by conducting research in a lab, the ecological validity is decreasing. The key to establishing a good research validity is to remain focused and consistent from start to finish. An extensive literature review, a good operationalisation of the main concepts and well designed interview guides are important tools to maintain focus and consistency (Boeije et al., 2009). Replicability is always harder to reach in qualitative research as a social setting can’t be frozen in time (Bryman, 2012). It is therefore important to be as transparent as possible by writing down every step of the research process. Especially the direct content analysis is a good tool in this regard as a coding scheme provides a clear guideline to analyse all documents and the interviews. The analysis was conducted through the use of Atlas.ti software. The coding scheme was built along the way with a focus on the main concepts of this research, as this allowed me to highlight any overlaps between interviews and the policy documents. Additionally, it allowed me to identify the most discussed and most important themes (Miles et al., 2014). Every research has its limits and its pitfalls in collecting data and the results that follow. Bryman (2012: p. 135) names four main ethical considerations:

1) Harm to participants; 2) Lack of informed consent;

3) Invasion of privacy; 4) Deception of the participants of the study. Keeping that in mind, in my reach outs I clearly stated what this research is about, and what knowledge I hoped to gain from the interview. With the permission of the interviewees, I recorded all interviews. These recordings were discarded after transcribing and analysing the interviews. All interviews were processed anonymously to maintain the privacy of the participants. That is also the reason why not all the field labs from which I interviewed representatives are mentioned, as the respondents’ information is easily found on the smart industry website.

3.5 Scope and limits of the research

The focus of this research is on the development of the fourth industrial revolution within the Dutch manufacturing industry. This is with the aim of providing real life context to the still mostly

theoretical assumptions with regard to the fourth industrial revolution. After all, this is still a relatively new concept. In this research I will try to identify to what extent these theoretical assumption coincide with real life developments and expectations from experts in the field. The generalisability of this research is limited because it only highlights the situation in the Netherlands. Figure 1 in the introduction shows that most European countries have development programs in place, besides a lot of other industrialised countries outside of Europe. There are a lot of differences between the size and scope of all these national industries, which might result in different priorities regarding the fourth industrial revolution. Nevertheless, the more general theoretical expectations are discussed and the results might therefore provide a decent understanding of the developments and results of the new industrial revolution.

(19)

19

4. The Dutch Smart Industry platform

The aim of this chapter is to describe the setting in which the fourth industrial revolution in the Netherlands is developing. This is important to provide some real life context to the theoretical expectations and assumptions, and also in which context my analysis is carried out. The Netherlands has set the goal to be, and remain at the forefront of the fourth industrial revolution. The Smart Industry platform is responsible for achieving this goal. This research is focussed on documents and respondents of this platform, so it is important to highlight the platform and research what its main operations and policies are. The Smart Industry platform focuses on the Dutch manufacturing

industry which is characterised by mostly high quality and low volume production, usually referred to as the ‘maakindustrie’. A key feature of this maakindustrie is that products are tailormade and build on a close producer – consumer relationship (Smart Industry, 2014a). There is a general realisation that if the Netherlands doesn’t invest in keeping the manufacturing industry up to date it might significantly hurt the Dutch economy and employment in the (near) future. Additionally, seizing the opportunities the fourth industrial revolution has to offer is not just important for economic growth and employment, but also for current and future societal issues. A main societal issue is for instance climate change. The new technologies will result in a more efficient production process and will therefore reduce the need for raw materials. Additionally, new technologies will also facilitate a more circular economy as all production stages will be better connected (Smart Industry, 2014a; 2018). Furthermore, through the use of new technologies developed in the fourth industrial revolution, incentives and opportunities for companies will arise to facilitate reshoring. This is also one of the main goals as reshoring is expected to be significantly beneficial for economic and employment growth (Smart Industry, 2014a; 2014b).

These developments need to be embedded within the Dutch Top Sectors. The Top Sectors policy builds on nine sectors in which the Netherlands is globally competitive. These include sectors such as the high tech systems and materials sector, agri-food and the chemical sector. These sectors are on a world class level and are the cornerstone for the Dutch economy and international competitiveness (see also https://www.hollandtradeandinvest.com/key-sectors). It is therefore crucial that the topsectors develop alongside the fourth industrial revolution (Smart Industry, 2014a; 2014b). Developing new technologies is just one part of the vision for the future. In order for these

technologies to be relevant it is important that the Dutch industry is aware of the possibilities of the new industrial revolution (Smart Industry, 2014a). This was the main goal for the Smart Industry platform the past five years. The awareness is highlighted by all interviewees from the field labs. One interviewee (4) estimates that about three quarters of the companies within the maakindustrie have an awareness that they need to invest in these developments. However, there is a broad distinction to be made. There are some who have a keen interest in the new technologies, and push the

developments forwards. A lot of companies also realise something has to happen sooner or later, but will just ride the wave for now. And then there are others who have a hard time grasping what is going on, and will struggle to stay in business in the near future (Respondent 3). So it is important to note that there is still a big difference between companies with regard to the capability and

willingness to adapt. It is up to the Smart Industry platform to provide the tools and knowledge for all these companies to help them innovate and adapt.

With the creation of the Implementation agenda 2018 – 2021 (2018) the aim has shifted more to the development and implementation of new technologies now that there is a broad consensus within the Dutch industry to transition to the fourth industrial revolution. Crucial to raising awareness, developing technology and now implementing it, are the 39 Smart Industry field labs.

(20)

20 In these ‘labs’ solutions are being developed, tested and implemented (see figure 5). These field labs are generally companies who have to make their own money to fund most of their target

development. Through the platform they for instance can get access to government subsidies, but also tap into knowledge and experiences from other field labs. Besides private field labs, there also field labs that are run by a collaboration of universities for example. In general the labs are a collaboration of government authorities, knowledge institutions and industry partners. Some examples to highlight the diversity amongst the field labs are the Ultra personalised products and services (UPPS), Smart dairy farming 3.0 and Industrial robotics (for more information see

https://www.smartindustry.nl/fieldlabs/). In pursuing these developments the Smart Industry platform is structured on three pillars; manufacturing technologies, digitisation and network centric production system. This model is presented in figure 6. Digitisation is primarily focused on developing high-end ICT networks that connect production- and product information with each other. The manufacturing technologies concern the development of new production methods such as 3D printing and industrial robotics. The network centric production system covers the changing organisation of production networks or value chains to facilitate the integration and connection of new technologies, human capital and business models (Smart industry, 2014a; 2014b; 2018). So the new production system is tying it all together, and is therefore the main focus of my analysis. To conclude, the importance of transitioning to a more digitised and automated industry has been recognised by the Netherlands with the introduction of the Smart Industry platform. This platform attempts to coordinate this transition by linking education, government and industry to each other with the aim of developing and implementing new production technologies, and educate the current and future workforce to work with these technologies. For the past five years the Smart Industry platform has gradually build up its platform and the scope of its goals and activities. This is likely to continue and the platform has set a clear agenda for the next few years to maintain the world class level of the Dutch manufacturing industry.

Figure 5: A field lab focused on 3D printing solutions. The second picture shows a robot arm that is converted to a 3D

(21)

21

Figure 6: Building blocks of Smart Industry development

(22)

22

5. Results of the research

In this chapter I will provide an analysis of the changes in production networks, and its main expected effects. The changing production network is the overarching segment of the fourth industrial

revolution as it includes the new technologies, new types of labour, new roles of suppliers, producers and consumers and so on. It is therefore the main focus of this thesis and it will be thoroughly discussed in this chapter.

5.1 The network centric - production system

The developments in production networks are captured by one main concept, ‘the network-centric

production system’. According to Smart Industry this is quite a radical change of the production

systems as we know them today. This new production system is facilitated by a continuous digitisation and automation of production. Companies within a network will be increasingly

connected to each other and collaborate on a much larger scale than ever before. Through the new technologies companies will share more information on production processes to allow for a stronger connection between companies in a production network (Smart Industry, 2014a; 2014b). The citation below highlights the radical change of the organisation of production.

In the coming decade, a network centric approach to production will replace linear production processes with intelligent and flexible network approaches. These networks will interconnect parts, products and machines across production plants, companies and value chains at a highly granular level. Production in existing value chains will be radically optimised in the network centric approach, but, more importantly, the notion of network centric production finally spells the end of the ‘value chain’ and the birth of the ‘value network. (Smart Industry, 2014a, p.6)

This citation also highlights the expectation that production networks will specialise to and within niche markets. This is already an ongoing process as many entrepreneurs are already specialising to a niche within a niche (Smart Industry, 2014a). The current developments in digitisation and

automation technologies offer an opportunity for these entrepreneurs to realise a cheaper, yet more flexible production process that is able to adapt quickly to changing customer demands (Smart Industry, 2014b). And flexibility is a key concept here. It is not just a result of the new technologies, it is also a necessity within the network centric production system. Meaning that consumers have increasing possibilities to set demands, and the entire network must be able to respond to these demands. If one supplier fails to sufficiently react, than the entire production process is delayed. This also exemplifies the necessity for increased collaboration between companies as they become more dependent on each other. The importance of flexibility is highlighted by Smart Industry as its goal ‘is

to ensure that by 2021 the Netherlands has the most flexible and digitally connected production network in Europe…’ (Smart Industry, 2018; p.24). This flexibility is also important for sustainable

and environmental targets. Quick reaction time to consumer demands and an increasing connectivity between companies means that everything can be made to order. Producers only buy the (raw) materials they need, and have no necessity for a large warehouse full of supplies anymore.

We do everything order specific. The material that we have in stock is just for urgent orders. That is something that needs to be delivered in just a few days. Everything a bit longer… the standard delivery time is a week. For that we purchase the materials the need as soon as we receive the client’s order. (…) For us the power of numbers also matters.

(23)

23

So we need less materials to cut our products in comparison to our competitors. That is economically and environmentally sustainable. (Respondent 2)

Moreover, this also means that companies are much better able to switch from producing product A to product B. The necessity for this flexibility also becomes clear due to rapidly changing demands consumers might set. A producer might one week need 100 units of a product from their suppliers, the next week it might be 500. So their supplier, and so on down the chain, must also have the capability to react to changing demands (respondent 4). A great example is the field lab The Smart Bending factory which is presented in box 1.

Vital to the new production system that is exemplified in Box 1 is the collaboration of companies within a production network. Businesses traditionally organise in a way to react to economic trends and other companies. However, the aligning of a production process is key within a network-centric production system. So companies should collaborate with-, rather than react or compete with other companies (Smart Industry, 2014a). Companies can then for instance develop products based on products of other companies within the value chain. They can also increasingly specialise, and through collaboration and openness with each other it allows the different steps of the production process to be fully aligned (Smart Industry, 2017). This also suggests that companies should not specifically focus on optimising their own business, but focus on the optimisation of the value chain instead. This is underlined by Smart Industry’s original report (2014a):

Smart Industry links companies in networks, with an active input in communication and organisation of the work of other organisations. Individual entrepreneurs have to shift their focus more and more to the added value of the value chain or network as a whole. Already, cloud services are emerging to organise information throughout the value chain, giving companies direct access to logistics information from organisations – upstream and downstream. This approach is expected to be extended to information about manufacturing, enabling active intercompany influence on production. (p. 14)

Box 1 - The Smart Bending Factory

This field lab is lead by 247 Tailorsteel (https://www.247tailorsteel.com/en) who sell basic metal products such as plates and tubes. They are frontrunners within their field due to their innovative way of production. The innovative part at first was that the process of laser cutting the plates was fully automated and digitised. This field lab is now coming to an end, and what they achieved is that they managed to also fully automate the bending of metal plates and tubes. They only have a very minimal amount of supplies in stock, which are only used for an order that has to be delivered on a very short notice. Their standard delivery time is about a week, and that is all the time they need to process the order, buy the supplies they need, produce the product and finally deliver it to their customer. Moreover, when metal plates are cut to size and shape, there will be leftover metal. This metal will go into their stock, and will be used later on for an order that requires smaller units. This means that this factory requires significantly less materials and produces less waste, which of course is a major contribution to a more sustainable production process.

(24)

24 The increased collaboration within-, and specialisation of production networks means that it is likely that production networks will fractionalise (Smart Industry, 2017). It provides an opening for new companies to move into the value chain. The possibility for new companies to move into a value chain is not just a result of increasing flexibility, it is also facilitated by it. As customer demands can much easier be set per order, there is not such a preference for big customers anymore. In one of the interviews it was mentioned that smaller customers, who usually set smaller orders, are now just as well served as big clients. In the old situation smaller clients often had to wait before their order was produced, as bigger clients were given priority.

When it is crowded in the market and customers have to manually place orders, you notice that small customers often have to wait longer. But a lot of small customers make a big one, so that is also an opportunity for revenue (Respondent 2).

Therefore it might become easier now for starting companies to move into a value chain, and as well because they have to collaborate more within a value chain now and therefore will not suffer as much from competition with larger companies. So it is very much a possibility that within the network centric production system we get to see more actors as part of the production process. However, despite these expectations that the Smart Industry platform makes regarding the fractionalisation of the production network, contrasting points were made in a few interviews.

My expectation is that there will be fewer actors because production will happen with more general technologies. (…) You can attach several applications to a robot arm such as milling tools, soldering iron, screwdriver etc. The robot arm and its movements however remains the same. (Respondent 5)

Meaning that a company is able to expand its product portfolio, rather than specialising towards one or a few products. This would make it harder for new companies to move into a value chain as bigger companies can also produce for niche markets. Another interviewee (6) mentioned that a lot of companies within the Dutch maakindustrie are small family owned companies. A lot of these small companies are bought by big companies further up the value chain, especially when there are no successors in the family to continue the company. Additionally, it is much easier to achieve a flexible production network if the amount of actors within the network is as few as possible. So this is quite the contrary belief to what is expressed in the policy documents. Rather than the fractionalisation of a production network, there is also the belief that production networks are more likely to integrate, and that the number of links within the network is declining.

Whichever way it might lead to, one thing is becoming clear. New roles are emerging within the network. These roles are more service orientated and act as a middle man between producer and supplier (Respondent 6). This is exemplified in Box 2. Furthermore, there is a current project in a field lab where logistics companies partly take on the role as the middleman between supplier and

producer for mostly fairly standard components. In other words, the producers places their order at the logistics company, and they then find a matching supplier. So it is not just creating new roles within the network, but also as expected, companies are expanding their portfolio by offering new kinds of services (Respondent 6), which is discussed in the following chapter.

Impacts of the fourth industrial revolution on production networks of the Dutch manufacturing industry : a study into the Dutch Smart Industry program as it tries to facilitate the new industrial revolution and maintain