SMART CITIES: COLLABORATION IS KEY. An explorative study on the development of network resources in smart city projects in the Netherlands

(1)

-

SMART CITIES:

COLLABORATION IS KEY.

An explorative study on the development of network resources

in smart city projects in the Netherlands

Student: C.L.M. Goossens

Student No.: s4343743

E-mail: clm.goossens@student.ru.nl

Program: Business Administration Specialization: Organizational Design and

Development Supervisor: Prof. dr. K. Lauche 2nd_examiner: _{R. Smals}

(2)

Abstract

Since the last decade, the interest in smart city projects has grown increasingly, however many projects face challenges upscaling, many related to the collaborative nature of smart city projects. Based on technological innovation system theory, a possible solution could be the development of network- and system resources. Therefore, an exploratory study is done on the development of particular network resources expected to stimulate the upscaling in smart city projects. Based on innovation literature, such as transition management, strategic niche management and technological innovation system theory, enriched with literature on network management, the following resources are expected to stimulate upscaling: shared vision, trust, shared knowledge, external relationships and user engagement. A qualitative multiple-case study design is used for studying the subject. Twelve interviews were held within eight smart city projects in five different cities in the Netherlands as well as with three smart city experts. The findings showed that even though the projects were diverse in nature, the development of network resources was comparable and showed many similarities. A theoretical framework was built on the development of network resources in smart city projects. Future research can validate or enrich the theoretical framework.

Keywords: smart city projects, upscaling, collaboration, network resources, technological innovation system

(3)

Acknowledgements

First of all, I would like to thank my supervisor, Kristina Lauche for all the meetings and feedback that helped me immensely with writing the final thesis report, but even more so for giving back my confidence in my own work. In addition I would like to thank my second examiner, Raphaël Smals for his feedback and suggestions that helped structure my perspective on the topic. Furthermore I would like to thank all the people who participated in the research and a special thanks to the projects of this study: AiREAS, Cycle Data, City Deal: Kennis Maken Ede/Wageningen, City Deal: Kennis Maken Nijmegen, Jouwlichtop040, Mijnbuurtje.nl, Positive Drive App and Smart Emission, who were all very open and helpful in support of my thesis. Moreover, I would like to thank my family and friends for all the love and support during my final year. However, the biggest thanks of all goes to Nick Jelicic, my boyfriend, for always being there for me and never letting me down, even when it felt like an emotional roller coaster.

(4)

1. Introduction

1.1 Background

All over the world a growing number of people are living in urban areas. This trend, called urbanization, is predicted to continue for the next decades as it is forecasted that 68% of the world’s population will live in urban areas by 2050 (United Nations, 2018). The same trend is occurring in the Netherlands. Dutch cities are predicted to continue growing for the next decade (CBS, 2016). The growing population living in urban areas results in a complexity of physical and material problems, such as traffic congestions, air pollution and aging infrastructures, as well as social problems, such as poverty and crime (Alawadhi, et al., 2012; Angelidou, 2015; Caragliu, Del Bo, & Nijkamp, 2011; Chourabi et al., 2012; Neirotti, De Marco, Cagliano, Mangano, & Scorrano, 2014; Washburn et al., 2010). These problems of urbanization are wicked in nature. According to Head and Alford (2015), wicked problems are generally associated with the following three characteristics: 1) (conflicting) concerns of multiple stakeholders, 2) scientific uncertainty and 3) institutional complexity (i.e. inter-organizational collaboration and multilevel governance). Because wicked problems are too complex to be solved by a single organization, inter-organizational, multi-disciplinary partnerships have been established in order to experiment with innovative solutions for the problems caused by urbanization (Alawadhi et al., 2012; Chourabi et al., 2012; Huxham & Vangen, 1996; Mora, Bolici, & Deakin, 2017; Neirotti et al., 2014). With the rise of urban challenges there is an urgent need for smart solutions to overcome these challenges. Smart city projects attempt to tackle the wicked problems of urbanization and aim to increase the quality of life through innovative solutions organized within inter-organizational partnerships (Komninos, 2018). Since the rise of smart city solutions, smart city initiatives have received increasingly more attention from businesses and the market for smart city solutions is expected to continue growing over the following years (Allied market research, 2017; Deloitte, 2015). Not only businesses see the potential of smart city initiatives, but also governments are increasingly interested in smart city solutions and support them by providing subsidies (van Winden, Oskam, van den Buuse, Schrama, & van Dijck, 2016; van Winden & van den Buuse, 2017). Despite smart city solutions gaining increasing attraction from businesses and governments alike, many of smart city projects are unable to scale up (Heiligenberg, Heimeriks, Hekkert, & van Oort, 2017; Vilajosana et al., 2013; van Winden & van den Buuse, 2017). Upscaling is the process of an

(8)

2 innovation spreading to other users and regional contexts (Heiligenberg et al, 2017). Various studies have been conducted into the barriers of upscaling in smart city projects (Alawadhi et al., 2012; Chourabi et al., 2012; European Commission, 2017; Saunders & Baeck, 2015; van Winden et al. 2016; van Winden & van den Buuse, 2017). However, there is a lack in the smart city literature on how smart city projects can overcome the challenges related to upscaling. Although the barriers to upscaling are diverse in nature, many of these barriers seem to be related to a lack of collaborative capabilities necessary for working in inter-organizational partnerships (see paragraph 1.3). These collaborative capabilities are called network resources within the technological innovation system literature (see paragraph 2.5) and play a key role in the success of upscaling for these type of innovations. Therefore, the goal of this study is to increase the scientific knowledge on the development of network resources within smart city projects. Furthermore, this study aims to enhance the knowledge that can be used in upscaling in smart city projects.

1.2 Definition of smart city projects

Although the concept of smart cities has existed in scientific literature for over three decades, the concept has remained fuzzy as there are still many disagreements about the definition of a smart city (Alawadhi et al., 2012; Angelidou, 2015; Chourabi et al., 2012; Neirotti et al., 2014). The concept of smart cities was first introduced in the literature in mid-80’s and interest in the subject has grown increasingly since then (Komninos & Mora, 2018). Throughout the years, the concept of smart cities has been enriched by various perspectives and studies within different contexts (e.g. urban development, social sciences, public administration and information science) (Alawadhi et al., 2012; Angelidou, 2015; Chourabi et al., 2012; Komninos & Mora, 2018; Neirotti et al., 2014). Smart city solutions cover a broad domain of urban activities: smart governance, smart people, smart living, smart mobility, smart economy and smart environment (Komninos, Pallot, & Schaffers, 2012; European parliament, 2014). Due to the multi-disciplinary nature of smart cities, the literature in this domain is fragmented (Mora et al., 2017) and definitions are not always consistent (Chourabi et al., 2012). Therefore, smart cities as a phenomena are hard to conceptualize. Although the exact definition of smart cities remains debatable, I noticed that in many smart city definitions the following three recurring characteristics can be distinguished: 1) quality of life, 2) technology and 3) collaboration.

(9)

3 1) Quality of life: Ever since the introduction of the smart city concept, the aim for better environmental and social performance has been an important element of the definition of smart cities (Komninos & Mora, 2018). Most definitions of smart cities include the ambition to increase the quality of life for the citizens of the city (Caragliu et al., 2011; Komninos et al., 2012; Komninos, 2018). Quality of life consists of aspects such as sustainability, well-being, healthcare, and education.

2) Technology: The concept of smart cities was first introduced in the same era as other, similar concepts such as intelligent city, cyber city and digital city (Komninos & Mora, 2018). In the beginning of smart city literature, most authors tended to emphasize the technological facets of smart cities (Komninos & Mora, 2018; Mora et al., 2017) and focused on aspects such as the role of (information) technologies and urban smartness (Alawadhi et al., 2011; Washburn et al., 2010). This emphasis on technology driven innovation has led to smart city projects to focus on interesting technologies rather than responding to the real needs of the citizens, and as a result, many of these projects have failed (Saunders & Baeck, 2015).

3) Collaboration: The literature on smart cities has grown rapidly since the first decade of the century (Komninos & Mora, 2018) and since then the focus has shifted towards a more enriched definition of smart cities, highlighting the importance of inter-organizational collaboration and citizen participation (Alawadhi et al., 2012; Angelidou, 2015; Caragliu et al., 2011; Chourabi et al., 2012; Saunders & Baeck, 2015; Komninos & Mora, 2018; Mora et al., 2017). Neirotti et al. (2014), argue that in order for technology to transform cities, investments in human capital are necessary. Therefore, many authors now take a more socio-technical perspective when defining smart cities and tend to emphasize on the importance of investments in human capital and various forms of networking (Alawadhi et al., 2012; Caragliu et al., 2011; Carvalho, 2014; Chourabi et al., 2012; Komninos & Mora, 2018; Saunders & Baeck, 2015).

During the ‘Innotep 2018’ conference in Nijmegen, a practical definition of smart cities was presented which contained the same three recurring characteristics of the smart cities definitions from the scientific literature:

(10)

4 ‘A city in which the municipality, residents, businesses and knowledge institutions jointly use technology and data in combination with traditional infrastructure and facilities in order to come to better decisions and achieve a higher quality of life.’ – Mettina Veenstra, Innotep 2018.

This suggests that, even though there is not one universal definition of smart city, most experts agree that the definition includes three important characteristics: higher quality of life, technology and collaboration. Therefore in this thesis smart city projects are defined as follows: ‘technological (and social) innovation projects aimed at increasing the quality of life in the city led by multi-actor partnerships’.

This thesis focusses on smart city projects in the Netherlands for the following reasons: 1) smart city strategies differ per continent due to contextual factors such as history, governance structures, culture, etc. (Neirotti et al., 2014), 2) Europe is the largest contributor to smart city literature (Mora et al., 2017) and 3) smart city projects in the Netherlands are known for their emphasis on bottom up participation and community development (Zygiaris, 2012).

1.3 Problem statement and research question

Even though smart city projects are attracting increased attention, upscaling is still hard to achieve for many smart city projects (Heiligenberg et al, 2017; van Winden & van den Buuse, 2017; Vilajosana et al., 2013). According to the European Parliament (2014), two-third of European smart city projects stagnate after the pilot phase and are unable to scale up. Different studies have investigated which barriers are preventing European smart city projects from upscaling. The barriers to upscaling for smart city projects are broad and diverse in nature. In order to analyze the barriers I’ve grouped these barriers into the following categories: barriers related to the partnership (e.g. lack of shared knowledge), barriers related to users (e.g. lack of user engagement), managerial barriers (e.g. resistance to change), financial barriers (e.g. lack of business model), governmental barriers (e.g. lack of institutional readiness) and technical barriers (e.g. lack of technical standards) (Alawadhi et al., 2012; Chourabi et al., 2012; European Commission, 2017; Saunders & Baeck, 2015; van Winden et al. 2016; van Winden & van den Buuse, 2017). An overview of the barriers related to upscaling in Smart city projects in Europe and the Netherlands is given below in table 1.

(11)

5

Barriers in the upscaling of smart city projects in Europe and the

Netherlands

van Winden et

al. (2016)

van Winden & van den Buuse.

(2017) European Commission (2017) Alawadhi et al. (2012) Saunders & Baeck. (2015) Chourabi et al. (2012) Partnership

Lack of shared vision X X X

Lack of clear ownership and responsibilities X X

Partnership doesn’t fit the scope of the project X X

Partnership is not open to new partners (with

different competencies) X

Lack of sharing knowledge X X X X

User engagement

Lack in user engagement (early on in the

project) X X X X X

Management

Resistance to change from the workforce of the

partners X

Lack in management of ambidexterity (i.e. balancing exploration and exploitation activities)

X

Financial

Lack of a viable business model (early on in the

project) X X

Lack of systematic measurement of financial

risk X X X

Government

Lack of institutional readiness and existing

regulatory, legal and policy frameworks. X X

Lack of coordination between different levels of

governance (e.g. EU, city administration) X

Project too heavily protected by government

funding X

Technology

Too much of a technology push X X

Lack of inter-operational systems and technical

standards X X X

Lack of good existing technical infrastructure

(e.g. good wireless infrastructure) X X

Table 1: barriers in the upscaling of smart city projects in Europe and the Netherlands.

After analyzing ‘table 1: barriers in the upscaling of smart city projects in Europe and the Netherlands’, I noticed that many of these barriers emerge from the collaborative nature of smart city projects. Barriers that arise within the partnership are directly related to problems in inter-organizational collaboration, such as the lack of a shared vision and shared knowledge (Alawadhi et al., 2012; Chourabi et al., 2012; Saunders & Baeck, 2015; van Winden et al. 2016; van Winden & van den Buuse, 2017). User engagement is a critical barrier in the upscaling of smart city projects, which is directly related to the lack of collaboration between project members and intended users for the new innovation (Alawadhi et al., 2012; Chourabi et al., 2012; European Commission, 2017; Saunders & Baeck, 2015; van Winden et al. 2016). Some of the governmental barriers, such as lack of institutional readiness, may be overcome through better collaboration between project partners and governmental institutions (Chourabi et al., 2012; van Winden & van den Buuse, 2017). Even some of the technical barriers, such as a lack of technical standards, may

(12)

6 be solved through better collaboration between smart city projects working on a similar innovation (Alawadhi et al., 2012; Chourabi et al., 2012; van Winden & van den Buuse, 2017). To gain a deeper understanding of the challenges related to upscaling in smart city projects, I participated in three consultation meetings in Breda, Nijmegen and Utrecht on the subject of barriers related to upscaling in smart city projects. Many of the barriers could be traced back to their collaborative nature. Some examples were: difficulties with sharing knowledge, lack of user engagement, absence of a national smart city program and fear for collaborating from businesses and governments. Smart city projects in the Netherlands seem to experience barriers to upscaling related to their collaborative nature in both literature and practice. Therefore, this study focuses on how smart city projects develop particular network resources expected to stimulate upscaling.

The research question of this thesis is:

 Which network resources are expected to stimulate upscaling for smart city projects and how are they developed?

1.4 Overview

In order to answer the research question, relevant innovation theories such as transition management, strategic niche management and technological innovation systems are described in chapter two. These innovation theories are complemented with network management theories and combined with the knowledge on the barriers to upscaling for smart city projects. As a result, five network resources are expected to stimulate upscaling for smart city projects: shared vision, trust shared knowledge, external relationships and user engagement. Chapter three discusses the qualitative methodology of the study. In this chapter, it is explained why an exploratory, multi-case study design is chosen to study how these resources are developed within smart city projects. Moreover, a concise description of the cases is given. The results of each network resource are analyzed and discussed in chapter four. Chapter five gives an answer to the research question based on the results. It also contains the discussion on the limitations of the present study and gives suggestions for further research. This chapter ends with the description of the practical implications of the current study.

(13)

7

2. Theoretical background

Because smart city projects experience challenges related to upscaling, it is important to understand what type of innovations pertain to smart city innovations. Therefore, the difference between incremental and radical innovations is discussed (see paragraph 2.1). Next, in order to understand how smart city innovations could accelerate, transition management is discussed to understand the wider context of radical (sustainable) innovations (see paragraph 2.2), complemented by strategic niche management to create an understanding of upscaling on the level of smart city projects (see paragraph 2.3). These theories are enriched by the literature on technological innovation systems, in order to understand how smart city projects can build support structures for the new innovation through the development of network and system resources (see paragraph 2.4). Based on these theories complemented by network management literature and the knowledge on the barriers to upscaling for smart city projects, five network resources are expected to stimulate upscaling within smart city projects: shared vision, trust, shared knowledge, external relationships and user engagement (see paragraph 2.5).

2.1 Incremental vs. radical innovations

To better understand the process of upscaling for smart city projects, it is necessary to evaluate what type of innovation smart city projects are involved with. An innovation is defined as ‘an idea, practice or material artifact perceived to be new by the relevant unit of adoption.’ (Dewar & Dutton, 1986). In the innovation literature, a distinction is made between incremental and radical innovations: incremental innovations are minor improvements to current technologies, while radical innovations pertain to technologies that are fundamentally different from existing technologies (Dahlin & Behrens, 2005; Dewar & Dutton, 1986; Ettlie, Bridges & O’Keefe, 1984; Musiolik, Markard, & Hekkert, 2012). Because radical innovations differ fundamentally from current technologies, their success depends on their capability of managing both technical- and market uncertainty (Chesbrough, 2003). Organizations following the traditional innovation model complete the whole innovation process internally as one autonomous organization (Chesbrough, 2003; Lee, Olson, & Trimi, 2012). However, recent streams of innovation literature (e.g. open innovation and co-innovation) suggest that collaborating in inter-organizational partnerships with a variety of partners, such as businesses, knowledge institutes and citizens may be more beneficial than the traditional innovation model (Chesbrough, 2003; Chiaroni, Chiesa & Frattini, 2011;

(14)

8 Desouza et al., 2009; Lee et al., 2012). Through inter-organizational partnerships the number of possible sources for information and innovation is increased and as a result technological- and market uncertainty are decreased (Chesbrough, 2003; Chiaroni et al., 2011; Desouza et al., 2009; Lee et al., 2012). Smart city projects are characterized by their highly experimental nature and the development of innovations that differ from existing technologies (Komninos et al., 2012; European parliament, 2014; Li et al., 2016; van Winden & van den Buuse, 2017) and therefore experience many uncertainties (European Parliament, 2014; European Commission, 2017). For this reason, smart city projects are best understood as radical innovations projects. In this light, it is not surprising that many smart city projects are organized within partnerships (van Winden, 2016; van Winden & van den Buuse, 2017) as inter-organizational collaboration has been theorized to decrease technical- and market uncertainty. Due to the technical- and market uncertainty involved with radical innovations, smart city projects need a different approach to upscaling than incremental innovations. Therefore, transition management is discussed in the next paragraph, as this theory focusses on how radical innovations can accelerate.

2.2 Transition management

In order to understand how smart city projects can accelerate and scale up, it is necessary to understand the wider context in which these type of innovations take place. One significant theory in the field of innovation literature explaining the wider context of upscaling of radical innovations is transition management. Transition management studies the development of long-term, sustainable transitions (Raven, 2016) within multi-actor collaborations through a socio-technical perspective (Loorbach, 2010). For this reason, some authors argue that the upscaling of smart city projects resembles the development of socio-technical transitions (Carvalho, 2014) making transition management particularly suitable for studying the upscaling of smart city projects.

Transition management is built upon a multi-level perspective of socio-technical transitions (Loorbach, 2010). The multi-level perspective consists of three levels: the macro level (i.e. landscape), the meso level (i.e. regime) and the micro level (i.e. niches). The macro level contains the socio-technical landscape, which is the external environment in which transitions takes place. The landscape cannot be directly influenced by actors from the meso and micro level (e.g. digitalization) (Loorbach & van Raak, 2006). The meso level contains the socio-technical regime

(15)

9 and is best described as the dominant socio-technical system consisting of a coherent set of rules, (technological) structures, cultures and practices that are institutionalized and hard to break through for new innovations (e.g. waste management regulations) (Raven, den Bosch, & Weterings, 2010; Schot & Geels, 2008). Because smart city projects are located in urban areas, they do not only operate within a functional regime (e.g. mobility), but simultaneously function within the ‘meta-regime’ of the city (i.e. patchwork of sectoral regimes congregated in the city) which the city administration regulates (Raven, 2016). The last level is the micro level, which is where experimental projects, such as smart city projects, take place in niches. Niches are protected environments where experiments can develop without pressure from the dominant regime (e.g. mainstream market competition). Because radical innovations are not able to compete against established technologies on the market and need time to progress, they are developed in niches (Raven et al., 2010; Schot & Geels, 2008). This process of managing transitions is complex and it can last decades before transitions are accepted by the socio-technical regime (Kemp, Loorbach, & Rotmans, 2007; Raven et al., 2010). The core idea of the multi-level perspective is that the innovations in niches can build up internal momentum through building and coordinating networks and coalitions, so that collectively over time these networks will become strong enough to influence the socio-technical regime (Kemp et al., 2007). When changes in the landscape create enough pressure on the regime, destabilization of the regime creates windows of opportunities for new innovations to overthrow or be incorporated within the socio-technical regime (Raven et al., 2010). Therefore, mutual adoption and alignment between the three levels is crucial, because without it, it is hard for an innovation to accelerate and scale up (Heiligenberg et al., 2017; Schot & Geels, 2008).

Based on transition management, the upscaling of smart city projects does not merely depend on the success of individual projects, but more importantly, on the mutual alignment between level of the niches, the socio-technical regime and the socio-technical landscape. In order to understand how smart city projects can stimulate upscaling, it is necessary to take a closer look at the level of the niches. Therefore, in the following paragraph, strategic niche management is discussed. Strategic niche management studies how niches can gain momentum and be able to compete with the established technologies from the socio-technical regime.

(16)

10

2.3 Strategic niche management

Because smart city solutions are considered radical innovations, smart city projects need a different approach to upscaling compared to incremental innovations. From transition management it is understood that smart city projects take place in a wider context and that the success of smart city projects depends on mutual alignment with the socio-technical- regime and landscape. Therefore, it is necessary to understand how smart city projects gain momentum in order to compete with established technologies from the socio-technical regime. Similar to transition management, strategic niche management is built upon the multi-level perspective (Loorbach, 2010) and gives an in-depth insight in the micro level of niches (Loorbach & van Raak, 2006; Schot & Geels, 2008). Strategic niche management focuses on how niches bridge the valley of death between R&D and market introduction (Loorbach & van Raak, 2006; Schot & Geels, 2008). Because smart city projects are similar to the micro-level of protected niches in strategic niche management and experience difficulties scaling up (i.e. bridging the valley of death), strategic niche management is particularly appropriate for studying the upscaling of smart city projects.

As previously discussed, smart city innovations are developed in protected niches (Schot & Geels, 2008; Raven et al., 2010). According to strategic niche management, protected niches are used for ‘mutual articulation and alignment of technology, demand and broader societal issues’ (Schot & Geels, 2008). This is in line with the socio-technical perspective on which strategic niche management is built, where technological innovations and social innovations co-evolve with each other (Loorbach & van Raak, 2006; Schot & Geels, 2008) and as a result decrease technical- and market uncertainty. According to strategic niche management, niches could function as proto-markets which in turn could aid the development of market niches (Schot & Geels, 2008). Market niches are specialized markets with stabilized, shared rules and are slowly exposed to more selection procedures (e.g. market competition) (Schot & Geels, 2008). Once the innovation has become competitive, market niches grow and market competition can be used for upscaling (Loorbach & van Raak, 2006; Raven, 2016), which in turn makes it easier to get accepted by the socio-technical regime (Raven et al., 2010).

Strategic niche management distinguished two levels of niches: the local niche level and the global niche level. On the local niche level, distinct local niches, such as smart city projects, are experimenting with new innovations (Schot & Geels, 2008). At first, there are no shared rules and

(17)

11 expectations for the new innovation are broad. However, as time passes, a sequence of niches around the world operating on the same type of innovation can collectively establish shared cognitive rules and specific expectations, developing a global niche (Schot & Geels, 2008). Global niches are emerging communities that share cognitive, formal and normative rules (Schot & Geels, 2008). Gradually, the global niche becomes more specific with stabilized rules, making it easier to create a market niche and start phasing out protection (e.g. subsidies) (Schot & Geels, 2008). This shows that the relationship between experimentation and upscaling for smart city projects is more complex than a simple linear process within a single organization and that upscaling requires series of related experiments in various niches (Heiligenberg et al., 2017).

From strategic niche management, it can be learned that in order for smart city projects to scale up, it is necessary to achieve a global niche level with shared rules and expectations. In order to achieve this, it is necessary for smart city projects to collaborate with similar projects and additional parties within the same domain. The development of a global niche level is very similar to the development of a support system for technological innovation systems. Therefore, in the following paragraph, transition management is discussed, with a focus on the development of support structures for radical innovations, such as smart city innovations.

2.4 Technological innovation systems

Based on strategic niche management, it is argued that smart city projects would benefit from developing a global niche level with shared rules and expectations in order to gain momentum and in turn become competitive with the socio-technical regime. Technological innovation system theory focusses on building support structures for radical innovations (Farla, Markard, Raven & Coenen, 2012; Markard & Worch, 2009; Musiolik et al., 2012; Musiolik, Markard, Hekkert, Furrer, 2018). Similar to both transition management and strategic niche management, technological innovation systems theory is based on the socio-technical perspective (Farla et al, 2012; Markard & Worch, 2009; Musiolik et al., 2012; Musiolik et al., 2018) and studies the development of support structures for new innovations through inter-organizational collaboration (Farla et al, 2012; Musiolik et al., 2012; Musiolik et al., 2018), making it a particularly suitable theory for studying smart city projects.

(18)

12 A technological innovation system is defined as:

“.. a set of networks of actors and institutions that jointly interact in a specific technological field and contribute to the generation, diffusion and utilization of variants of a new technology and/or a new product” (Musiolik et al., 2012, p. 1033).

Technological innovation systems are based on the resource-based perspective, which is a theory in management studies explaining the competitive advantage of organizations (Markard & Worch, 2009; Musiolik et al., 2018). In the traditional resource-based perspective, resources are strategically (and organically) developed and accumulated within one autonomous organization to improve the efficiency and effectiveness of achieving its organizational goals, potentially resulting in competitive advantage if executed well. (Chesbrough, 2003; Markard & Worch, 2009; Lee et al., 2012). The network perspective, however, sees organizations in relationships with each other and developing and accessing particular resources through their networks (Glatter, 2003; Musiolik et al., 2012; Musiolik et al., 2018; Zaheer, Gözübüyük, & Milanov, 2010). The technological innovation system literature combines the resource-based perspective with the network approach and focusses on the progress of radical innovations through the development of supportive resources within multiple networks and actors interacting with one other (Musiolik et al., 2012; Musiolik et al., 2018). Resources are divided in two groups: tangible assets (e.g. equipment, finances) and intangible assets, which are further divided into structural- (e.g. knowledge, skills, culture, routines) and relational assets (e.g. customer loyalty, reputation) (Musiolik et al, 2012). Resources are the target of strategic activities (i.e. the deliberate creation of particular resources), however, these resources simultaneously constrain which strategic activities are within the realm of possibilities (Farla et al., 2012; Musiolik et al, 2018).

Because regime actors benefit from valuable system resources within the socio-technical regime, such as: ‘1) capabilities and technical knowledge, 2) identity and mission, 3) beliefs and cognitive frames and 4) regulations and other formal policies.’ (Farla et al., 2012, p. 995), overthrowing the socio-technical regime or becoming accepted by it is difficult for new innovations. Since new innovations do not have existing support structures to rely on, it is necessary to build new support structures to guide the actions of system actors developing the new innovation, also called system building (Farla et al, 2012; Musiolik et al., 2012; Musiolik et al, 2018). System building is done through formal structures (i.e. organizational structures with clearly identifiable members that

(19)

13 organize themselves to achieve a common objective) and/or through informal structures facilitating the exchange of knowledge and other resources between actors (Musiolik et al., 2012; Santoro, Borges, & Rezende, 2006). Through system building over time, system actors are able to reduce technical- and market uncertainty and build a more competing technological innovation system (Musiolik et al, 2018). Building support structures for technological innovation systems and achieving global niche levels is very similar, as both concepts entail the interaction between actors and networks within the same innovation (Musiolik et al., 2012; Schot & Geels, 2008) and through these interactions, shared rules and other supportive resources are developed increasing the potential for the new innovation to become competitive with the socio-technical regime (Musiolik et al., 2012; Schot & Geels, 2008).

How successful a network is at system building depends on the resources made available for- and/or developed within the network (Musiolik et al., 2012). In the literature, a distinction is made between three types of resources: 1) organizational resources, 2) network resources and 3) system resources. Organizational resources are produced and owned by one organization and are readily available to the network after joining. However, these resources are no longer available to members of the network after the organization exits the network (Farla et al, 2012; Musiolik et al., 2012; Musiolik et al, 2018). Network resources are developed and accumulated over time within the network through collaboration between network partners and thus are not readily available at the start of the network. Once network resources are established, all network members have access to these resources, even after an organization leaves the network (Farla et al, 2012; Musiolik et al., 2012; Musiolik et al, 2018). System resources are developed and accumulated over time through interplay between different system actors and networks and are not readily available to the technological innovation system. Once system resources are established, all system members have access to these resources (Farla et al, 2012; Musiolik et al., 2012; Musiolik et al, 2018). In the context of smart city projects, resources can be understood as follows: organizational resources are owned by partners of the project, while network resources are developed through collaboration between project partners. System resources, on the other hand, are established through interplay between the different smart city projects and external actors within the same technological innovation system. Through interplay between the three levels of resources, a supportive environment for the

(20)

14 new innovation is build, which in turn helps tackle particular challenges the new innovation faces (e.g. change in regulations through reputation) (Musiolik et al, 2012; Musiolik et al, 2018).

Technological innovation systems give insights in the need for developing a support system for the new innovation in order for smart city projects to scale up. In technical innovation systems theory, a support system is established through the development of system resources. A crucial step in building system resources is the development of network resources. As discussed in the problem statement, many problems related to upscaling in smart city projects are due to their collaborative nature. It seems like smart city projects have difficulties developing network resources. The lack of network resources could explain why many smart city projects do not scale up as network resources are necessary to build system resources. Therefore, in the following paragraph, based on previously discussed literature and network literature, five network resources are discussed. These five network resources are expected to stimulate the upscaling of smart city projects.

2.5 Expected network resources

The goal of the present thesis is to answer the research question ‘Which network resources are expected to stimulate upscaling for smart city projects and how are they developed?’. According to technological innovation systems theory upscaling is stimulated through the development of system resources. A necessary step in the development of system resources is the development of network resources. In the problem statement, an overview of the barriers to upscaling in smart city projects are presented in table 1. Although the barriers to upscaling in smart city projects are diverse in nature, many of these barriers seemed related to incapability to develop particular network resources for smart city projects (e.g. lack of shared vision, lack of shared knowledge, lack of user engagement). Based on the literature on transition management, strategic niche management, technological innovation systems and network management as well as the barriers from table 1 the following network resources are expected to stimulate upscaling in smart city projects: shared vision, trust, shared knowledge, external relationships and user engagement. In the following paragraph, these five network resources are discussed.

(21)

15

2.5.1 Shared vision

The first network resource expected to stimulate upscaling in smart city projects is a shared vision. One of the barriers to upscaling in smart city projects is the lack of a shared vision (Chourabi et al, 2012, van Winden et al, 2016, van Winden, & van den Buuse, 2017). In both transition management and strategic niche management, a shared vision is crucial to the successful development of radical innovations, as it provides guidance to the learning processes and trajectory of the new innovation, as well as serving as a cognitive model on which choices related to the innovation are made (e.g. technology design) (Heiligenberg et al., 2017; Kemp et al., 2007 Loorbach & van Raak, 2006; Raven et al., 2010; Schot & Geels, 2008). Furthermore, the process of visioning helps to attract attention, resources and new actors to the innovation, because when people believe in the shared vision, they are willing to invest and admit resources to the project (Raven et al., 2010). In the technological innovation systems literature a collective vision is considered to be a valuable resource as it functions as a coordinating mechanism for strategies of network- and system actors and aids the development of shared cognitive rules (i.e. shared knowledge) for the new innovation at network and system level (Farla et al, 2012; Musiolik et al., 2012; Musiolik et al., 2018; Schot & Geels, 2008). In network management literature a shared vision is essential to the success of inter-organizational collaboration (Glatter, 2003; Huxham & Vangen, 1996; Mandell & Steelman, 2010; Ospina & Saz-Carranza, 2010; Zaheer et al., 2010) as it helps understanding the goal of the collaboration (Mandell & Steelman, 2010; Ospina & Saz-Carranza, 2010). The following aspects are mentioned in the literature for the development of a shared vision:

- Facilitate interaction between partners: Through interaction between members of the network a platform for talking about common goals is created (Ospina & Saz-Carranza, 2010).

- Articulation of expectations: Through the articulation of expectations, actors are able to share and align expectations with each other in order to form a shared vision (Raven et al., 2010). Through discussing the expectations of the project, the scope of the project becomes more clear. If these expectations are implemented within the shared vision, it can give guidance and structure to the project (Farla, et al, 2012).

- Compromise on individual priorities: Trough compromise on individual and organizational goals a realistic shared vision for all partners can be developed (Huxham & Vangen, 1996).

(22)

16 - Adaptable vision: Due to the technical- and market uncertainty involved with radical innovations, a vision adaptable to new information is necessary to prevent socio-technical lock-in (Kemp et al., 2007; Loorbach & van Raak, 2006; Loorbach, 2010; Raven et al., 2010).

- Fit with long-term regional vision: Sustainability experiments (e.g. smart city projects) are theorized to benefit from alignment with the long-term regional vision, rather than alignment with national- or global visions (Heiligenberg et al., 2017).

2.5.2 Trust

The second network resource expected to stimulate the upscaling in smart city projects is trust. For the development of (other) network- and system resources in technological innovation systems trust is considered an important network resource as it fosters joint action and better collaboration between partners (Musiolik et al., 2012; Musiolik et al., 2018). Likewise, in the network management literature, trust is seen as crucial for effective collaboration and achieving shared goals (Glatter, 2003; Huxham & Vangen, 1996; Mandell & Steelman, 2010; Zaheer et al, 2010). At first, the process of building trust takes risk (Huxham & Vangen, 1996) and should be nurtured throughout the lifespan of the network (Glatter, 2003). The history of previous relationships has an influence on trust (e.g. good or bad experiences) (Glatter, 2003; Mandell & Steelman, 2010). Strong ties in the network increases tacit knowledge transfer between network members (Zaheer et al, 2010), which is important for the development of shared knowledge (see next paragraph 2.5.3). The following aspects are mentioned in the literature for the development of trust:

- Cultivate personal relationships: In order to increase trust within the network, cultivation of personal relationships is necessary to establish strong ties between the members of the network (Huxham & Vangen, 1996; Mandell & Steelman, 2010; Musiolik et al., 2012 Ospina & Saz-Carranza, 2010; Zaheer et al, 2010).

- Participation: Developing a trusting relatonship implies taking risk (Huxham & Vangen, 1996), as it is unknown if the other party can be relied upon to carry out the collective tasks (Mandell & Steelman, 2010). Therefore, when all partners participate in the project and contribute to the shared vision, trust in each other is likely to increase (Huxham & Vangen, 1996; Mandell & Steelman, 2010).

(23)

17

2.5.3 Shared knowledge

The third network resource expected to stimulate upscaling in smart city projects is shared knowledge. This is reflected in the smart city literature, as a lack of shared knowledge between partners is a direct barrier to upscaling in smart city projects (Alawadhi et al., 2012; Saunders & Baeck, 2015; van Winden et al., 2016; van Winden & van den Buuse, 2017). Learning is considered crucial for radical innovations for two reasons: 1) it reduces technical- and market uncertainty through aligning technological design with social preferences (Chesbrough, 2003; Desouza, 2009; Raven et al., 2010; Schot & Geels, 2008) and 2) it is necessary for the development of shared knowledge about the new innovation (Musiolik et al., 2012; Musiolik et al., 2018). Especially learning at the system level can help develop shared rules and cognitive frames (i.e. shared knowledge) (Musiolik et al., 2012; Raven et al., 2010; Schot & Geels, 2008). Learning from (and collaborating with) citizen groups and knowledge institutes are considered an important aspect of the success of sustainable innovation projects (Heiligenberg et al, 2017). The following aspects are mentioned in the literature for the development of shared knowledge:

- Trust: The stronger the ties are within the network, the higher the quality of the information and knowledge being shared between partners (Zaheer et al, 2010).

- Reflexive activities: Reflexive activities (e.g. evaluation and assessment of ongoing policies) are an important aspect of learning and help prevent socio-technical lock-in for the new innovation (Loorbach, 2010; Raven et al., 2010).

- First- and second order learning (in multiple dimensions): Successful innovation projects focus on first order learning (i.e. facts and data learning) as well as second order learning (i.e. changing cognitive frames) (Heiligenberg et al, 2017; Raven et al., 2010; Schot & Geels, 2008), since the nature of the problem and its solutions might be incorrect (Raven et al., 2010). Lack of second order learning is an often reported reason for the failure of niche experiments (Schot & Geels, 2008). Learning processes should take place in multiple dimensions such as technical aspects and user preferences (Schot, & Geels, 2008).

- Visioning: Second order learning is hard to achieve when the process of visioning is not done properly (Schot, & Geels, 2008).

(24)

18

2.5.4 External relationships

The fourth network resource expected to stimulate the upscaling of smart city projects is external relationships. One of the barriers in the upscaling of smart city projects is the lack of institutional readiness (Chourabi et al, 2012; van Winden & van den Buuse., 2017). Although developing external relationships might not directly aid institutional readiness, based on the knowledge from transition management it could be argued that the development of external relationships would be beneficial for two reasons: 1) in transition management, it is favorable for the success of the new innovation that the socio-technical regime is adequately open towards the particular innovation (Raven et al., 2010). Therefore, in transition management relationships with key players from the regime play an important role in the success of the new innovation (Heiligenberg et al., 2017;Raven et al., 2010). 2) The importance of involving external stakeholders is not only limited to the transition literature. According to strategic niche management many failed niche experiments can be traced back to ‘minimal involvement of regime actors which resulted in lack of resources and institutional embedding’ (Schot, & Geels, 2008). Therefore, external relationships with regime actors are considered as an important network resource. A second important external relationship is the relationship system actors. System actors are the other projects and additional parties within the same technological innovation system and are necessary in order to build a support system for the new innovation (Musiolik et al., 2012; Musiolik et al., 2018; Raven et al., 2010; Schot, & Geels, 2008). Reputation is considered important for developing external relationships:

- Reputation: Gaining reputation makes the network look more credible and legitimate in the eyes of external actors (Musiolik et al., 2012; Musiolik et al., 2018; Ospina & Saz-Carranza, 2010). Through reputation policy makers are more likely to adapt new regulations in support of the new innovation (Musiolik et al., 2012; Musiolik et al., 2018).

2.5.5 User engagement

The last network resource expected to stimulate the upscaling of smart city projects that is discussed in this study is user engagement. This is reflected in the smart city literature, as the lack in user engagement is one of the barriers to upscaling of smart city projects (Alawadhi et al.,2012; Chourabi et al., 2012; European Commission, 2017; Saunders, & Baeck, 2015; van Winden et al., 2016). In the study by Heiligenberg et al., (2017) user engagement was considered the biggest barrier in the upscaling of sustainable innovation projects. Users are defined as all organizations

(25)

19 and individuals that expect to benefit from using the new innovation, as opposed to selling the innovation (Hippel, 2005). In strategic niche management, user engagement is considered important because through user engagement early on in the project, it is easier to align user preferences with technical design (Desouza, 2009; Raven et al., 2010; Schot & Geels, 2008). In the innovation literature, there are no further insights on how user engagement is developed in innovative projects.

2.6 Literature overview

Radical innovations are part of a bigger socio-technical transition and need a different approach to upscaling than incremental innovations. Transition management shows that the upscaling of smart city projects is not merely dependent on the success of individual projects, but more importantly, on the mutual alignment between smart city projects, the socio-technical regime and landscape. Strategic niche management shows that in order for smart city projects to scale up, a global niche level with shared cognitive rules should be established in order to gain momentum and become competitive against the socio-technical regime. The development of a global niche level is very similar to the development of a support system for new innovations within technological innovation systems theory. This theory explains how support structures for new innovations can be built through the development of system resources. A crucial step in building system resources is the development of network resources. Many of the barriers in upscaling for smart city projects are related to the collaborative nature of smart city projects. It seems like many smart city projects experience difficulties developing particular network resources. Based on previously mentioned literature, complemented with network management and the knowledge on the barriers to upscaling of smart city projects, five network resources are expected to stimulate upscaling: shared vision, trust, shared knowledge and external partnerships. This thesis studies the development of network resources within smart city projects.

(26)

20

3. Methodology

In this chapter the methodology is discussed. First, the overall research strategy of this study is discussed, explaining why a qualitative, explorative, multiple-case study is the best design to answer the research question. After that, an explanation is given for choosing semi-structured interviews as a data collection method, followed by an explanation about the case selection. Next, a short description of the eight selected cases is presented. In addition, the data analysis method is explained followed with an explanation of how the quality of the research methods are upheld in this study. This chapter is concluded with an explanation about the research ethics of this study.

3.1 Overall research strategy

The aim of this study is to evaluate which network resources are expected to stimulate the upscaling of smart city projects. Furthermore, this study looks at how these network resources are developed within smart city projects. In order to give an answer to the research question, in-depth research of the social processes developing network resources in smart city projects is required. The context of this study is highly specific (i.e. smart city projects in the Netherlands) in light of the prior literature. The development of network resources have not yet been studied in the context of smart city projects. Therefore, a qualitative research strategy is chosen for this study. A qualitative research strategy is particularly suitable for studying social processes as qualitative research is more sensitive to the complexities involved with social processes (Babbie, 2013; Iacono, Brown, & Holtham, 2011; Schell, 1992). Qualitative research helps to examine ‘concepts in terms of their meaning and interpretation in specific contexts of inquiry’ (Ketokokivi & Choi, 2014, p. 233) and is thus highly suited for studying the development of network resources in the context of smart city projects in the Netherlands.

A multiple-case study is selected as the most appropriate research design for this study. A case study is described as:

‘an empirical inquiry that investigates a contemporary phenomenon in depth and within its real-life context, especially when the boundaries between phenomenon and context are not clearly evident’ (Yin, 2003, p. 13).

There are three criteria that determine if a case study is an appropriate design for studying the development of network resources in smart city projects: First, the case-study design is an

(27)

21 appropriate research strategy for studies with ‘how’ or ‘why’- research questions (Schell, 1992; Yin, 2003). This criteria is met since the research question of this study is ‘.. how are certain network resources developed in smart city projects?’. The second criteria for choosing a case-study design is related to the lack of controllability of the chosen subject (Schell, 1992; Yin, 2003). Since smart city projects are real-life cases over which a researcher has no control, this criteria is met. Lastly, the third criteria for determining the appropriateness of a case-study design is the degree of contemporary events (Schell, 1992; Yin, 2003). Both the domain of smart cities as well as the development of network resources are relatively new subjects to study and are presently studied. Therefore, the development of network resources in smart city projects is an highly contemporary event, which implies that the last criteria for choosing a case-study design is also met.

Case studies are commonly used for developing new theory (Iacono et al., 2011; Ketokokivi & Choi, 2014). As there are few studies on the development of network resources, and especially within smart city context, this study aims to generate new theories on this subject. Because the main focus of this study is to explore how network resources are developed in the context of smart city projects in the Netherlands, an exploratory research design is chosen. Such a research design is particularly appropriate when the subject of research is relatively new and a-priori theories are lacking (Babbie, 2013). A key feature of exploratory research is the use of inductive reasoning in order to develop theoretical generalizations (Eisenhardt, 1989; Ketokokivi & Choi, 2014). An inductive approach is suitable when the subject of study lacks a-priori theory and exploration of data is necessary in order to develop theories about the subject (Saunders, Lewis, & Tornhill, 2009). Since there is little literature on the development of network resources, especially in smart city projects, it is first necessary to analyze the data in order to develop theories.

3.2 Data collection method

There are six types of data sources which case studies use to collect evidence ‘documents, archival records, interviews, direct observation, participant-observation and physical artifacts’ (Yin, 2003, p. 83). Although triangulation of sources is recommended when conducting case study research, due to time constraints and lack of accessibility, semi-structured in-depth interviews were used as the only data collection method of this study (see Appendix 1: interview guide). Semi-structured interviews are a data collection method where the researcher prepares a list with open-ended semi-structured questions, which is very suitable for the studying complex social behaviors and

(28)

22 experiences as it allows for exploring unexpected issues (Longhurst, 2009). The logic of structured interviews fits well with the inductive nature of this study, as the interview is semi-structured and open-ended it leaves more space for exploring unknown variables (Boeije, 2014). As this way of interviewing tends to unfold in a conversational manner, if done well, semi-structured interviews can lead to more honest answers as this data collection method contributes to increased levels of trust between the researcher and the respondent (Longhurst, 2009). For this reason, the interviews was audio recorded, so the researcher was able to focus completely on the conversation with the respondent (Longhurst, 2009) as this helped feel the respondents more at ease.

3.3 Case selection

The multiple-case study design was chosen as the research design for this study, because patterns found in multiple cases are more robust than a pattern found in a single case. For this reason, a multiple-case study design is an appropriate method for developing theories in inductive research (Eisenhardt, 1989). Furthermore, the multiple-case study design is particularly fitting for studying complex topics involving many actors (Schell, 1992). Because the subject of the thesis (i.e. the development of particular network resources in the context of smart city projects in the Netherlands) has a high degree of complexity and involves many actors, the multi-case study design is deemed to produce more compelling and robust evidence (Yin, 2003). The evidence resulting from the multiple-case study design is more robust due to the underlying replication logic (Yin, 2003). Contrary to quantitative research, the primary purpose of qualitative research is not necessarily to generalize the results to a larger population, but rather to generalize to theoretical propositions (Yin, 2003) and to deepen the understanding of particular social phenomena (Babbie, 2013; Schell, 1992). For this study, purposive sampling, following the replication logic, was used to select cases, as purposive sampling is an applicable technique to use when particular types of cases are selected for in-depth investigation in a specific context (Almutairi, Gardner, & McCarthy., 2013; Mohd Ishak, & Abu Bakar, 2014). According to the replication logic, cases are selected based on either 1) the prediction of similar results (i.e. literal replication) or 2) the prediction of contrasting results but for predictable reasons (i.e. theoretical replication) (Hak & Dul, 2009; Yin, 2003). This is in line with the inductive reasoning of the study as an important step of replication logic is the development of a strong theoretical framework (Yin, 2003). The

(29)

multiple-23 case design was used for a comparison analysis between cases (Schell, 1992) in order to develop theories about the establishment of network resources in smart city projects based on recurring patterns between cases. If patterns are similar between cases, then the study has provided compelling evidence in support of the new theory. To increase the strength of the comparison analysis a broad range cases was selected for comparison based on a-priori criteria (see paragraph 3.3.1).

Because the researcher was unfamiliar with the field of smart city projects, a snowball sampling technique was used to collect cases that fit the theoretical criteria (see paragraph 3.3.1). Snowballing is a sampling technique where contacts in the field help direct the researcher to further possible contacts (Longhurst, 2009). The process of snowballing was started simultaneously with the literature research by attending three consultation workshops on the barriers for upscaling for smart city projects in Breda, Nijmegen and Utrecht, as well as visiting conferences about smart city projects. These consultation meetings and conferences were very helpful in giving direction to the research as well as meeting potential cases for the study. During the process of snowballing, three smart city experts were interviewed on the topic of the upscaling and asked were to give their expert opinion on the subject from a broader domain perspective in order to create a better understanding of the subject studied.

As this study focusses on the development of network resources within smart city projects, it was important to select the right respondents. The first and foremost criteria was that the respondents had enough experience and knowledge (Longhurst, 2009) on the project. Therefore, representatives from smart city projects were chosen that played a central role in the project. From each case, one respondent was interviewed, with the exception of the case ‘Jouwlichtop040’, where two respondents were interviewed. In total twelve interviews were held.

3.3.1 Criteria for case selection

Following the replication logic cases were selected based on the following criteria: technological innovation, network and quality of life.

1) Technological innovation criterion: The first important criterion for selecting cases was the technological innovation criterion. Based on the definition of smart city projects one of the three criteria for selecting smart city projects was the development of technological innovations. Smart city projects develop technological innovations within a broad range of

(30)

24 urban domains: smart economy, smart mobility, smart environment, smart people, smart living and smart governance (Caragliu et al., 2011). In order to increase the robustness of the results through theoretical replication, two cases within a similar domain were selected to see if similar patterns arise within and between different domains.

2) Network criterion: Another important element of the definition of smart city projects was inter-organizational partnerships. Therefore, the second criterion for selecting cases was the network criterion. In order to study the development of network resources within smart city projects it was necessary to select cases operating within a network of partners.

3) Quality of life criterion: The last criterion for selecting cases was the aim at a higher quality of life. One key component of the definition of smart city projects is the aim at a higher quality of life within the city. Therefore, cases were selected based that aimed at achieving a higher quality of life for the users.

3.4 Case description

In this study eight different smart city projects from five different cities in the Netherlands are selected as cases for the study.

3.4.1 AiREAS (Eindhoven)

‘AiREAS’ is a smart city project located in Eindhoven aimed at realizing a cleaner city. The network consists of more than 100 individuals from all kinds of organizations (e.g. private/public organizations, knowledge institutes, citizens). The first and most notable project of ‘AiREAS’ is the development of the ‘innovative air measurement system’ which has been introduced to the city of Eindhoven in 2013. The users of this project are both the municipality and the citizens of Eindhoven.

- Technological innovation criterion: ‘AiREAS’ is developing an air measurement system, as well as other projects including changing citizen behavior.

- Network criterion: ‘AiREAS’ operates in a network with multiple project owners including citizens, knowledge institutes, profit- and public organizations.

- Quality of life criterion: The goal of ‘AiREAS’ is to achieve a cleaner city for its citizens and operates in the smart environment domain.

(31)

25

3.4.2 Smart Emission (Nijmegen)

‘Smart Emission’, also called ‘the citizens sensor network’, is located in Nijmegen and aims to map the air quality of the city as well as some other elements such as sound. The project started in 2016 and has since completed their first project. Because the innovation is not yet finished, a follow-up project is planned in the future. The users of this project are the municipality of Nijmegen as well as the citizens.

- Technological innovation criterion: ‘Smart Emission’ is developing an air quality measurement system.

- Network criterion: ‘Smart Emission’ operates in a network with multiple project owners, such as the municipality of Nijmegen, Radboud University and profit organizations. Users play an important role within this network.

- Quality of life criterion: ‘Smart emission’ is active in the smart environment domain aimed at improving the air quality in Nijmegen.

3.4.3 City Deal: ‘kennis maken’ (Nijmegen & Wageningen)

‘City deals’ are a national initiative aimed at stimulating collaboration and learning between different projects within the same domain, such as mobility and waste management. Dutch cities are free to choose which city deal they want to participate in. The ‘City Deal: kennis maken’ is aimed at tackling a wide range of societal challenges through collaboration between knowledge institutes and the municipality. Two participating cities Nijmegen and Wageningen are selected as cases for this study. Although these two cases are not involved with developing technological innovations they are interesting to include in this study for two reasons: 1) these projects specifically focus on how to develop a well-functioning collaboration and 2) the city deal initiatives focus on collaboration between projects in a broader system, similar to the development of system resources from chapter two.

- Technological innovation criterion: Both cases are not developing a technological innovation, but a social innovation. The ‘City Deal: kennis maken’ is aimed at developing collaborative processes to solve societal challenges of the municipalities.