How to overcome the challenges of Internet of Things to ensure successful technology integration: A case study at an Aerospace manufacturer

(1)

How to overcome the challenges of

Internet of Things to ensure successful

technology integration

A case study at an Aerospace manufacturer

Viktor Berger

Sakib Chowdhury

Industrial and Management Engineering, master's level 2021

Luleå University of Technology

(2)

ABSTRACT

Purpose - The purpose of this study is to investigate how the challenges of Internet of Things

that manufacturers can influence can be overcome.

Theoretical foundation - This study conducted an extensive literature review to identify and

understand the challenges to Internet of Things and actions to overcome them. 11 technical challenges, 13 organisational and 6 resource availability challenges were identified. 4 actions were identified.

Method - To fulfil the purpose, an embedded multiple case study at a global Aerospace

manufacturer was conducted. 7 unstructured interviews, 12 semi-structured interviews and a survey were conducted. Respondents were picked due to their experience in Internet of Things projects and relevant technologies. The survey was conducted to evaluate the challenges’ relevance to high-technology manufacturers, on a 7-point Likert scale. The semi-structured interviews aimed to find actions to overcoming the challenges relevant to high-technology manufacturers.

Findings - The evaluation of the challenges relevance to high-technology manufacturers

resulted in 10 common, 9 occasional and 2 uncommon challenges. 12 actions to overcoming the challenges and their tasks were identified.

Theoretical contribution - The study provides a comprehensive list of potential challenges to

Internet of Things. It evaluated the challenges’ impact on high-technology manufacturers, thus challenging and validating Internet of Things challenges presented in literature. It provides a set of actions and a framework which aids in overcoming challenges that impact high-technology manufacturers’ Internet of Things initiatives, thus contributing to digital change management. Finally, it aids in the progress towards concretising Industry 4.0 and its trends toward connectivity, intelligence, and flexible automation.

Practical contribution - The study provides an increased understanding of potential challenges

of Internet of Things, and recommendations to how high-technology manufacturer can overcome the challenges. A framework is provided which gives an overview of which actions, and subsequent tasks, to take to overcome a specific challenge.

Keywords - Internet of Things, Identifying challenges, Technical challenges, Organisational challenges, Resource availability challenges, Overcoming challenges, High-technology, Manufacturers, Industry 4.0, Connectivity, Intelligence, Flexible automation, Digitalisation, Innovation

(3)

ACKNOWLEDGEMENTS

This report represents the master’s thesis of Viktor Berger and Sakib Chowdhury. This study is the epitome of our Master of Science in Industrial Engineering and Management, a master’s degree in Innovation and Strategic Business Development at Luleå University of Technology (LTU).

This degree project was carried out during the spring of 2021 in association with a global Aerospace manufacturer. We would therefore like to thank our supervisors and colleagues at the case organisation that provided us with the foundation, feedback, and respondents that we required to conduct this study. Additionally, we would like to thank the supervisors for inviting us to contribute to the organisation’s Internet of Things standard project.

Furthermore, we would like to express our gratitude to Mats Westerberg, our supervisor at LTU, for the invaluable feedback and subsequent discussions that guided us during the degree project. We would also like to thank our fellow students that took time from their own degree projects to provide feedback and increase the quality of this study.

Finally, we would like to acknowledge our families and friends for the immeasurable support we were given during our studies.

Viktor Berger Sakib Chowdhury

(4)

TABLE OF TABLES

Table 1: Challenges of Internet of Things and their various causes. ... 2

Table 2: Potential benefits of Internet of Things and their origin. ... 2

Table 3: Shortcomings mentioned in literature and this study's aim. ... 3

Table 4: Previous similar studies. ... 6

Table 5: Overview of the technical challenges, in order of appearance. ... 7

Table 6: Overview of the organisational challenges, in order of appearance. ... 11

Table 7: Overview of the resource availability challenges, in order of appearance. ... 16

Table 8: Chosen methods. ... 21

Table 9: Respondents in unstructured interviews in phase one. ... 24

Table 10: Respondents in semi-structured interviews in phase two. ... 25

Table 11: 7-point Likert scale used to evaluate the challenges. ... 27

Table 12: Thematic analysis example. ... 29

Table 13: Rating of the technical challenges. ... 33

Table 14: Rating of the organisational challenges. ... 34

Table 15: Rating of the resource availability challenges. ... 36

Table 16: The challenges deemed relevant by the case organisation and their frequency. ... 37

Table 17: Overview of the actions. ... 38

Table 18: Quotes relevant to adapt solution and the challenges they help overcome. ... 38

Table 19: Quotes relevant to assigning resources and the challenges they help overcome. . 39

Table 20: Quotes relevant to clarifying responsibilities and the challenges they help overcome. ... 41

(8)

Table 21: Quotes relevant to collaborate and the challenges they help overcome. ... 43

Table 22: Quotes relevant to educate and the challenges they help overcome. ... 44

Table 23: Quotes relevant to prestudy and the challenges they help overcome. ... 46

Table 24: Quotes relevant to communicate and the challenges they help overcome. ... 50

Table 25: Quotes relevant to adapt culture and the challenges they help overcome. ... 52

Table 26: Quotes relevant to create reward system and the challenges they help overcome. ... 53

Table 27: Quotes relevant create security measures and the challenges they help overcome. ... 54

Table 28: Quotes relevant to decide technology features and the challenges they help overcome. ... 55

Table 29: Quotes relevant to support and the challenges they help overcome. ... 57

Table 30:Theoretical contributions. ... 60 Table 31: Thematic analysis. ... VII

(9)

1. INTRODUCTION

The manufacturing industry’s digitalisation progress lag, which according to Paiola & Gebauer (2020) depends on the necessity of adapting the business’ traditional culture to fit the digitalised manufacturing environment. Since there is a limit to how much organisations can gain from conventional methods, digital innovations are required to stay competitive. Due to the lack of progress, manufacturers are not benefitting from digitalisation’s opportunities to the extent they could. Literature regarding digitalisation technologies varies in availability and maturity. Buzzwords and the great number of technologies associated with digitalising manufacturing industries further dilute the comprehension of them. One such buzzword is Industry 4.0, the fourth industrial revolution, that entails trends such as connectivity, intelligence, and flexible automation (WEF, 2019). Frank et al. (2018), Oztemel & Gursev (2020) and Lu (2017) presents a wide array of Industry 4.0 technologies, one of them being Internet of Things, commonly referred to as IoT.

To aid in concretising the current digitalisation trends, and to delimitate the study, it centres on Internet of Things. This improves its potential to provide practically useful knowledge for future projects. This choice was further based on (1) the number of articles written on the topic, (2) the case organisation’s various initiatives in the area and (3) the case organisation’s potential to benefit from the study’s focus. This delimitation is also in line with Moeuf et al. (2019) that recommends practitioners to discover the opportunities of the current digitalisation trends by implementing Internet of Things. Adapted from Oztemel & Gursev (2020), this study defines Internet of Things as connecting physical objects to one another, to systems and other technologies with the aim of collecting and sharing data and allowing for data-driven decision making as well as remote real-time visualisation and control.

1.1 Problem discussion

Internet of Things entail considerable challenges such as financial barriers, lack of technical capabilities and data security issues to name a few (Boehmer et al., 2019; Haddud et al., 2017; Ingemarsdotter et al., 2021; Paiola & Gebauer, 2020; Raj et al., 2020; Sisinni et al., 2018; Sung, 2018). A variety of challenges of Internet of Things and their potential causes can be seen in Table 1.

(10)

Table 1: Challenges of Internet of Things and their various causes.

Challenge Explanation

Data needs Misalignment of data needs between functions or lacking understanding of end-users

Scalability Limited capacity of device management systems

Talent shift Competing for talented individuals with the ability to create appropriate solutions

The challenges of Internet of Things need to be overcome to fully benefit from the technology, a few benefits and their origins can be seen in Table 2.

Table 2: Potential benefits of Internet of Things and their origin.

Benefit Origin

Improved efficiency Connected sensors relay information regarding overall equipment efficiency

Reduced downtime Connected sensors relay information valuable to maintenance, e.g., vibrations or heat

Improved monitoring and control A multitude of objects can be monitored and controlled from one place

Improved decision making More data to base decisions on

The shortcomings in literature regarding studies of Internet of Things and how this study will aim to solve these are presented in Table 3.

(11)

Table 3: Shortcomings mentioned in literature and this study's aim.

Article Mentioned shortcoming Current study’s aim

Sony & Naik (2019) Lack of case studies

This study is a case study of a high-technology manufacturing organisation Kang et al. (2016) Characteristics and requirements of

businesses must be considered

Fatorachian & Kazemi (2021) Organisational and cultural factors need to be considered

This study will investigate the challenges presented in literature and aim to evaluate their relevance to practitioners

Goad et al. (2021); Raj et al. (2020) Challenges in literature need to be validated or evaluated

Haddud et al. (2017) Challenges must be validated by conducting a survey towards practitioners

Haddud et al. (2017); Xu et al. (2018) Must investigate how challenges can be

overcome _{To aid organisations in benefitting from} Internet of Things, this study will aim to provide recommendations to how the relevant challenges can be overcome Osterrieder et al. (2020)

Behavioural, economic, and managerial aspects need to be considered to overcome challenges

Chiarini et al. (2020)

Researchers and practitioners experience difficulties in approaching novel digitalisation technologies

This study will aim to provide a framework or a roadmap that can guide organisations in approaching Internet of Things. Oztemel & Gursev (2020); Schumacher et

al. (2016)

Need for assessment and evaluation methodologies

Kang et al. (2016) Need to establish standard models and services

Sung, 2018 Lack of a common narrative to guide organisations

As implied from previous remarks, the implementation of digital technologies force change upon manufacturers’ way of doing business. This calls for an increased understanding of change management among manufacturers aiming to integrate Internet of Things with their business. However, there are few studies relating to how organisations are digitally transformed (Warner

(12)

& Wäger, 2019), this may be the reason for the lack of frameworks or roadmaps in literature. A framework or roadmap would aid organisations in understanding how to smooth the transition to a digital work environment (Ghobakhloo, 2018). This study defines change management as the process that supports organisations, and their people, in conducting and maintaining organisational change that is consistent before, during and after its implementation (Kramer & Magee, 1990).

1.2 Purpose

In line with previous discussions, the purpose of this study is to investigate how the challenges of Internet of Things that manufacturers can influence can be overcome. The following research questions have therefore been formulated:

RQ1: What are the challenges of using Internet of Things to digitalise and how should they be prioritised by managers in high-technology manufacturing organisations?

RQ2: How can the challenges of Internet of Things be overcome by managers in high-technology manufacturing organisations?

(13)

2 THEORETICAL FOUNDATION

The theoretical foundation highlights the challenges of Internet of Things implementation. The construction of the theoretical foundation is illustrated in Figure 1.

Clarification of concept Literature review

Internet of Things

RQ1 RQ2

Challenges of Internet of Things Overcoming challenges of Internet of Things Figure 1: Construction of the theoretical foundation.

2.1 Internet of Things

Internet of Things is first and foremost related to connectivity, as previously defined it is about connecting physical objects to one another, to systems and other technologies with the aim of collecting and sharing data and allowing for data-driven decision making as well as remote real-time visualisation and control. It can be connected to several technologies; the related technologies are presented in Figure 2.

(14)

It is related to intelligence as the data that is collected allows for data-driven decision making, and if e.g., integrated with machine learning and artificial intelligence massive amounts of data can be processed and analysed in real-time. If integrated with augmented or virtual reality, machines could be analysed remotely and previously obstructed areas could be inspected without stops to production, as sensors could provide data in real-time. Moreover, it relates to flexible automation, such as when integrated with additive manufacturing. Information regarding a specific part could be available directly on the product itself, for example through a printed Quick Response-code (QR code) that provides information regarding maintenance, material or similar.

2.2 Challenges of Internet of Things

Studies regarding the challenges of Internet of Things have been conducted previously, however, in a different manner than that of this study, see Table 4.

Table 4: Previous similar studies.

Article Investigated Method Result Shortcoming

Haddud et al. (2017)

Studied potential benefits and challenges with Internet of Things with a focus on Supply Chain Management

Quantitative study; survey; 87 academics in North America, Asia, Europe, South America, Africa and Australia;

Divide challenges into technical, organisational and resource availability challenges; 17 benefits and 15 challenges are related to individual organisations; 18 benefits and 15 challenges are related to the entire supply chain

Need to be verified

Sisinni et al. (2018)

Studied the technical aspects of Internet of Things

Literature review

Mentions opportunities and challenges regarding energy efficiency, real-time performance, coexistence, interoperability, and security and privacy

Lacks many of the challenges that Haddud et al. (2017) brought up

Paiola & Gebauer (2020)

Studied opportunities and challenges for Business-to-Business firms with a focus on service-oriented business models

Qualitative study; interviews; in 2016 and 2017

Opportunities and challenges to digital disintermediation; opportunities and challenges of using Internet of Things to either provide improvements to customers’ business processes like remote condition monitoring or to aid them in achieving specific business outcomes such as to monetise data

Determines and adds further organisational challenges; lack the technical and resource availability challenges of previous studies

(15)

While the previous studies are similar to this study, they are not collectively exhaustive as they only present some of the challenges established in literature. They do not fully evaluate the challenges of Internet of Things implementations nor provide suggestions of how to overcome the challenges.

The challenges of Internet of Things are grouped into technical, organisational and resource availability challenges, in accordance with Haddud et al. (2017).

2.2.1 Technical challenges

An overview of the technical challenges discovered in literature and further described below can be seen in Table 5.

Table 5: Overview of the technical challenges, in order of appearance.

Technical challenge

1. Cyber security 2. Data availability and accuracy

3. Data storage and distribution 4. Data processing and analysis

5. Software complexity 6. Infrastructure

7. Interoperability and compatibility 8. Usability

9. Scalability 10. Energy efficiency

11. Production flexibility and technology integration

Cyber security is a cause for concern for digitalising organisations. With devices such as smart

sensors that collect and uploads data to a data storage via a network, there is a risk of device and network security vulnerabilities (Bibby & Dehe, 2018; Haddud et al., 2017; Sharma et al., 2020; Xu et al., 2018; Machado et al., 2019; Ancarani et al., 2019; Kamble et al., 2020) as security and privacy measures are commonly not present in such devices (Goad et al., 2021). Small devices

(16)

may lack the computing and storage capacity for complex security software (Sisinni et al., 2018). Vulnerabilities increase the more systems an organisation uses (Raj et al., 2020; Alcácer & Cruz-Machado, 2019), as organisations commonly use several systems, the security of the connected systems is only as high as the least secure one. Additionally, old devices or lack of software updates creates possible vulnerabilities (Alcácer & Cruz-Machado, 2019). It is possible for attackers to exploit the vulnerabilities to gain access to sensitive data (Chiarini et al., 2020; Ghobakhloo, 2018; Luthra & Mangla, 2018). There is a need to protect sensitive data as it may include intellectual property, weapons manufacturing, or other confidential information (Sung, 2018). Organisations commonly underestimate the risks of cyber-attacks (Moeuf, et al., 2019), and with increased digitalisation, cyber-attacks have become more common. Work required to adhere to the security concerns of the industry can be extensive (Xu et al., 2014) as organisations need to do a lot of the research themselves since involving third parties in the security of the organisations creates additional risks (Raj et al., 2020). There is also a lack of standards, laws and regulations (Calabrese et al., 2020; Xu et al., 2014) regarding cyber security to guide organisations. Due to constant innovations, new encryption and security measures need to be created to eliminate vulnerabilities in devices, networks, and storage (Sisinni et al., 2018; Lu, 2017).

Data availability and accuracy is affected by the sensor technology used (Awan et al., 2021),

such as the quality of the data, the type of data collected and whether the data is delivered in real-time or not (Ingemarsdotter et al., 2021; Haddud et al., 2017). However, it can be difficult to change what data is collected over time, and therefore, choosing the wrong sensor technology can cause poor data availability and reliability (Sharma et al., 2020; Luthra & Mangla, 2018; Ingemarsdotter et al., 2021; Sisinni et al., 2018). The availability and accuracy of the data is also dependent on where the sensors are placed, such as their location in the manufacturing process or on the products (Bibby & Dehe, 2018; Alcácer & Cruz-Machado, 2019). Moreover, data has a limited useful life (Alcácer & Cruz-Machado, 2019) and the sensor technology and the network stability affect whether the data is updated in time or delivered in real-time (Sahal et al., 2020). Another issue here lies in the fact that devices may cause electromagnetic interference (Sisinni et al., 2018) as cheaper alternatives may leak power over to adjacent channels, thus causing interference with devices on the adjacent channels.

Data storage and distribution can affect the reliability of the data (Fakhar Manesh et al., 2021),

as it can become difficult to handle large quantities of data without sufficient storage options. Large volumes of data, especially pictures and videos (Urbina Coronado et al., 2018), causes

(17)

problems for unprepared organisations as the storage needs quickly becomes massive (Haddud et al., 2017; Lu, 2017; Alcácer & Cruz-Machado, 2019; Kamble et al., 2020; Raj et al., 2020; Sahal et al., 2020). As distribution of data between units within the organisation becomes essential (Gooneratneet al., 2020), especially in critical machine-to-machine communications (Sung, 2018) or safety-critical data (Lu et al., 2020), large volumes of data can cause delays and data loss if the amount of data distributed reaches a point above the capacity of the network (Xu et al., 2014; Xu et al., 2018; Oztemel & Gursev, 2020; Olsen & Tomlin, 2020). Such connectivity issues can eventually cause system failures (Sharma et al., 2020) that could potentially cause stops in production which would be devastating for a manufacturer.

Data processing and analysis must have the capacity to handle large volumes of data

(Dalenogare et al., 2018), whether it is structured or unstructured. To achieve transparency and productivity through large volumes of data, it first needs to be processed (Lu, 2017) which requires new processing and analysis tools (Yin et al., 2017; Bibby & Dehe, 2018; Xu et al., 2018; Haddud et al., 2017) such as big data management systems. It requires digital data analytics techniques used by data scientists (Xu et al., 2018; Chiarini et al., 2020), as there is a difficulty in extracting useful data due to the large volumes (Raj et al., 2020; Sahal et al., 2020; Olsen & Tomlin, 2020). It may even require artificial intelligence (Rauch et al., 2020; Oztemel & Gursev, 2020) to process and analyse the data with the speed required to act upon it (Alcácer & Cruz-Machado, 2019). There is a need for low latency operations here as well (Sahal et al., 2020), for the system to respond to changes in a timely manner.

Software complexity can cause issues depending on the needs of the organisation (Luthra &

Mangla, 2018), as applications and algorithm coding can be a barrier to using data productively (Agostini & Filippini, 2019; Haddud et al., 2017; Olsen & Tomlin, 2020). It can add to the learning curve for developers as there is commonly a need for substantial custom code (Sahal et al., 2020). Organisations may need to hire new individuals with the right expertise (Machado et al., 2019) or find external software. Software complexity can also lie in the management of all the devices, such as addressing, identification and optimisation (Xu et al., 2014; Xu et al., 2018).

Infrastructure is a challenge (Sharma et al., 2020; Oztemel & Gursev, 2020; Calabrese et al.,

2020; Raj et al., 2020; Karadayi-Usta, 2020; Ghobakhloo, 2018) as a lack thereof requires upgrades (Mittal et al., 2018; Xu et al., 2018; Yin et al., 2017) such as universal communication systems (Bibby & Dehe, 2018) and networks (Haddud et al., 2017) with speeds that enable data transmissions required for large data volumes (Agostini & Filippini, 2019). The data transmissions

(18)

are also affected by the communications and signal coverage within the organisation (Haddud et al., 2017), such as at the shop floor.

Interoperability and compatibility issues which can affect data sharing between systems and

(Urbina Coronado et al., 2018) may occur due to a lack of data communication standards (Lu et al., 2020; Chiarini et al., 2020; Xu et al., 2014; Sharma et al., 2020), aged machines (Chiarini et al., 2020) or legacy systems (Gooneratneet al., 2020). These issues can also be caused by sensors, networks and systems brought in from different external organisations or consultants (Haddud et al., 2017). Previously used systems may have over-engineered features (Kohtamäki et al., 2020) which are not easily adaptable to new technology or methods of working. Thus, the lack of interoperability and compatibility can affect the cost and complexity of implementing new technology (Sisinni et al., 2018). There is a lack of standards and certifications for Internet of Things (Raj et al., 2020) and the increasing rate of innovation makes it difficult to create standards (Xu et al., 2014; Xu et al., 2018). The absence of standards for Internet of Things, or not following them, can cause issues with interoperability and compatibility (Schumacher et al., 2016; Awan et al., 2021) as it increases the risk of several different methods being used across the organisation. Furthermore, lack of standards can cause data issues (Lu, 2017) such as inputting manual data differently into Internet of Things systems at different areas of the organisation.

Usability is therefore an additional challenge. There is data that may have to be entered

manually, such as metadata that needs operators or other essential support staff to enter it accurately for it to be processed correctly by the system (Ingemarsdotter et al., 2021; Urbina Coronado et al., 2018). User-friendly software is needed (Nižetić et al., 2020) to remove distractions that potentially lead to workplace accidents (Oztemel & Gursev, 2020) and to ensure staff knows what action to take on what data or error code (Ingemarsdotter et al., 2021). In multinational or global organisations, usability is affected by the number of languages included in software or hardware-related documents (Lu, 2017).

Scalability need to be considered when choosing technology, as it is limited in some hardware

and software solutions (Kamble et al., 2020; Chiarini et al., 2020). Scalability issues can occur with device management systems as their capacity may be limited to a specific number of devices (Xu et al., 2018). Similarly, data storage has limited scalability, such as SQL-based solutions (Sahal et al., 2020).

Energy efficiency needs to be addressed for various reasons. Some devices have to run on

(19)

consumption (Nižetić et al., 2020) and may well limit Internet of Things application and adoption in manufacturing.

Production flexibility and technology integration are two challenges that depend on one

another. It is necessary to maintain the integrity of the manufacturing processes, as stops in production can quickly become expensive (Ghobakhloo, 2018; Sung, 2018). Operational complexities and a lack of readiness for innovation can cause wasteful processes to be digitalised (Agostini & Filippini, 2019; Bag et al., 2021; Karadayi-Usta, 2020), it is therefore a challenge to ensure production systems are mature enough to handle Internet of Things implementation (Alcácer & Cruz-Machado, 2019). Moreover, there is the possibility that some products, processes, or machines are unable to be simulated or integrated with Internet of Things (Moeuf, et al., 2019), such as due to customer requirements. Another issue here lies in deciding when to test the integration of the technology, it can be challenging as testing it in a premature stage can cause expensive delays across the supply chain (Raj et al., 2020).

2.2.2 Organisational challenges

An overview of the organisational challenges discovered in literature and further described below can be seen in Table 6.

Table 6: Overview of the organisational challenges, in order of appearance.

Organisational challenge

1. Resistance to change 2. Data needs

3. Strategy 4. Prioritisation of tasks

5. Creating a sense of ownership and responsibility 6. Collaboration

7. Technology comprehension 8. Data interpretation

9. A talent shift 10. Employee displacements

11. Security procedures 12. Incentives

(20)

Resistance to change can affect the commitment of end-users as well as people involved in

Internet of Things projects (Boehmer et al., 2019; Schumacher et al., 2016; Luthra & Mangla, 2018; Mittal et al., 2018; Haddud et al., 2017). End-users’ resistance can be caused by fear of job destruction (Ancarani et al., 2019), such as through automation (Stock et al., 2018; Sung, 2018; Raj et al., 2020). Involving people in Internet of Things projects can be challenging (Gooneratneet al., 2020; Rauch et al., 2020) as manufacturers with long standing traditions and ways of working tend to resist change (Calabrese et al., 2020; Raj et al., 2020; Ingemarsdotter et al., 2021). User resistance also forms due to privacy issues (Kamble et al., 2020; Oztemel & Gursev, 2020; Xu et al., 2014; Xu et al., 2018) as continuous data collection of employees and their processes can be regarded as surveillance (Moeuf, et al., 2019; Rauch et al., 2020; Sung, 2018; Sisinni et al., 2018) and as employees may be uncertain of what data is collected and its use (Goad et al., 2021; Machado et al., 2019).

Data needs can diverge between units and functions in organisations (Machado et al., 2019)

and it can therefore be hard to decide what data to collect. Additionally, there are numerous alternatives for data collection (Chiarini et al., 2020), choosing the appropriate one that fit the needs of the organisation can be a significant task (Olsen & Tomlin, 2020). Understanding end-user needs, their relations to one another and the flow of information between them is essential to guarantee that data is generated for a purpose (Agostini & Filippini, 2019; Alcácer & Cruz-Machado, 2019; Boehmer et al., 2019), otherwise organisations risk limiting the scope of the technology implementation to certain users (Ingemarsdotter et al., 2021), and thereby its benefits. Without a clear strategy, data collection is driven by what is technically possible rather than the needs of the organisation (Ingemarsdotter et al., 2021; Karadayi-Usta, 2020).

Strategy is therefore a challenge as short-term strategies increase the risk for unsuccessful

technology implementation (Moeuf, et al., 2019). Internet of Things, and associated digital strategy, need to be embedded in the central mind-set of the organisation (Machado et al., 2019). If a digital strategy is lacking, it must be established (Moeuf, et al., 2019), as the need for the technology and the continuous innovation mind-set must be communicated across all levels of the organisation (Agostini & Filippini, 2019; Raj et al., 2020; Schumacher et al., 2016) for the technology to succeed (Mittal et al., 2018). A lack of understanding about the environment where the technology is to be implemented further risk decreasing its benefits (Olsen & Tomlin, 2020). Furthermore, the lack of roadmaps provided, in literature and from governments, can

(21)

increase the difficulty of implementing Internet of Things (Luthra & Mangla, 2018; Sharma et al., 2020).

Prioritisation of tasks is another challenge that will be affected by the strategy. Without

central coordination that creates a clear consensus of what tasks should be prioritised across the organisation, frequent changes in priorities will stall projects (Awan et al., 2021; Machado et al., 2019; Moeuf, et al., 2019; Schumacher et al., 2016; Sony & Naik, 2019). Patience for projects and transformations is limited and stalled or drawn-out processes can affect the commitment of people involved (Karadayi-Usta, 2020). Additionally, a challenge lies in creating clear priorities for support staff regarding which alarms and error codes to focus on (Ingemarsdotter et al., 2021).

Creating a sense of ownership and responsibility in stakeholders for the implemented

technology can be the difference between success and failure. As the return on investment for new technology take time (Wambaet al., 2017), conservative middle- and first-line managers must be supported by top management to adopt and implement new technology (Machado et al., 2019; Moeuf, et al., 2019; Sony & Naik, 2019). The ambition of the project and its delivery relies extensively on how well middle- and first-line managers embrace the technology (Bibby & Dehe, 2018; Ingemarsdotter et al., 2021; Schumacher et al., 2016), since their task is to stimulate personal development and increase participation among end-users (Agostini & Filippini, 2019).

Collaboration can be impeded by hierarchical organisational structures (Moeuf, et al., 2019;

Schumacher et al., 2016), as it can limit the transparency between different areas in the organisation, their capabilities, and current priorities (Raj et al., 2020; Machado et al., 2019; Luthra & Mangla, 2018). If different areas involved do not share priorities, conflicts may occur which can hinder collaboration (Ingemarsdotter et al., 2021). In multinational and global organisations, a lack of standardisation can further decrease collaboration (Karadayi-Usta, 2020), due to language barriers, distance or due to incompatible solutions used in different countries (Machado et al., 2019). Limited communication between people with manufacturing expertise and developers may lead to results-focused instead of needs-focused Internet of Things implementation (Ingemarsdotter et al., 2021). Collaborating with end-users across the organisation leads to mutual benefits and an increased understanding of the adoption challenges (Lu, 2017), which highlights the need to overcome hierarchical limitations and language barriers.

(22)

Collaborations with end-users are needed to find appropriate solutions that fit the needs of the organisation (Agostini & Filippini, 2019).

Technology comprehension of managers involved in the implementation of Internet of Things

affect its success (Bibby & Dehe, 2018; Schumacher et al., 2016). As Internet of Things is a new concept for many manufacturers (Boehmer et al., 2019), their comprehension of the technology is generally lacking (Ancarani et al., 2019; Ghobakhloo, 2018; Haddud et al., 2017; Machado et al., 2019) which could increase the difficulty of finding the appropriate solutions and devices that fit the needs of the organisation (Machado et al., 2019). They may lack a comprehensive picture of the technology and what change it implies to operations (Calabrese et al., 2020; Núñez-Merino et al., 2020). Many stakeholders feel uncomfortable because of their lacking experience of Internet of Things (Chiarello et al., 2018), and thus, require time for professional development.

Data interpretation, as it can be difficult to create consensus of how data should be interpreted

and how people are expected to act upon the data. Data interpretation capabilities is furthermore affected by users’ backgrounds, such as experience with related digital technologies, education, and professional experience, and are based on how data is presented (Haddud et al., 2017; Ingemarsdotter et al., 2021; Paiola & Gebauer, 2020).

A talent shift may be required in the organisation (Ghobakhloo, 2018). To construct the

applications required to collect, store, and analyse data, the organisation may need expertise within several emerging digitalisation areas (Haddud et al., 2017; Paiola & Gebauer, 2020), e.g., network infrastructure, telecommunications, data centres or Internet of Things. Thus, as organisations undergo transformations using digital technologies, they must compete for experienced individuals. It might be difficult to find the talent required to utilise Internet of Things to its full extent (Karadayi-Usta, 2020; Moeuf, et al., 2019).

Employee displacements are due to affect the organisation (Haddud et al., 2017; Raj et al.,

2020) as Internet of Things may remove the need for direct social interactions (Nižetić et al., 2020). Organisational changes to operations, such as automation of repetitive and low value-added tasks may require employees to undergo training in order for them to focus on higher value-added tasks or new functions (Calabrese et al., 2020; Moeuf, et al., 2019; Stock et al., 2018; Sung, 2018).

Security procedures are needed to keep data safe by avoiding intentional or unintentional

(23)

al., 2017). Such security measures should not be underestimated (Ancarani et al., 2019; Machado et al., 2019), and must be adhered to, to safeguard data and privacy (Kamble et al., 2020; Lu, 2017).

Incentives, as a lack of comprehension of the benefits of Internet of Things can prevent people

involved in technology implementation from committing fully to the project (Ancarani et al., 2019; Chiarini et al., 2020; Luthra & Mangla, 2018; Schumacher et al., 2016), especially when considering creating long term, lasting benefits (Sony & Naik, 2019). Lack of experience with similar projects creates further uncertainty with regards to the benefits, as well as increases the possibility of people not even considering that the technology can aid the organisation (Ingemarsdotter et al., 2021). Even if the technology has partly been implemented in the organisation, its fragmentation leads to a lack of overview of what it may give to other areas in the organisation (Raj et al., 2020) and what needs the technology can satisfy (Machado et al., 2019). This is because the benefits from Internet of Things are difficult to measure (Rauch et al., 2020). The risks involved with implementing Internet of Things may still be unclear for organisations, thereby reducing the willingness to implement it (Boehmer et al., 2019; Paiola & Gebauer, 2020). Likewise, the operational costs for the technology is unclear, such as energy needs, maintenance or similar (Sharma et al., 2020). As it can take time for digital transformations to give a return on investment (Wambaet al., 2017), return on investment examples in literature and reports are lacking (Haddud et al., 2017; Paiola & Gebauer, 2020). Therefore, creating uncertainty about the outcomes of Internet of Things (Núñez-Merino et al., 2020) which may be the reason for an unclear business use case (Dalenogare et al., 2018; Ingemarsdotter et al., 2021). Due to the lack of road maps or narratives, visualising the results and the benefits of Internet of Things is difficult (Calabrese et al., 2020; Sung, 2018). Not understanding the value of data creates additional challenges (Moeuf, et al., 2019), such as the need for an incentive system (Machado et al., 2019) to motivate reluctant individuals. Due to these issues, organisations need to create consensus across the organisation about the benefits of Internet of Things (Awan et al., 2021).

Sustainability is another organisational concern, as the use of data centres potentially increase

organisations’ emission of carbon dioxide (Stock et al., 2018). Long-term effects of Internet of Things has not been fully delineated (Nižetić et al., 2020), the recycling of electronic waste is therefore another issue as organisations must decide if this will be conducted in-house or by external solutions provider. There are currently no recycling options for Radio Frequency

(24)

Identification tags (Stock et al., 2018) which can cause problems for organisations wanting to track products and equipment.

2.2.3 Resource availability challenges

An overview of the resource availability challenges discovered in literature and further described below can be seen in Table 7.

Table 7: Overview of the resource availability challenges, in order of appearance.

Resource availability challenge

1. Energy 2. Raw material

3. Pre-existing technology 4. Financial resources

5. Expertise 6. Training

Energy needs increase with the use of Internet of Things (Sharma et al., 2020; Stock et al.,

2018), such as through sensors, antennas, expanded networks or data centres. Depending on the size of the organisation and its needs, energy consumption could increase a noticeable amount (Nižetić et al., 2020). As the world is getting closer to a shortage on energy and as data storage, processing and analytics require significant energy (Arshad et al., 2017), organisations may require expertise and use of energy harvesting methods such as vibration, heat, light and wireless (Kamalinejadet al., 2015).

Raw material needs that may affect the implementation of Internet of Things, due to global

needs for electronics the increased consumption of related raw materials (Nižetić et al., 2020) has led to a shortage of several materials required to create sensors. Various materials required to create sensors include aluminium, copper and silver used for Radio Frequency Identification and arsenic and gallium used in semiconductors, where several of these have a shortage, or are predicted to in the near future (Stock et al., 2018). A shortage of these raw materials leads to a

(25)

lack of advanced electronics components (Nižetić et al., 2020), as e.g., semiconductors are used for integrated circuits.

Pre-existing technology is a potential barrier as the lack of advanced or advancements in

electronics can hinder the implementation of Internet of Things (Nižetić et al., 2020). As previously mentioned, if access to e.g., telecommunications, data storage, data processing and analytics tools is lacking then projects may end up stalled or prolonged and end users may be unable to use the technology productively (Haddud et al., 2017; Ingemarsdotter et al., 2021).

Financial resources need to be suitable for the implementation and the following maintenance

of Internet of Things (Bag et al., 2021; Ghobakhloo, 2018; Haddud et al., 2017; Moeuf, et al., 2019; Raj et al., 2020). Insufficient initial investments can halt or prolong projects (Alcácer & Cruz-Machado, 2019; Calabrese et al., 2020), e.g., due to the lack of a comprehensive pilot study that miss interoperability issues with pre-existing technology (Sisinni et al., 2018). As new technology implementations commonly require significant investments (Kamble et al., 2020; Kohtamäki et al., 2020; Sony & Naik, 2019; Sung, 2018), a lack of investment from stakeholders can cause conflicts among potential end users (Bibby & Dehe, 2018; Haddud et al., 2017; Tortorella et al., 2020) about where to implement the technology. Moreover, whether development of Internet of Things will be done within the organisation or by an external solutions provider, it will require customisation to fit the organisation (Lu et al., 2020; Schumacher et al., 2016). Such customisation can furthermore increase maintenance cost (Machado et al., 2019). The availability of infrastructure, hardware and software furthermore affects the financial resources required for Internet of Things implementation (Luthra & Mangla, 2018).

Expertise of stakeholders affects the implementation of Internet of Things. Managers need

relevant skills and the right capabilities to integrate business processes with Internet of Things (Haddud et al., 2017; Machado et al., 2019; Sung, 2018), such as experience from similar projects (Alcácer & Cruz-Machado, 2019) and the ability to collect the needs of the organisation (Kamble et al., 2020). Without these skills, the implementation may lack credibility (Boehmer et al., 2019) and understanding of adoption challenges (Luthra & Mangla, 2018) and thus lose support from end users (Calabrese et al., 2020). Planners require some technical skills as they need an understanding of the challenges and risks with Internet of Things implementation and adoption (Olsen & Tomlin, 2020). Additionally, support staff need the skills and knowledge to maintain the implemented technology (Agostini & Filippini, 2019; Ghobakhloo, 2018; Haddud et al., 2017), such as digital and technical skills (Machado et al., 2019; Moeuf, et al., 2019; Chiarini et

(26)

al., 2020) from experience in calibration or programming (Alcácer & Cruz-Machado, 2019; Olsen & Tomlin, 2020). End users’ needs to be adaptable and have the possibility to gain knowledge of the technology beforehand to exploit the data acquired using relevant processing and analysis tools (Moeuf, et al., 2019) as skill gaps will affect the benefits gained from Internet of Things (Bag et al., 2021; Sharma et al., 2020).

Training is therefore another challenge if there is a lack of relevant skills among managers,

developers, and end users (Ingemarsdotter et al., 2021; Kamble et al., 2020). Managers may require training in digital management, such as its tools and trends, to understand the needs Internet of Things can satisfy, thus, ensuring they recognise potential end users (Agostini & Filippini, 2019; Chiarini et al., 2020; Wambaet al., 2017). Developers need the technical skill set to ensure they understand how to best create or customise infrastructure, data management systems and decision support systems (Raj et al., 2020; Wambaet al., 2017). Furthermore, end users need to be informed on how to use the technology that will be implemented (Kohtamäki et al., 2020; Schumacher et al., 2016). Training therefore needs to be planned extensively to ensure it fits the specific needs of the stakeholders (Fantini, Pinzone, & Taisch, 2020; Moeuf, et al., 2019) as a lack of an advanced education system can hinder the implementation and the adoption of Internet of Things (Karadayi-Usta, 2020).

2.3 Overcoming challenges of Internet of Things

While existing research is limited regarding change management during digitalisation using Internet of Things, several of the challenges presented in literature are to different degrees self-explanatory through an increased understanding of them. They could potentially be considered drivers, success factors or enablers in their own right, such as Infrastructure, Usability, Creating a sense of ownership and responsibility, Collaboration, Incentives, and Training. However, Goad et al. (2021), Haddud et al. (2017) and Raj et al. (2020) argued that the challenges they brought up had to be validated by practitioners.

There are studies indirectly related to the topic through their studies of similar technology integrations, such as technologies related to intelligence and flexible automation. Moeuf et al. (2019) presented success relating to implementing Industry 4.0 projects in small and medium sized enterprises. Silverio-Fernandez et al. (2019) evaluated success when implementing smart devices in the construction industry. Andersen (2018) discussed manager involvement impact on

(27)

success in organisational integration of information and communication technologies. Eaves et al. (2018) investigated knowledge sharing for successful information technology deployment.

2.3.1 Actions

A rigorous pre-study should be conducted, perhaps with the support of academics with expertise

in the area, to find and define user-needs and relevant technology (Moeuf, et al., 2019). The pre-study should allow for early identification of legacy challenges, e.g., to prevent overly complex systems, dependency on individual knowledge holders (Eaves et al., 2018) or interoperability issues that prevent scalability (Silverio-Fernandez et al., 2019). It should also include a cost-benefit analysis to ensure the benefits outweigh the costs and to further increase incentives (Silverio-Fernandez et al., 2019).

Sponsorship from top or senior managers that understand and creates a consensus across the

organisation of the benefits by showing successful, preferably local, cases (Silverio-Fernandez et al., 2019). Additionally, managers must reach consensus among themselves and estimate opportunities, communicate the objectives, and clarify them along the hierarchical line (Moeuf, et al., 2019). The organisation should move towards a cross-functional communication and collaboration by having top or senior management coordinate different business functions towards the same objective (Eaves et al., 2018).

Be attentive to cultural integration, avoid unclear terminologies and have communication

channels open to prevent a view of the new organisation as a removal of confidence, clarity or a sense of belonging (Eaves et al., 2018). Keeping negotiations open would develop technology acceptance and consensus among both managers and employees on responsibilities, training, and prioritisation of work (Andersen, 2018). Organisations may benefit from creating awareness starting with technology enthusiasts or younger generations and incrementally moving towards those who prefer traditional ways of working, thus, increasing the speed of technology awareness and adoption (Silverio-Fernandez et al., 2019). To further improve interaction between functions and levels, events should be held where everyone is welcome to present their feedback (Eaves et al., 2018). A strategy towards continuous improvement would promote employees to embrace a technology, its routines, and tools (Moeuf, et al., 2019). A move from a competitive to collaborative work environment could also prevent knowledge being withheld or hoarded

(28)

due to its personal value (Eaves et al., 2018). This is necessary, as establishing new routines that fit the new technology is essential for success (Andersen, 2018).

Employee training (Moeuf, et al., 2019), both formal and situational, with sufficient plan

where time, space and support is allocated and adapted to individual end-users work situations to aid them in understanding and utilising the technology (Andersen, 2018). This should include educating end-users about the tools they are expected to use, moreover, to ease this process, a sufficient degree of usability is required to provide a positive user experience (Silverio-Fernandez et al., 2019). The technology should be simplistic, to reduce the expertise required to integrate and use it and to promote it by quickly providing incentives (Moeuf, et al., 2019).

Due to the limited literature on change management during digitalisation using Internet of Things, exploratory data collection is required to answer RQ2 and to understand how to overcome the challenges of Internet of Things.

(29)

3 METHOD

The three phases completed to fulfil the research purpose are presented here. A summary of the chosen methods is seen in Table 8 and an overview of the process is seen in Figure 3.

Table 8: Chosen methods.

Method element Chosen method

Research type Exploratory

Research approach Abductive

Research strategy Embedded multiple case study

Data collection Literature review, interviews & survey

Sampling, survey Convenience & snowball

Sampling, interviews Systematic & convenience

Analysis, survey Mean value, maximum given value & comparative

Analysis, interviews Thematic

Figure 3: Overview of the methodology.

3.1 Research approach & strategy

This study aimed to benefit from both the qualitative and the quantitative methods and to mitigate their individual disadvantages, the study thereby adopted a mixed-methods research

(30)

approach in accordance with Johnson & Onwuegbuzie’s (2004) definition. The purpose of this study is to further the understanding of how organisations can overcome the challenges of Internet of Things. It was therefore essential to (1) evaluate the generalisability of the challenges found in literature using quantitative deductive methods and (2) gain a deeper insight into how organisations can overcome the challenges of Internet of Things using qualitative inductive methods. This means that the qualitative data collected has the purpose of explaining the quantitative data. The study therefore used abductive reasoning as well as systematic combining as described by Dubois & Gadde (2002) which is an iterative process with the aim of remaining agile regarding the collection of data and the analysis of gathered data. This iterative process was well suited considering the mixed-methods approach and explanatory nature of this study. This iterative process meant continuously reviewing the initial research questions, the theoretical framework, and the data collection.

To answer the research questions, an embedded multiple case study was conducted at a global tier 1 aerospace supplier with over 15 000 employees. The case study was chosen because of the insight that studying Internet of Things in its real-life setting would bring. Because of the complex manufacturing setting, the traditional culture, and the global business of the organisation, it has just begun its work with global digitalisation strategies. The organisation is mainly focused on using well-proven digitalisation technologies due to quality requirements, costs related to production downtime and the need to protecting sensitive data and confidentiality. It has however gone through many local Internet of Things implementations at its manufacturing sites across the globe, including all related technologies presented in Figure 2 to varying degrees. Because of its many Internet of Things projects and its endeavours to digitalise, it was viewed as a great source for data collection and its possibilities to benefit from this study were regarded as high. The multiple case study was seen as beneficial since it allowed for perspectives from manufacturing sites with different level of process and digitalisation maturity, thereby giving insights into thought processes of individuals in different stages of Internet of Things projects at the case organisation. To gain a wider breadth of perspectives, this study aimed to include respondents from all levels within the organisation, whether at a strategic, management or operational level.

3.2 Data collection

The study began in January 2021 by collecting data in three phases from project areas seen in Figure 4, ending in May 2021.

(31)

Figure 4: Projects gathered data from.

The organisation appointed a supervisor for this study that aided in the data collection process. The first phase involved exploratory unstructured interviews, the second phase, exploratory structured interviews, and an exploratory survey and the third phase, validatory semi-structured interviews. Both primary and secondary data was collected and analysed. Primary data was collected through interviews and a survey. Secondary data was collected from internal documents regarding Internet of Things implementation projects and external documents concerning industry-specific Internet of things implementation issues. A total of 19 interviews were held of which 7 were unstructured ranging from 30 to 75 minutes and 12 were structured ranging from 45 to 75 minutes. 76802 words were transcribed from the semi-structured interviews. 23 internal and 5 external documents were reviewed. A major part of the study was conducted at the case company’s office in Sweden, however, due to the ongoing pandemic and the various geographic locations of interviewees, most interviews were conducted through video conferencing software.

3.2.1 Phase 1: Exploratory unstructured interviews

As a big part of the study was conducted at the case company’s office, many formal as well as informal meetings took place. Meetings such as technology demonstrations, weekly project reviews and department reviews occurred on a regular basis. A lot of time went into networking within the case organisation to map previous projects related to Internet of Things, with the aim of locating potential interviewees that held essential knowledge of the topic. Due to the confidentiality requirements of the Aerospace industry, this took considerable time. As a result, seven unstructured interviews of exploratory character were held with individuals experienced in Internet of Things projects and related technologies, see Table 9.

(32)

Table 9: Respondents in unstructured interviews in phase one.

No. Respondent code Position/Expertise Country Type Duration (minutes)

1 R01 Industry 4.0 Coordinator Sweden Skype 60

2 R02 Engineering Manager Sweden Skype 60

3 R03 Enterprise & Data Architect United States Starleaf 75

4 R04 Digital & Industry 4.0 Architect United Kingdom Starleaf 30

5 R05 IT Integrations Specialist United Kingdom Starleaf 30

6 R06 Methodology specialist, Quality Sweden Face-to-face 45

7 R07 Manufacturing Operations Management Sweden Face-to-face 60

These interviews had the purpose to outline the data collection for this study and to help map Internet of Things projects within the case organisation. The mapping was needed to understand the potential of different samples and key individuals at the case organisation that could confirm the validity of the findings of this study. The respondents of the unstructured interviews were chosen based on the supervisor’s recommendation and subsequent snowball sampling.

In connection to these interviews, it was common that documents regarding previous and current digitalisation projects were shared which increased the relevance of the interview and the understanding of the held discussions. Information was also shared regarding e.g., the Aerospace industry, consulting firms’ reports, data management, industrial internet, enterprise architecture and standards. For the first phase, Sharepoint was used for knowledge distribution and the contents of articles and documents were organised in Excel. The limitations to the data collection during this phase were due to the understanding of technical terms which to some extent affected the transparency of data in documents and which in some interviews increased the duration between e.g., learning common abbreviations and having insightful discussions.

(33)

3.2.2 Phase 2: Exploratory semi-structured interviews

The interviews in the second phase had the aim of understanding digital change management, such as the rate of success of previous digitalisation projects, the challenges encountered and how they were be overcome. This was done as there is a lack of roadmaps and guidelines in the existing literature regarding digital transformations of organisations, specifically regarding Internet of Things. Semi-structured interviews were decided to be the most fitting for this purpose, as to ensure the questions were adapted to the interviewees’ experiences and to create an in-depth discussion.

The interviewees and the content of the interviews were chosen based on the interviews conducted in the first phase, their recommendations, and internal documents, i.e., snowball sampling and convenience sampling. Twelve interviews were conducted with thirteen unique interviewees that have experience from working in local, multinational, and global projects, see Table 10.

Table 10: Respondents in semi-structured interviews in phase two.

No. Respondent

code Position/Expertise Country Type

Duration (minutes)

Transcribed words

1 R06 Methodology specialist, Quality Sweden Skype 45 3744

2 R08 Project Manager, Product Lifecycle Management Sweden

Face-to-face 70 7985

3 R02 Engineering Manager Sweden Skype 75 6146

4 R01 Industry 4.0 Coordinator Sweden Skype 40 3366

5 R09 Product Lifecycle Management, previously Head

of Department of Manufacturing Engineering Sweden Skype 75 9804

6 R10 Head of Department, Manufacturing Engineering Sweden Skype 75 7245

7 R11 Manufacturing Engineer, Autonomous Guided

(34)

8 R05 IT Integrations Specialist United

Kingdom Starleaf 60 5035

9 R03 Business Architect, previously restructuring United

States Starleaf 55 8254

10

R12 IT Architect Sweden Skype 60

6525

R13 Applications Architect Sweden Skype 60

11 R14 Project Manager Norway Skype 70 6430

12 R04 Industry 4.0 & Digital Architect, Business Excellence

United

Kingdom Starleaf 70 8218

While the interviews were adapted with regard to the interviewees experience with change management in previous digitalisation projects and their expertise, they were sent a general outline of the interview beforehand. Thus, ensuring transparency and increasing interviewees’ possibility to prepare before the interview. While preparation may reduce spontaneity, follow-up questions asked during the interview, as aligned with the semi-structured approach, were aimed at mitigating this risk. They were also informed beforehand that their responses would be anonymised, that the recordings would exclusively be used for this study and that they would be able to withdraw their answers, up until the point the study is published. This was communicated with the aim of putting the interviewees at ease, creating a more relaxed discussion during the interview. The outline was designed to explore (1) their current role and experience with digitalisation projects, (2) challenges encountered during change, (3) how challenges were overcome, i.e., change management. The interviews were conducted as an iterative process. The interviews were recorded and follow up questions were asked to clarify uncertainties. The interviews were later transcribed. After every interview, they were asked if they had anything to add or whether any question should be added to the next interview, hence, the outline was updated after every interview that recommended an additional question or reformulation.

(35)

3.2.3 Phase 2: Exploratory survey

The survey was conducted concurrently with the interviews in phase two. It was created and shared with the goal of validating the challenges found in literature together with the ones brought up during interviews in phase one. The challenges were evaluated according to a 7-point Likert scale, seen in Table 11.

Table 11: 7-point Likert scale used to evaluate the challenges.

1 2 3 4 5 6 7 Strongly disagree Disagree Somewhat disagree Neither agree

nor disagree Somewhat agree Agree Strongly agree The survey is presented in Appendix II. The survey was also meant to provide more extensive insight into how Internet of Things is comprehended within the case organisation. It also asked whether they were available for follow-up interviews. The survey was constructed so that each respondent remained anonymous, and if they signed up for the upcoming interviews, their previous contribution was kept confidential and could not be traced to them. The respondents’ names were not used, instead each contribution was assigned a code to keep them anonymous during data processing and analysis. Respondents were chosen from different levels within the organisation, to ensure that every necessary perspective was considered, from those who see a need, to those who implement technology and finally those who will use it. A systematic sampling method was used to reach those who manage projects of related technologies, those who collect data from the technologies or those who use it on an operational level. As the case organisation is a global actor, convenience sampling was also used to ensure that no language barrier skewed the results from the survey.

3.2.4 Phase 3: Validatory semi-structured interviews

The semi-structured interviews in the third phase were conducted to verify the data received from previous interviews. Furthermore, they had the aim of validating how the most significant or most common challenges, derived from literature, and evaluated by the survey in phase two, could be overcome.

(36)

3.3 Data analysis

The analysis was conducted with regard to the three phases of data collection.

3.3.1 Analysis of unstructured interviews in phase one

Content analysis was used to approach the information received from interviews, documents, meetings, and technology demonstrations in the first phase. The unstructured interviews aided in the comprehension of Internet of Things, the organisation’s and individual’s perspectives of it and their view on the implementation challenges. The interviews gave insight into local as well as global workplace practices that helped in future networking within the case organisation, thus increased awareness of this study and availability of respondents. Moreover, insight into the case organisation’s previous digitalisation projects and the extent of Internet of Things implementations proved valuable in understanding who to interview in phase two. It furthermore aided in the creation of the outline for those interviews.

3.3.2 Analysis of semi-structured interviews in phase two

The interviews in phase two were analysed thematically to understand how the challenges for novel digitalisation technologies and specifically Internet of Things could be overcome. Thematic analysis is fitting as it is a flexible and easy to learn method that can be used to simplify large data such as that from qualitative data collection (Braun & Clarke, 2006). It is therefore fitting to fulfil explorative research questions like that of RQ2.

The theoretical foundation of this study created a basis for the thematic analysis by increasing the understanding of the challenges that must be overcome. It therefore eased the investigation of how the challenges could be overcome.

The content of the interviews was analysed in the language the interviews were held in, to decrease the risk of misunderstandings. Transcribed interviews were read, and essential quotes

(37)

were highlighted and coded. Themes were created by grouping similar codes in an excel spreadsheet, see Table 12.

Table 12: Thematic analysis example.

Quote Code Respondent Sub-theme Theme

“Working to overcome the challenges and problems they face today. Develop based on the problems you are trying to solve.”

Develop the technology with

regards to problems solved R04 Problem focused solutions

Solution “People think it's fun with e.g., screens,

where you get to see a modern

workplace.” Modern design R12 & R13 Modern workplace

Sub-themes that were not mutually exclusive were merged and were later validated against the collected data by ensuring they adhered to the previously created codes. Thereafter, the relevance of the themes was evaluated on the number of codes they were based on as well as their motivations. They were furthermore evaluated based on their contributions to fulfilling the RQ2, thus, clarifying if more interviews or another method of data collection was required. 183 relevant quotes were identified as relevant to overcoming challenges and were grouped to create 51 sub-themes and 12 themes, these are presented in Table 31. These sub-themes and themes helped identify 85 quotes which each represented a unique task to overcome one or more challenge(s) to Internet of Things.

3.3.3 Analysis of survey in phase two

The data collected from the survey in phase two provided a foundation to how high-technology manufacturers can prioritise challenges presented in literature. Challenges were rated with regard to the others in their group as derived from the theoretical foundation, i.e., technical, organisational or resource availability aspects. The mean values and maximum given values were presented in radar charts as they were regarded suitable due to the quick overview and comparison possibilities. Furthermore, radar charts were suitable due to the length of challenges’ names, as it removed the need to code or create abbreviations for the challenges, thereby increasing readability. The mean values of the challenges were compared to the 7-point Likert Scale in Table 11 to rate the challenges’ relevance. Additionally, the maximum given values were taken into consideration to understand the relationship between the challenges’ frequency and significance, to estimate the consensus of respondents. A low mean value and a high maximum given value can show that a challenge is uncommon or avoidable, but significant if encountered.

How to overcome the challenges of Internet of Things to ensure successful technology integration: A case study at an Aerospace manufacturer