DETECTING DETERMINABLE CONSTRUCTS TO THE IMPLEMENTATION OF UNMANNED AIRCRAFT SYSTEMS BY USING THE CONSOLIDATED FRAMEWORK FOR IMPLEMENTATION RESEARCH WHEN THE SKY IS NOT THE LIMIT:

(1)

WHEN THE SKY IS NOT THE LIMIT:

DETECTING DETERMINABLE CONSTRUCTS TO THE

IMPLEMENTATION OF UNMANNED AIRCRAFT SYSTEMS BY USING

THE CONSOLIDATED FRAMEWORK FOR IMPLEMENTATION

RESEARCH

By Amber Mooibroek

University of Groningen

Combined thesis for MSc Business Administration (O&MC) and MSc Controlling (A&C)

June 2019

E-mail address:

a.e.mooibroek@student.rug.nl

Student number: […]

Thesis supervisor and assessor: Dr. B. Crom

Co-assessor: Prof. Dr. E.P. Jansen

(2)

When the sky is not the limit:

Detecting determinable constructs to the implementation of unmanned aircraft systems by using the Consolidated Framework for Implementation Research

Research summary: The implementation of disruptive interventions could enhance the quality and effectiveness of healthcare. Yet, healthcare organizations often experience difficulties when implementing these types of interventions. The purpose of this paper was to study in-depth which constructs could be determinable for the implementation of a disruptive technology in healthcare. We studied neglected and criticized determinable constructs in an understudied implementation phase and extended a prominent model in the field: The Consolidated Framework for Implementation Research. We conducted a case study in the Netherlands where multiple organizations were involved. Our findings suggest that the constructs of complexity, stakeholder engagement, cosmopolitanism and public attention are determinable for the successful implementation of disruptive technologies. Complexity was found to barrier implementation. Complexity was found to be caused by the degree of disruptiveness and the regulatory demands that come with employing the technology. Conversely, stakeholder engagement, cosmopolitanism and public attention were found to facilitate implementation. Our study suggests that the number of engaged stakeholders could lower the facilitating effect of stakeholder engagement. We present evidence that new and existing CFIR constructs interact in rich and complex ways. We recommend additional research to capture other determinable constructs to the implementation of disruptive interventions.

Managerial summary: Unmanned aircraft systems (UASs) could provide several potential benefits in modern healthcare systems. The systems could provide faster, extended urgent care at lower costs. However, the introduction and first operationalization (i.e. implementation) of these systems is subjected to several uncertainties. We provide insights about these uncertainties by illustrating specific matters that come with UAS implementation. We identified two matters that are problematic for the implementation of these systems in an early phase: the chicken-and-egg matter and the integration matter. The matters capture the disruptive possibilities and difficulties that come with introducing the systems in healthcare. Further, we found that strong networked stakeholder engagement and specific public attention could mitigate these matters and facilitate the introduction and operationalization of the systems. Our findings implicate what should be considered by managers when trying to implement UASs. This thesis can guide managers in identifying (positive and negative) factors that affect the achievement of organizational goals by implementing disruptive interventions.

Keywords: Early implementation, determinable constructs, unmanned aircraft systems, healthcare, CFIR, stakeholders and disruptiveness.

(3)

1. Introduction

When an emergency occurs, care must be provided fast since quick response care can save lives and increase long-term health outcomes (Van de Voorde, Gautama, Momont, Lonescu, De Paepe and Fraeyman, 2017). Common emergency healthcare care systems, such as in the Netherlands, are first aid posts of physicians and hospitals in combination with ambulances or trauma helicopters. These types of systems ensure a large network of emergency medical services, but at high operational costs (Van de voorde et al., 2017). With the rise of new technologies, a disruptive provider and means arrived in emergency healthcare (Hwang & Christensen, 2008; Howard, Murashov & Branche, 2018). Unmanned aerial vehicles and unmanned aircraft systems (abbreviated: UAVs and UASs) show to be beneficial for various scientific and public safety purposes (Clothier, Greer, Greer & Mehta, 2015; Balasingam,2017). UAVs, such as drones, can revolutionize healthcare since they offer an inexpensive solution to expanding emergency care (Bhatt, Pourmand & Sikka, 2018). UASs can play an important role in the supply of medicines and blood with speed and efficiency (Wen, Zhang & Wong, 2016). UASs can provide humanitarian aid after disasters have occurred or can be used to supply hospitals, remote clinics or hard-to-reach patients (Balasingam, 2017).

UASs provide many possible benefits over manned systems, such as having smaller platforms, being more (cost) efficient and enabling longer mission times (Dixon, Wickens, & Chang, 2005). Van de Voorde et al., (2017) refer to UASs for emergency missions as a ‘golden bullet’ and translate the benefits of UASs as a way to save lives and quality of adjusted life years. The application of UASs has been explored in the military (Dixon et al., 2005; Balasingam, 2017) and other branches for recreational, public and commercial use (Howard et al., 2018), though literature about the implementation of UASs in healthcare systems is scarce. This scarcity may be due to the relative newness of the technology in the health industry or the risks that are involved with such systems (Clarke, 2014). However, UASs could transform healthcare in the 21st_{century, therefore should attention be granted to UAS}

implementation (Balasingam, 2017).

While the healthcare industry is dynamic and organizations are continuously pressured to innovate (Ilott, Gerrish, Booth & Field, 2012), up till this day, it has not been achieved to implement UASs in modern healthcare systems on a large scale. Especially in Western areas of the world, such as the United States and Europe, remains the use of UASs for medical missions limited (Balasingam, 2017). UAS implementation in healthcare comes with hurdles such as regulatory limitations and monitoring issues (Balasingam, 2017). The existence of strict technological and practical requirements troubles the implementation as well (Dixon et al., 2005). The idea to implement UASs for medical missions is submitted to discussions between health providers, (local) governments, developers and other parties involved. Moreover, there are many other factors that could affect the successful implementation of new interventions such as UASs (Damschroder, Aron, Keith, Kirsh, Alexander & Lowery, 2009).

(6)

This paper aims to research these implementation factors, from now on termed as the constructs, that can determine the implementation success of UASs in the Netherlands. UASs are not implemented, however several Dutch organizations ambition to implement such systems. One of those organizations is Internship Organization1_{and provides us the opportunity to study the implementation of UASs in this}

early phase. The organization is a subsidiary of a large academic hospital in the Netherlands. Internship Organization is a healthcare provider that focuses on emergency care. To implement UASs in the Netherlands several organizations must engage. We will zoom in on this engagement, cosmopolitanism, public attention and the complexity of UASs. We intent to provide implementers with an overview that depicts the identified determinable constructs for the implementation of UASs.

The research question that centers this study is two-folded: which determinable constructs can

be detected in the pre-implementation of UASs? And how do these detected constructs determine implementation success? We conceptualize implementation as the planned process and systematic

introduction of value-proven interventions, and that these interventions are given a structural place in the practice of healthcare organizations (ZON, 1999; Grol, Wensing, Eccles & Davis 2013). In this study the intervention is a form of disruptive innovation. UASs are any means of transport through the sky without a human pilot on board (Clarke, 2014). Constructs are those factors within a determinable framework that can affect implementation outcomes (Nilsen, 2015). Constructs that are determinable are those that can facilitate or barrier implementation outcomes (Damschroder et al., 2009). For implementation outcomes, we focus on the ‘successfulness of the implementation’. We focus on the implementation of UASs for medical missions, as those UASs that transport emergency medications, blood products and other urgent care products. We use the Consolidated Framework for Implementation Research (CFIR) to detect potential determinable constructs (Damschroder et al., 2009). Further, we use the concepts of disruptive innovations and stakeholders in our study to extent current findings about the CFIR. To detect and understand determinable constructs we conduct a case study which is approached through an interpretive and iterative process.

Contribution

This paper contributes to existing literature in several ways. First, this paper goes beyond current literature that explores the application possibilities of UASs in healthcare. For instance, Balasingam (2017) discussed the possible pros and cons of UAS applications. Dixon et al., (2005) provided a more in-depth research about the use of UASs and the changing role of pilots and their workload. Howard et al., (2018) explored what UASs could mean for worker safety. These papers concerning the possible application of UASs in healthcare settings have in common that they briefly explain the extent of the related problems and possibilities of using UASs. It should be noted that UASs are not a simple intervention and it may not be easily implemented (Clarke 2014; Motlagh, Taleb & Arouk, 2016).

(7)

Several factors, internally or externally, could make it difficult to implement such systems (Damschroder et al., 2009). Howard et al., (2018) argue that it is too early to be certain about the benefits of UASs and that it is time to identify any risk such as determinable constructs associated with implementing these systems. Mapping constructs is important since the barriers and costs of prototype development and field testing are high (Van de Voorde et al., 2017).

We contribute to the field of implementation research by trying to extend a prominent model in the field: the CFIR (Damschroder et al., 2009). The framework has been praised for its extensiveness, applicability and structure (Breimaier, heckermann, Halfens & Lohrmann, 2015; Keith, Crosson, O’Malley, Cromp and Taylor, 2017). The CFIR includes important implementation and adaptation theories and deals with inconsistent conceptualizations throughout the field (Damschroder et al., 2009; Powell, Proctor & Glass, 2014). However, as with any framework, shortcomings are identified. For instance, the CFIR is not exhaustive, meaning that implementation constructs (i.e. determinants) are missing (Breimaier et al., 2015). One of those missing constructs which needs to be studied is the engagement of stakeholders. It is suggested that stakeholders can affect the implementation success of an intervention to a certain extent (Brett, Staniszewska, Mockford, Seers, Herron-Marx & Bayliss, 2010; Breimaier et al., 2015; Weir, Newham, Dunlop & Bennie, 2019). To grant a deeper understanding of stakeholder engagement, we include the existing CFIR construct of cosmopolitanism. This construct is about the degree of a networked relation among external organizations (i.e. stakeholders), and the strength of this network (Damschroder et al., 2009). Another overlooked CFIR construct is public attention. The public (i.e. society) is neglected in the current framework and forms a significant research gap since high public resistance can limit the implementation of interventions (Brett et al., 2010; Ilott et al., 2012). We contribute to the field by focusing on these two new constructs (i.e. stakeholder engagement and public attention) for the CFIR. The research setting of UASs implementation in the Netherlands provides an opportunity to study these two new proposed constructs. This is because implementation of UASs in the Netherlands asks for engagement between multiple stakeholders. Incorporating public attention is relevant, because it is expected that the society will resist the implementation of UASs (Clothier et al., 2015; Khan, Tausif & Malik, 2019).

Besides missing constructs are some CFIR constructs criticized. One of these criticized constructs for being too broad is complexity. The current definition of the construct appears to be comprised out of so many terms that it has become ambiguous (Ilott et al., 2012). In the CFIR, complexity is conceptualized as the perceived difficulty of implementation, reflected by the duration, scope, radicalness, disruptiveness, centrality, intricacy and number of steps required to implement (Greenhalgh, Robert, Macfarlane, Bate & Kyriakidou, 2004; Grol, Bosch, Hulscher, Eccles & Wensing, 2007; Damschroder et al., 2009). We will focus on one of these terms, namely disruptiveness. This is because UASs for medical missions are viewed to be disruptive (Srinivasan, 2013). Studies that applied the CFIR have in common that incremental interventions are studied. Examples of incremental

(8)

intervention studies are: a large-scale weight management program (Damschroder & Lowery, 2013), a new guideline for a hospital-based nursing practice (Breimaier, heckermann, Halfens & Lohrmann (2015), a telephone lifestyle coaching program (Damschroder, Reardon, Sperber, Robinson, Fickel & Oddone, 2017), a mobile-based tele monitoring program (Ware, Ross, Cafazzo, Laporte, Gordon & Seto, 2018) and a perioperative music initiative (Carter, Pyati, Kanach, Maxwell, Belden, Shea… & Raghunathan, 2018). As mentioned, healthcare organizations continuously innovate their processes (Ilott et al., 2012), however the field of implementation studies predominantly focuses on interventions of incremental nature. Health organizations can significantly change the business and achieve long-term goals when disruptive interventions are implemented (Christensen, Bohmer & Kenagy, 2000). Yet, health organizations often experience difficulties or even fail to implement disruptive interventions (Hwang & Christensen, 2008). We seek to examine the construct of complexity more in-depth with respect to disruptiveness and contribute to the conceptualization of the construct. To the best of our knowledge is our study one of the first to apply the CFIR to study the implementation of a disruptive intervention.

We contribute to the field of implementation research since we are one of the first to study the pre-implementation phase by using the CFIR. Damschroder et al., (2009) argue that the CFIR can be used throughout all implementation phases of pre-implementation, during and after implementation. However, most conducted research applied the CFIR during or after implementation. The pre-implementation phase seems heavily understudied and forms a significant research gap (Kelley, Yankey, Birken, Abadie and Damschroder, 2016). We will seize this research opportunity by studying the implementation of UASs in this early stage. This appears to strengthen the relevance to study the new proposed construct of stakeholder engagement, since Breimaier et al., (2015) and Weir et al., (2019) argue that it is very important to include stakeholders in the first phase of implementation.

This paper is practically relevant since our study provides insights about constructs that can barrier or facilitate the successful implementation of UASs. Identifying and understanding constructs that barrier or facilitate implementation is a way to assess the uncertainties, risks and opportunities that come with introducing and operationalizing these systems. This study can provide managers, who aim to implement UASs, insights about which positive and negative matters should grant attention when up taking these systems.

To summarize, this paper studies specific implementation constructs that could be determinable for the successful implementation of UASs. The remaining of this paper constitutes of four sections. In the following (second) section, the theoretical background will be presented. In the third section we elaborate on the research method. The fourth section presents the research findings. Fifth and finally, we include the discussion and conclusion of this study.

(9)

2. Theoretical background

2.1 Implementation and success

Implementation is the critical gateway between the (strategic) decision to adopt an intervention and the routine use of that intervention (Klein & Sorra, 1996). Implementation is a period in time where targeted stakeholders become increasingly skillful, consistent and committed in their use of an intervention (Klein & Sorra, 1996). Implementation research entails the scientific study of methods to promote the systematic uptake of clinical findings and other evidence-based practices into routine practices to improve the quality, effectiveness, reliability, safety, appropriateness and efficiency of healthcare (Eccles, Armstrong, Baker, Cleary, Davies, Davies, ... & Logan, 2009). The overall goal within the field is to better understand why implementation is successful in some organizations, while fails in others (Burnes, 2004; Eccles et al., 2009; Kirk, Kelley, Yankey, Birken, Abadie, & Damschroder, 2016).

To what extent interventions are implemented successfully is conceptualized in different ways. Damschroder et al., (2009) refer to implementation effectiveness, whereas others refer to implementation outcomes. Implementation outcomes include adherence to guidelines and research use, the application of research-based knowledge in practice, and the success of the implementation (Nilsen, 2015). Noteworthy is that term ‘success’ is used widely throughout implementation research but is almost never defined. Linder and Peters (1987) argue that the term ‘success’ has more definitions, which is partly dependent on research stream. According to the authors, success is defined to meet certain patterns of fit. This pattern of fit is subjective to the evaluation of the researcher rather than general pre-specified criteria. Thomas and Fernández (2008) argue that for a project the definition of success is multidimensional and extends beyond technical performance, cost, and duration towards the dimensions of user satisfaction and perceived benefits. Following the above-mentioned scholars, we view implementation success as the degree to which pre-defined expectations and goals are achieved, as defined by the implementer and or user of the intervention.

2.2 Implementation theories, frameworks and models

Within implementation research a variety of theories, models and frameworks exist to assess and guide implementation (Damschroder et al., 2009). Implementation theories, models and frameworks have three overarching aims. First, to describe andguide the process of translating research into practice. Second, to understand and explain what factors influence implementation outcomes. Third, to evaluate implementation (Nilsen, 2015). We pursue the second aim since we focus on detecting and understanding determinable constructs to the implementation of UASs. The theoretical approaches that could satisfy our research aim are classic theories, implementation theories and determinant frameworks

(10)

(Nilsen, 2015). For the purpose of this study, we will use a determinant framework (i.e. CFIR) which will be elaborated on extensively in the next section.

Determinant frameworks describe constructs that are hypothesized or have been found to influence implementation. These determinant frameworks are integrative in nature, meaning that implementation is seen as a multi-dimensional phenomenon with interacting components (Nilsen, 2015). Some frameworks have been empirically tested, such as the NASS-framework (Greenhalgh, Wherton, Papoutsi, Lynch, Hughes..., A'Court, & Shaw, 2017), where others have not such as the Implementation of Change Model (Grol et al., 2013). Grol et al., (2013) discuss that the model is highly conceptual and that in ‘real situations’ additional steps may need to be added. A problem with these determinant frameworks, also within implementation theories in general are the overlapping domains and findings but inconsistent conceptualizations (Damschroder et al., 2009; Nilsen, 2015). To tackle this issue, Damschroder et al., (2009) conducted an extensive meta-analysis of implementation frameworks, models and theories and formed the CFIR.

2.3 The Consolidated Framework for Implementation Research

The CFIR is based on 19 theories, models and frameworks, and facilitates research findings into (healthcare) practice. The framework can be used to guide implementation in multiple settings and is far-reaching; it can serve as a predictable means to ensure successful implementation (Damschroder et al., 2009). The framework consists of five domains and these domains interact in rich and complex ways. The five domains are: characteristics of the intervention (1), outer setting (2), inner setting (3), individuals involved (4), and implementation process (5). The five domains are built out of several (sub) constructs (Damschroder et al., 2009).

The first domain, characteristics of the intervention, includes the different properties of an intervention, in our research characterized as the properties of UASs. These properties are translated into the constructs of intervention source, evidence strength and quality, relative advantage, adaptability, trialability, complexity, design quality and packaging, and cost (Damschroder et al., 2009). The second domain, outer setting is about the economic, political, and social context where an organization is present. The inner setting (third domain) is about the structure, political, and cultural context through which the implementation process proceeds (Pettigrew, Woodman & Cameron, 2001). For the implementation of UASs the outer setting deals with the country of the Netherlands, but we also include the European Union. This is because UAS regulation and legislation is formed on European level (Balasingam, 2017). The inner setting can be different for each of the (Dutch) organizations that are involved in the implementation of UASs. Changes in the outer setting can affect implementation and are often mediated through effects of the inner setting (Damschroder et al., 2009). Constructs of the outer setting are patient needs and resources, cosmopolitanism, peer pressure, and external policies and incentives. The constructs that are part of the inner setting are structural characteristics, network and

(11)

communication, culture, and implementation climate (Damschroder et al., 2009). The fourth domain, individual characteristics tackles the role and agency that several individuals in an organization have when new interventions are implemented. For this study, we look at the characteristics of the representatives of the organizations that are involved in the implementation of UASs. Constructs in the domain are knowledge and beliefs about intervention, self-efficacy, individual stage of change, individual identification with the organization, and other personal attributes (Damschroder et al., 2009). The final domain of process includes the activities of the implementation process and translates the process activities into the constructs of planning, engaging, executing, reflecting and evaluating (Damschroder et al., 2009). Below, we have depicted an overview of the CFIR domains and constructs in figure 2.1.

Figure 2.1 The CFIR based on Damschroder et al., (2009)

The CFIR holds some strengths. First, the authors have tackled the problem of inconsistent conceptualizations in implementation research (Damschroder et al., 2009). Second, the CFIR is formed after an extensive analysis; prominent models and theories of the field are included (Powell et al., 2014). Third, the CFIR is an applicable and useful tool to guide implementation and reveals the constructs that could determine implementation success (Breimaier et al., 2015; Keith et al., 2017). These strengths provide us with a clear and evident base to detect the potential determinable constructs to UAS

Intervention characteristics Intervention source Evidence strength & quality Relative advantage Adaptability Trialability Complexity Design quality & packaging Cost

Outer setting Individual’ Process characteristics

Inner setting

Patient needs & resources Cosmopolitanis m Peer pressure External policies and incentives Planning Engaging Executing Reflecting & evaluating Structural characteristics Network & communication Culture Implementation climate Knowledge & beliefs about intervention Self-efficacy Individual stage of change Individual identification with organization Other personal attributes

(12)

implementation. Powell et al., (2014) argue that the CFIR provides one of the most comprehensive overviews of key theories, therefore is a guiding force in implementation research and practice.

In defiance of positive reactions, the framework has shortcomings. For instance, the CFIR had not been empirically tested when published (Damschroder et al., 2009). Present-day, we see that the framework has been applied or reviewed in studies (Damschroder & Lowery, 2013; Damschroder et al., 2017; Ware et al., 2018; Carter et al., 2018). Previous mentioned scholars encourage additional empirical research where the CFIR is applied. This is because implementation success can differ per setting or implementation phase (Damschroder et al., 2009). Besides, the framework does not seem to be exhaustive; implementation constructs are missing or underdeveloped (Brett et al., 2010; Ilott et al., 2012; Breimaier et al., 2015; Weir et al., 2019). With this mind, we take upon new and criticized constructs in an understudied implementation phase, and like the CFIR itself, represent them in a conceptual model at the end of this section.

To put it briefly, we see that the CFIR is well-received. The framework is a suitable basis for our study because it provides consistent, standardized conceptualizations and includes prominent implementation theories. However, we also see that the framework is flawed and incomplete. The CFIR entails twenty-five constructs (thirty-nine, including sub constructs). Studying all the constructs goes beyond the scope of this thesis. We will focus on those constructs that are understudied or criticized such as stakeholder engagement, public attention and complexity. This is relevant because theoretical and empirical notions indicate that these constructs could be determinable for successful UAS implementation. Research has suggested that the existing domains and constructs of the CFIR interact in rich and complex ways (Damschroder et al., 2009), therefore we will also study existing constructs such as cosmopolitanism. We expect that missing constructs can be defined and understood when we link them with existing constructs.

2.4 Potential determinable CFIR constructs for UAS implementation

Determinant constructs are those constructs that barrier or facilitate the successful implementation of new interventions such as UASs (Damschroder et al., 2009). The choice to focus on particular constructs is in line with existing literature (Breimaier et al., 2015; Kirk et al., 2016). Damschroder and Hagedorn (2011) argue that the choice to study specific CFIR constructs should depend on the highest applicability for the study. This applicability depends on the specific implementation setting and phase. It is important to justify and explain the rationale for selecting constructs (Damschroder et al., 2009). In the remaining of this section we will explain on which notions we base our decision to study the constructs of complexity, stakeholder engagement, cosmopolitanism and public attention. For the completeness of this section, we included an overview of all CFIR construct definitions in appendix I.

(13)

2.4.1 The determinable construct complexity. Interventions are often complex and multi-faced with many interacting components (Damschroder et al., 2009). A very complex intervention can be more difficult to implement, therefore complexity barriers implementation (Ware et al., 2018). The complexity of an intervention is conceptualized in the CFIR as the perceived difficulty of implementation, reflected by the duration, scope, radicalness, disruptiveness, centrality, intricacy and number of steps required to implement (Greenhalgh et al., 2004; Grol, et al., 2007). Ilott et al., (2012) argue that the construct of complexity is too broad and must be simplified. This simplification is needed because the broad term has shown to lead to insufficient information gathering. Some concepts are duplicated. For instance, we do not understand the logic to include both terms of radicalness and disruptiveness. Existing literature indicates that the concepts are more or less the same and closely intertwined (Hacklin, Raurich & Marxt, 2004; Markides, 2006).

We aim to capture and understand complexity by focusing on the concept of disruptiveness (thus radicalness). This is relevant because UASs for medical missions are a form of disruptive interventions. The vehicles are expected to disrupt business models of healthcare organizations (Hwang & Christensen, 2008). Disruptive innovations can be distinguished from incremental innovations. Incremental innovations are described as the improvement of technology performance or product feature enhancement, whereas disruptive innovations refer to new technologies and products (Hacklin et al., 2014). Disruptive innovations require three characteristics; they should be technology enabled, they should innovate the business model and should constitute so-called value networks (Srinivasan, 2013). Radical innovations disrupt consumers as they introduce new products and value propositions that disturb prevailing consumer habits and behaviors in major ways (Markides, 2006). UASs for medical missions are a form of disruptive innovation since the systems are technology enabled, can reform or introduce new business, and are value networked. UASs are a new way to provide care towards the citizens in the Netherlands. Hwang and Christensen (2008) discuss how healthcare organizations often fail to implement disruptive interventions, while it is perceived as extremely important to include these innovations to perform successfully on the long-term (Christensen et al., 2000).

We include the term duration as a reflection and conformation of disruptiveness, thus complexity. This is because there is a link between the disruptiveness of an intervention and the duration of implementation. Greenhalgh et al., (2004) argue that disruptive interventions ask for significant reorientation and new non-routine processes to change the organization its activities and to departure from existing activities. This significant reorientation costs time. It is expected that the implementation duration for UASs will be long because of strict regulatory demands (Clarke et al., 2014; Balasingam, 2017). Besides, we expect that the duration is long, because extensive system testing is needed (Van de Voorde et al., 2017).

(14)

We do not include complexity terms of steps, scope, centrality and intricacy because there are no theoretical underpinnings that these concepts underlie the determinable construct complexity for the implementation of UASs at this moment. It is too early to speak about the number of steps, scope or centrality in this early phase (Howard et al., 2018). We do not include the term intricacy because this concept refers to clinical procedures (Damschroder et al., 2009). An UAS for medical missions is not a clinical procedure.

Altogether, the CFIR suggests that the existing construct of complexity can be determinable for UAS implementation (Damschroder et al., 2009). A high degree of complexity barriers implementation (Ware et al., 2018). We expect that the barrier of complexity is caused by the degree of disruptiveness. We expect that disruptiveness; thus complexity is visible in the long (expected) duration of UAS implementation (Greenhalgh et al., 2004). To better understand how disruptiveness leads to complexity and is determinable for UAS implementation, we formed the following sub question:

Sub question 1: How does disruptiveness lead to complexity and therefore barriers successful UAS

implementation?

2.4.2 The new determinable construct stakeholder engagement. The term stakeholder is often used in implementation research. For instance, stakeholders are the ones who should become increasingly skilled with a certain intervention (Klein & Sorra, 1996). Damschroder et al., (2009) often refer to ‘stakeholders their perceptions’. A stakeholder is anyone or any group who can affect or is affected by the achievement of the organization's objectives (Freeman, 1984). Stakeholder examples are employees, customers, suppliers, financiers and the society. Parmar, Freeman, Harrison, Wicks, Purnell and De Colle (2010) argue that successful organizations create value for all key stakeholders. However, this value should not only be created for all the stakeholders, it can be created with stakeholders (Freeman, Harrison & Zyglidopolous, 2018). This is relevant for our study since the implementation of UASs only seems realizable with the combined efforts of stakeholders. We expect that a implementing organization can create value by implementing UASs for their stakeholders with (some of) their stakeholders. Following Freeman (1984), for an UAS implementing organization a stakeholder is any other organization or individual who can affect or is affected by the implementation of UASs. We merely focus on those (key) stakeholders that are discussed by the implementing organizations as an important stakeholder. Identified key stakeholders for implementing organizations are regulators and legislators, developers, operators, (other) healthcare organizations, employees, patients and the society (Clarke, 2014; Clothier et al., 2015; Balasingam, 2017).

There is discontent about how the CFIR includes stakeholders into the model. Constructs that give ‘real’ attention to how stakeholders barrier or facilitate successful implementation are absent (Breimaier et al., 2015). We follow Weir et al., (2019) who suggest that engagement of stakeholders can be determinable for successful implementation. Thus, we aim to study the new proposed construct of

(15)

stakeholder engagement. Greenwood (2007) argues that stakeholder engagement are the practices that a company (i.e. organization) undertakes to involve its stakeholders in a positive manner in organizational activities. For this study we view stakeholder engagement as the practices that a implementing organization undertakes to involve its (key) stakeholders to implement UASs. As suggested by Breimaier et al., (2015) we view considering and openly discussing the aims, needs and wishes of stakeholders as stakeholder engagement activities. Openly discussing aims, wishes and needs of stakeholders can lead to stakeholder involvement and acceptance. An increased involvement and acceptance could facilitate successful implementation (Breimaier et al., 2015). Besides, involving stakeholders could contribute to detecting additional barriers and facilitators to implementation (i.e. other determinable constructs) (Breimaier et al., 2015). Following the same authors, we view stakeholders’ aims as the idea that someone or an organization has with respect to what they want to achieve with UAS implementation. We see stakeholders’ wishes and needs as what is (continuously) considered during the implementation of UASs.

One might argue that attention is given to stakeholders and their involvement in the construct of engaging (process domain) in the current CFIR (Breimaier et al., 2015). The engaging construct is about attracting and involving appropriate individuals in the implementation (Damschroder et al., 2009). However, this construct only grants attention to those stakeholders that are involved with actual implementation. Implementers must also engage with other stakeholders, for instance those stakeholders that form a large or strong group and can resist to accept the implementation of interventions (Breimaier et al., 2015), such as regulators or the public (Ilott et al., 2012). We include a specific construct for engaging with society, namely public attention. Weir et al., (2019) advocated the importance of stakeholder engagement in the pre-implementation phase. Accordingly, engaging in an early phase can be positive for implementation (Breimaier et al., 2015). Weir et al., (2019) suggested future research about engagement in an early phase. We address this suggestion with this study and formed the following sub question:

Sub question 2: How does stakeholder engagement facilitate successful UAS implementation?

2.4.3 Stakeholder engagement with respect to cosmopolitanism. While the stakeholder engagement construct misses in the current CFIR (Breimaier et al., 2015; Weir et al., 2019). The framework does give attention to cosmopolitanism which is about the degree to which an organization is networked with other external organizations (Damschroder et al., 2009). The Dutch healthcare system, including governmental agencies has a dense network with healthcare providers (hospitals, GPs, Ambulance care) and healthcare purchasers (insurance companies) (Mossialos, Wenzl, Osborn, Sarnak, 2016). We distinct cosmopolitanism from stakeholder engagement since stakeholder engagement is about the activities an implementer organization undertakes to involve its stakeholders (Greenwood, 2007). This means that an implementer organization could engage with single stakeholders. Whereas cosmopolitanism is about a network between multiple external organizations, thus a network between

(16)

multiple stakeholders. We view stakeholder engagement and cosmopolitanism as separate constructs; however, we expect that there is a logical interaction between the constructs as indicated by Damschroder et al., 2009 and Nilsen (2015). Cosmopolitanism is expected to be an extensive form of stakeholder engagement.

According to Greenhalgh et al., (2004) a situation where organizations support external relations in a network could facilitate implementation. This suggests that stakeholders who engage in a network could facilitate UAS implementation. To understand the determinable character of cosmopolitanism one must consider ‘social capital’. Social capital is about the extend and quality of external relationships (Brehem & Rahn, 1997). Components of social capital are the degree of a shared vision among the network, and the degree of information sharing in the network (Damschroder et al., 2009). We expect that stakeholders who engage in a high-quality network facilitates implementation. To gain an understanding of what cosmopolitanism means for the successful implementation of UASs and what a strong relation constitutes, we formed the sub question of:

Sub question 3: How does cosmopolitanism facilitate successful UAS implementation?

2.4.4 The new determinable construct public attention. The society is an important stakeholder to consider in the implementation of UASs. Brett et al., (2010) and Ilott et al., (2012) argue that the engagement of citizens, therefore society misses in the CFIR. However, the CFIR does include a construct about the needs (and resources) of patients. This construct is conceptualized as understanding and prioritizing the need of patients, as well as knowing the barriers and facilitators to meet those needs (Damschroder et al., 2009). Building upon Damschroder et al., (2009) we consider a new construct, where we focus on the needs of society and meeting those needs. We call this construct public attention. Studying public attention is relevant because it is expected that the society will resist the implementation of UASs in the Netherlands. PytlikZillig, Duncan, Elbaum and Detweiler (2018) discuss that the attitudes of citizens can have large effects on the trajectory of UAS technology. Ramadan, Fara and Mrad (2017) argue that consumers can have hesitations about unmanned aerial delivery systems because of perceived privacy risks and safety risks. Clothier et al., (2015) researched the public acceptance of drone technologies in Australia and found that overall the public was not resistant, nor in favor of these type of technologies. The authors did find that the perceived privacy of citizens can be a significant issue when implementing such systems as argued by Ramadan et al., (2017). This is also confirmed by Khan et al., (2019) who found that in urban areas in Pakistan concerns about privacy are critical when assessing UAS delivery.

Public perceptions can facilitate or hinder the implementation of UASs (PytlikZillig et al., 2018). We expect that these perceptions can be understood when giving attention to the public. Additionally, we anticipate that giving attention to the public can facilitate UAS implementation. Implementing organizations could increase positive public perceptions and tackle negative perceptions

(17)

by engaging with the public (Greenwood, 2007). To better understand what public attention entails for UAS implementation, we formed the final sub question of:

Sub question 4: How does public attention facilitate successful UAS implementation?

2.5. Revisiting the research question

This section consisted of an investigation of current implementation models and focused on the CFIR. Revisiting the central research question: which determinable constructs can be detected in the

pre-implementation of UASs? And how do these detected constructs determine implementation success?

We view the CFIR as a suitable model to detect and explore which constructs could be determinable for successful UAS implementation. We expect that the degree of complexity can be determinable for the implementation success because of the disruptiveness of UASs. It is expected that the degree of disruptiveness, hence complexity is reflected by the expected duration of the implementation process.

We also expect that the engagement of stakeholders will be determinable for the successful implementation of UASs. Engaging with stakeholders could facilitate implementation. Stakeholders who have strong networked relations (i.e. a high degree of cosmopolitanism) are likely to implement UASs more successfully as well. One specific stakeholder that needs to gain attention in the implementation of UASs is the public. It is expected that the needs of the public must be understood and prioritized, thus grant attention, to successfully implement UASs.

We have included a deductive conceptual model (figure 2.2) that shows which determinable constructs we will study. In chapter 4 we will see to what extend our sub questions are answered by empirical evidence. First, in the following chapter, we will elaborate on how we studied determinable constructs.

(18)

3. Methodology

3.1 Research method

We conducted interpretive qualitative research by using a case study. Interpretive research is a research tradition that underlines the local intersubjective realities composed from subjective and objective meanings as perceived by actors (Gephart, 2004). Interpretive research is a way to recover and understand situated meanings and systemic divergences in meanings. Interpretive research is viewed as suitable since it is our goal to understand the complexity, engagement of stakeholders, cosmopolitanism and public attention for the specific implementation of UASs in the Netherlands. We studied determinable constructs in an iterative process, meaning that we repeated the research steps of consulting theory, collecting data and analyzing throughout the study (Glaser & Strauss, 1967). We conducted a case study where multiple organizations were involved. Including more organizations and their representatives will lead to findings that are more generalizable than conducting a single case study (Weick, 2007).

3.2 Sampling

Pettigrew (1988) argues that for the sample selection of cases it makes sense to study the extreme situations, in which the process of interest is observable in a transparent manner. We have discussed in the previous chapter that several organizations (i.e. stakeholders) are involved in the implementation of UASs in the Netherlands. Identified stakeholders at this moment of implementation are regulators and legislators on European, national and local level, the society, developers, operators and potential users (i.e. healthcare providers or care facilitators). Since data at all the involved organizations cannot be gathered due to limited time, we chose to collect most of the data at the ‘core involved organizations’, which provided us with the most observable and transparent process of implementation. Core involved organizations are developers, operators, healthcare providers and care facilitators. These organizations can technically realize UAS implementation. The organizations provide a transparent and observable implementation process. Besides, the organizations engage with each other to implement UASs. Putting those organizations in the sample can extend theory such as implementation determinants (Pettigrew, 1988). Therefore, is our sample based on theoretical sampling as prescribed by Eisenhardt (1989).

3.3 Data collection

We collected several types of data for the purpose of this case study. We gathered data in the form semi-structured interviews, informal and formal meeting observations, surveys and archival documents. Collecting several forms of data provides stronger substation of constructs due to triangulation (Eisenhardt, 1989). Triangulation enhances the credibility and validity of our findings (O'Donoghue & Punch, 2003). We collected data at different moments in time and places. Data was

(19)

collected from the beginning of February, until the beginning of May and we collected data throughout the Netherlands. Collecting data at different times and places reduces circumstance bias and enhances the reliability of data (van Aken, Berends & van der Bij, 2012). For an overview of all the collected data we included figure 3.1. Because of confidentiality and privacy reasons we anonymized all the data and findings about participating organizations and their representatives. To enhance the understanding of the sample, an overview of organizational characteristics is listed in appendix II.

Figure 3.1 Overview of data collection.

A significant proportion of the primary data was collected with semi-structured, in-depth interviews. The duration of the interviews ranged between 60 to 75 minutes and resulted in 122 pages of transcript. Targeted participants were those representatives with a managerial function, such as director, founder, innovation manager, process manager or project leader of core involved organizations. We met organizational representatives during formal meetings in February and March 2019. We recruited organizational representatives as participants by inviting them via e-mail to participate in the study (Hennink, Hutter & Bailey, 2010).

We attended four formal meetings between several organizations (i.e. stakeholder groups). These meetings lasted approximately 60 to 110 minutes. We took field notes during the meetings after

Semi-structured interviews and surveys

- Innovation Manager organization A - General manager organization B and B1 - Operations Manager organization C - Director and project leader organization

D

- Director organization E - General manager organization F - Process manager organization G

Informal meetings

- Two introductory conversations with Innovation Manager A

- Conversations about development of implementation with several participants (coffee, carpooling) with A and B/ B1

- Evaluation talks after formal meetings

- Company tour at organizations E and F

Formal meetings (observations)

- Meeting between organizations A, B, C and E

- Meeting at organization H between organizations A, B, C, D, G, I, J and others.

- Event meeting at organization H: open for all interested organizations.

- Meeting between organizations A, B, C and F

Archival

- Project proposals

- Project documentation (meetings and learning goals)

- Project evaluations - Declarations of intent forms - Collaboration forms

(20)

receiving verbal consent of the meeting’ attendees. During the observations we performed passive participation, meaning that the researcher did not interact or participate in the meetings, but observed and recorded from a nearby point (Hennink et al., 2010). Informal observations were those meetings with the company supervisor and conversations with the employees of the internship company. Some participants started the interview with a company tour, showing the departments and main processes of the organization. Notes were taken during these informal meetings.

3.3.1 Interview Protocol. To enhance data-collection reliability we conducted in-depth interviews by using a semi-structured protocol (Yin, 1994; Van Aken et al., 2012; Ahaus, n.d.). The protocol was based on open questions and these questions were predominantly based on the CFIR and the concepts of stakeholders and disruptive innovations. An example of a question is the one for complexity: ‘how complicated is the intervention?’ (CFIR Domains, n.d.). At the beginning of the interview we asked the participant about the purpose of implementing UASs. At the end of the interview we asked the participants how the ideal situation when UAS are fully implemented is seen (i.e. hypothetical situation). These questions helped us to understand what organizational goals UAS implementation should achieve and when implementation is perceived as successful. Other introductory questions and background information of concepts were included. This helped participants to be more familiar with the concepts and helped us in getting information in the direction of the CFIR constructs (Walle, 2015). We used a general interview protocol but made some adjustments for specific participant groups. For example, the question about how the implementation of UASs serves the patients was adjusted. Since this question is suitable for healthcare provider’ representatives, but less suitable for the operator, developer or logistic organization. Since it was a semi-structured interview some subjects were discussed more extensive by stakeholders (i.e. participants). We have discussed the technical complexity with three participants extensively (A, C, D), but this was less discussed by other organizational representatives. However, these participants (B, E, F) elaborated on healthcare specific demands to UAS implementation. The general interview protocol is included in appendix III. Questions were formed to be open and ‘free to fill in’ to make sure that participants feel like they could express important implementation matters (Hennink et al., 2010). We asked additional questions by using a probing technique. Anonymity was guaranteed to gain trust. We tried to eliminate issues of ‘response-behavior’ by making sure that participants felt free to express how they feel about implementation issues (Hennink et al., 2010). Sound recordings were made after permission of the participants was received. Before starting the interview, participants were asked to give their written consent about participating in the study. We used a consent form where we stated our expectations of the participant and what the participant could expect from us (the researcher). The used consent form is listed in appendix IV.

3.3.2 Starting point case study: Internship organization. Internship Organization is a subsidiary of a Dutch academic hospital. The organization provides emergency and traumatic care to citizens in two provinces. The organization has provided us the opportunity to study determinable

(21)

constructs to the implementation of UASs. Our study began with attending the formal meetings that formed observations. Internship Organization made it possible for us to attend these meetings. Further we were able to collect archival data at the organization. Several types of archival data were collected such as project proposals, evaluative reports, learning goals documentation, and collaboration forms. 3.3.3 Indication survey. In this study we focused on specific CFIR constructs namely complexity, stakeholder engagement, cosmopolitanism and public attention. Accordingly, we limited interview-questions about other determinable CFIR constructs. However, we did aim to obtain some information about the other constructs since other constructs might be determinable for the implementation success of UASs. Therefore, we added a digital survey about the remaining CFIR (sub) constructs. The survey was collected to get indications about other important implementation constructs. The collected surveys do not fulfill any statistical purpose in this study. The participants were asked to fill in the survey a week before conducting the in-depth interview. The survey was sent in advance because it provided us with some ideas about points of attention as seen by the participant. The survey constituted 45 statements that could be answered on a five-level Likert-type scale. We adjusted for some of the CFIR constructs in the survey by eliminating some constructs or by splitting constructs into more questions. For instance, we eliminated questions about the ‘source of intervention’ since we already knew that UASs will be developed outside ‘potential user organizations’. For some (sub) constructs distinctions were made. For example, the sub construct of leadership engagement and commitment. We did the same by splitting questions about quantitative and qualitative feedback. We choose to split these (sub) constructs because scholars, such as Ilott et al., (2012) and Breimaier et al., (2015), suggested that refinement and limitations in the broad terminologies during data collection should be made. The survey is listed in appendix V.

3.4 Data analysis

The interview transcripts and field notes of formal meetings were transformed to Rich Text Format. Consecutively, the transcripts and notes were coded in a software analytic tool, namely ATLAS t.i. The choice to work with this program helped us in coding and analyzing data in a structured manner and enhances the reliability of our study (Van Aken et al., 2012). We have used several types of coding, such as open coding, axial and selective coding (Wolfswinkel, Furtmueller, Wilderom, 2013; Glaser & Strauss, 1967). We combined deductive and inductive codes following Hennink et al., (2010). The deductive codes were based on theoretical concepts and inductive codes arose by reading the transcripts. We choose to combine codes, because theory building occurs in an ongoing dialogue between pre-existing theory and new empirical generated insights (Liamputtong & Ezzy, 2005). We have included our codebook as well as the process of coding and analyzing in appendix VI.

The collected data in the form of surveys was analyzed in Excel. The survey results were not statistically analyzed, because we did not pursue any form of quantitative research. The main purpose

(22)

of the survey was to shed light on constructs that were not studied in-depth (Eisenhardt 1989; Yin, 2014). We analyzed the archival data by reading analytically and allocated code groups in a more pragmatic manner; we used a color scheme and allocated the codes of ‘barrier’ or ‘facilitator’ to the emergent conceptual codes regarding complexity, stakeholder engagement, cosmopolitanism and public attention.

(23)

4. Findings

In this section the findings are presented. These findings are based on the analyses of the in-depth interviews, observations, archival data and surveys. Most of this chapter is devoted to presenting the findings about the implementation determinants complexity, stakeholder engagement, cosmopolitanism and public attention. At the end of this chapter we will briefly discuss the indications and conformations that were found after analyzing the surveys.

4.1 Understanding complexity for UAS implementation

In section two we outlined our expectations about the complexity construct. We expected that UASs are difficult to implement because the systems are disruptive. Due to disruptiveness, we expected that the implementation duration will be long for UASs. To gain a deeper understanding of the potential determinable construct we presented the sub question: how does disruptiveness lead to complexity and

therefore barriers successful UAS implementation?

Before we move to complexity because of disruptiveness, ‘complexity in general’ must be discussed. We found that the degree of UAS complexity is high. One major component of this complexity is technical complexity. Especially the representatives of the operating and manufacturing organizations (C, D) elaborated on the technical complexity of UASs. Technical complexity was expressed as the required technological properties of UASs and the difficulty of meeting those requirements. The technical complexity of UASs is high because the systems need to contain multiple sub-systems to ensure safe flights. For instance, the systems should contain software to stabilize vehicles, operating software to ensure ground control management, and communication systems (Representative D). Besides, the systems need to meet certain requirements that are prescribed by healthcare providers that aim to transport medical products with UASs (Representatives A, E, F). European regulation and legislation contain strict safety requirements to make sure that UASs will technically function in a safe and appropriate manner. Most representatives discussed that regulation influences the technical complexity and hinders implementation in this phase. One major barrier due to current regulation at this moment is that ‘beyond visual line of sight’ flights, abbreviated as BVLOS, are prohibited (Representatives D, G, Archival documents).

The representatives elaborated on the effect of regulation when discussing complexity. However, the effect also disseminates in an existing CFIR construct: external policy and incentives. This construct is about the degree to which external strategies spread interventions and includes regulation (Damschroder et al. 2009). As mentioned, current regulation barriers implementation (Balasingam, 2017). Thus, it can be said that external policies and incentives barrier implementation. Yet, during data collection we heard statements about ‘ongoing development’ in the pre-implementation phase. This ongoing development is about technological and regulatory development. Regulatory development should lead to new regulation that becomes effective in 2020. This regulation should make it possible

(24)

to fly BVLOS when complied with strict safety requirements (Representatives A, C, D, Observations, Archival documents). This indicates that external policies not necessarily barriers successful (future) UAS implementation in a direct line. However, as discussed by the representatives: regulatory requirements enhance the technical complexity of UASs, therefore can barrier implementation indirectly. To make a long story short: the existing CFIR construct of external policies was found to interact with the construct complexity.

4.1.1 Complexity due to disruptiveness. Now that we have captured the general complexity of UASs and the role of regulation (i.e. external policies) on this complexity we resume with the findings about the earlier detected potential determinable constructs. We found that UASs are viewed as ‘new technologies’ in healthcare settings and that this leads to complexity:

‘This is complex, because it is new. There are so many situations where you have certain logistic issues in an organization which have the same base [...] however this is an entirely different structure. And that is what leads

to complexity.’ (Representative B)

‘Complex because it is new’ refers to the disruptive character of UASs. Since, disruptive interventions are new forms of technology (Hacklin et al., 2014). Srinivasan (2013) argued that interventions are disruptive when it constitutes new technologies, is value networked and leads to new business models. Several representatives spoke about how UASs could change business models. Hence, how UASs could replace traditional ways of transporting healthcare products:

‘This can turn the world upside down […]. Everything changes, but I am not that scared of the existing disappearing […]. Only it will ask, and this is what I can do about it, something of our own organization. […] So, what is disruptive about it; it will change the way you think and maybe it is possible to transport everything

through the air.’ (Representative E)

‘I call it disruptive opportunities, and the disruptiveness is not because of the technology, but because of the business models.’ (Representative A)

Some representatives expressed worries about implementing such new systems because of needed changes in the current organizational processes:

‘I think that it is really, really, really complex. Because, […] the flexibility and the stretch in the current organization at the several departments is gone. There have been so many reorganizations and budget cuts that

the current work packages […] are full.’ (Representative B)

Similar worries were pointed out by representatives E and F as it is expected that the organizations need new staff to work with such systems. Besides, the representatives argued that employees must be extensively trained to work with these systems. Accordingly, training and employing new staff is complex, because transitioning is viewed to be difficult (Representative E, F). Also, representatives B and E spoke about how current employees will not be happy with this transition, because of the

(25)

possibility to lose their jobs. Internal resistance in the organization is expected. It seems that in the pre-implementation phase it is difficult to get staff on board because of ‘the new and unknown meaning of UASs’:

‘I see that people get scared. […] Drones? Well it is far away from their view of logistic processes in an organization. […] when you elaborate on it, they say this could be interesting. And if you ask more about it,

trying to make agreements they say we will need to discuss this by ourselves.’ (Representative B)

When analyzing complexity due to disruptiveness we found emerging concepts. As mentioned, representatives referred to ‘ongoing development’, ‘the unknown’ and ‘newness’ of UASs. All these concepts refer to disruptiveness, thus complexity, as illustrated by representative D:

‘Thus, the technology […] which is new, needs to be completely developed and tested. And it does not exist yet. […] The regulation is in the beginning phase and must be developed. […] the development is quite complex.’

The three concepts mentioned above capture an emerging concept: ‘the chicken-and-egg matter’. The matter has been mentioned by name by three representatives (A, C and D). The matter refers to the absence of current UAS business models in the healthcare industry and the need to develop these models. For the development of UAS business models ‘mutual dependencies’ exist among stakeholders who aim to implement the systems. The chicken-and-egg matter illustrates how disruptive UASs are in the Dutch healthcare system. When we asked about the purpose and aim of implementing UASs for medical missions, representative C answered:

‘Eventually do we need to have an answer on that question […] it is still about searching where the value is that you add with this new modality and I believe that it is not answered yet. The clearer a problem is, the easier a solution […] we do not have that yet. […] And that is the chicken-and-egg. Of course, when you have a solution

the demand will follow. […] this service is not crystallized yet […]. The outcomes are still unsure.’

The remaining representatives (B, E, F, G) also referred indirectly to the chicken-and-egg matter. For instance, representative B believed that clients (i.e. healthcare providers) do not know that they will demand UASs in the future. Representative G referred to a non-existent UAS market for medical missions at this moment in time.

For the creation of the UAS market and business models, mutual dependencies exist between the implementing stakeholders. We will discuss this more in the paragraph about understanding stakeholder engagement.

Another emerging concept that considers UAS complexity and disruptiveness was found: ‘the integration matter’. All the representatives mentioned the integration matter during some point in the interviews. We define this emerging concept as a difficulty to place an intervention in an existent setting. The integration matter regarding UASs distinguishes an aerial integration issue and a civilization integration issue. The aerial integration issue is about the difficulty of placing UASs in the existing

(26)

setting of aviation. Whereas the civilization integration issue refers to the difficulty of placing UASs in the existing public space. Some representatives refer to the integration matter as the biggest barrier to implementation:

‘The biggest challenge at this moment and the biggest complexity is the integration of this new means in a very crowded society. Crowded in the air as well on the ground. […] Just dropping a new drone airport […] and suddenly is the civilization bothered with 1000 flights a day. If you want to do this in the Netherlands, you need to comply with very high safety requirements and that is where the complexity lies. […] and because drones are small, land and start everywhere, and fly relatively low you are also going to integrate the systems in a public

space.’ (Representative A)

The aerial integration issue interacts with technical complexity, thus external policies (which are under development). According to representative C, legislative and regulatory facilitation is needed for aerial integration:

‘Complexity lies in the fact that a lot of matters need to be arranged in the Netherlands if a drone wants to fly safely through the airspace. […] This is very complex. But this is something actually that is out of our hands.’

The civilization integration issue was found to interact with public attention. Therefore, we will discuss this issue more in-depth in the paragraph about the new construct.

4.1.2 Complexity reflected by expected duration. We found that the expected duration of implementation will take several years according to all the representatives. How many years differed. Arguably, the duration of implementation depends on what the organization aims to achieve with UASs. For instance, some organizations (E and G) aim to implement UASs for missions on pre-set routes between point A and B. Whereas others (Organizations A, C and D) aim to implement UASs where an UAV can fly from point A to X. Implementation duration also depends on whether an UAV needs to fly at night and, or during poor weather conditions (Representative A). It is expected that UASs that will fly over pre-set routes can be implemented faster; in a period of two to five years. More difficult missions are expected to be fully operationalized in five to ten years. Several representatives argued that the implementation of UASs depends on the development of regulation, legislation and UAS business model(s). Representative A argued that regulators determine the implementation pace. We found that an important component of UAS development and its duration is the interaction with ‘cascading risky development’:

‘A building plan that builds upon gaining experiences and this experience is translated into a SORA […] and a SORA means that you establish a standard mission. This standard operation analyses the complete risk of the mission and determines for which requirements you need to comply to carry out the mission.’ (Representative A)

Thus, specific operations risks assessment (SORAs) are an important component of the UAS development. The representatives of organization A, C and D spoke extensively about cascading and risky learning and the role of regulation in the development of SORAs. We found that including SORAs

DETECTING DETERMINABLE CONSTRUCTS TO THE IMPLEMENTATION OF UNMANNED AIRCRAFT SYSTEMS BY USING THE CONSOLIDATED FRAMEWORK FOR IMPLEMENTATION RESEARCH WHEN THE SKY IS NOT THE LIMIT:

WHEN THE SKY IS NOT THE LIMIT: