Counter the Counterfeiters Examining Blockchain’s Suitability in Industrial Supply Chains Master’s Thesis in the Master’s Programme Management and Economics of Innovation MARTIN BJÖNTEGAARD LOVISA HOLMGREN Department of Technology Management and Economics Division of Entrepreneurship and Strategy CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2019 Report No. E 2019:083 Master’s thesis E2019:083 Counter the Counterfeiters Examining Blockchain’s Suitability in Industrial Supply Chains Martin Bjöntegaard Lovisa Holmgren DF Department of Technology Management and Economics Division of Entrepreneurship and Strategy Chalmers University of Technology Gothenburg, Sweden 2019 Counter the Counterfeiters Examining Blockchain’s Suitability in Industrial Supply Chains Martin Bjöntegaard, Lovisa Holmgren © MARTIN BJÖNTEGAARD, LOVISA HOLMGREN 2019. Supervisor: Charlotta Kronblad, Division of Entrepreneurship and Strategy Examiner: Joakim Björkdahl, Department of Technology Management and Eco- nomics Master’s Thesis E2019:083 Department of Technology Management and Economics Division of Entrepreneurship and Strategy Chalmers University of Technology SE-412 96 Gothenburg Telephone +46 31 772 1000 iii Abstract The problem with counterfeiting of physical products is increasing worldwide and affects global manufacturing companies’ supply chains to a large extent. Blockchain is a relatively new technology that is involved in many supply chains projects today and could potentially help to mitigate this problem. The purpose of this thesis is to examine the applicability of blockchain technology in industrial supply chains to assess its suitability to prevent counterfeiting of phys- ical products. The authors will also provide SKF with further implications and recommendations based on the suitability. The study was performed through the use of a qualitative approach, with an abduc- tive process that allowed the authors to iterate between theory and social observa- tions. First, a theoretical framework was created to enable a good understanding of the blockchain technology and global supply chains in general. Second, interviews, a case study of SKF’s extensive supply chain together with secondary data constituted the data collection to enable the authors to answer the underlying research ques- tions. Finally, the collected data was compared with the theoretical framework and formed a basis for the analysis and final conclusion to answer the general research question. The research reveals that there exist no universal definition of blockchain but the benefits of using the technology in a supply chain could be many. Transparency, de- centralized power, immutability and security just to mention a few. However, there are also several challenges connected with this that needs to be taken into account. A large amount of actors are hard to coordinate, high variation of digitalization between the actors and the difficulties to tag the unique product could be seen as the main challenges. At first, blockchain seems suitable to implement in a supply chain due to its many benefits, but soon the challenges outweigh these benefits leading to the conclusion that it is not suitable for the intended purpose. iv Acknowledgements First of all, we would like to thank our supervisor at Chalmers, Charlotta Kronblad, for helping us along the way with constructive feedback and recommendations when needed. A sincere thank you is also addressed to SKF and especially the Group Brand Protection Unit for making this thesis possible and giving us the opportunity to work alongside them every day. This includes thanking our supervisor at SKF, Petter Rönnborg, and the unit’s director, Johan Bravert. Furthermore, we are very grateful for the opportunity to interview all the blockchain experts, SKF employees and industry experts and receiving their valuable knowl- edge, which contributed a great deal to the result of this study. We also thank everyone who helped us with this thesis in general through feedback and insights. Finally, a generous thank you to Chalmers and our fellow students for helping us through these five years, it has been marvelous. Lovisa Holmgren & Martin Bjöntegaard Gothenburg, May 2019 v Contents 1 Introduction 1 1.1 What Is Counterfeiting? . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Counter the Counterfeiters . . . . . . . . . . . . . . . . . . . . . . . . 2 1.3 Blockchain and Counterfeiting . . . . . . . . . . . . . . . . . . . . . . 3 1.4 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.5 Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.6 Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.7 Delimitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2 Method 7 2.1 Research Strategy and Design . . . . . . . . . . . . . . . . . . . . . . 7 2.1.1 Case Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2.2 Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.2.1 Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.2.2 Secondary Data . . . . . . . . . . . . . . . . . . . . . . . . . . 11 2.3 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 2.4 Research Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.4.1 Reliability and Validity . . . . . . . . . . . . . . . . . . . . . . 13 2.4.2 Ethics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 3 Theoretical framework 15 3.1 Introduction to the Theoretical Framework . . . . . . . . . . . . . . . 15 3.2 Blockchain Technology . . . . . . . . . . . . . . . . . . . . . . . . . . 16 3.2.1 Describing Blockchain as Applied in Bitcoin . . . . . . . . . . 16 3.2.1.1 Network . . . . . . . . . . . . . . . . . . . . . . . . . 17 3.2.1.2 Transactions . . . . . . . . . . . . . . . . . . . . . . 18 3.2.1.3 Blocks . . . . . . . . . . . . . . . . . . . . . . . . . . 20 3.2.1.4 Proof-of-Work . . . . . . . . . . . . . . . . . . . . . . 20 3.2.1.5 Incentives in a Decentralized System . . . . . . . . . 21 3.2.2 Private and Public Blockchains . . . . . . . . . . . . . . . . . 22 3.2.3 Smart Contracts . . . . . . . . . . . . . . . . . . . . . . . . . 23 3.3 Techniques to Create Unique Identities . . . . . . . . . . . . . . . . . 23 3.4 Global Supply Chains . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 3.5 Digitalization of Supply Chains . . . . . . . . . . . . . . . . . . . . . 26 3.5.1 What Is Digitalization? . . . . . . . . . . . . . . . . . . . . . . 26 vi Contents 3.5.2 How Digitalization Has Changed Supply Chains . . . . . . . . 27 3.5.3 Demands, Benefits and Risks of Digital Supply Chains . . . . 28 3.5.4 Possible Role for Blockchain in Digital Supply Chains . . . . . 29 3.6 Chain of Custody . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 4 Empirical Data and Findings 32 4.1 Expert Interview Data . . . . . . . . . . . . . . . . . . . . . . . . . . 32 4.1.1 What Is Blockchain and What Are the Applications Used To- day? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 4.1.2 Can a Physical Product and the Digital Blockchain Be Irre- vocably Linked Together? . . . . . . . . . . . . . . . . . . . . 34 4.1.3 What Can Be the Main Benefits of Applying Blockchain in a Supply Chain? . . . . . . . . . . . . . . . . . . . . . . . . . . 36 4.1.4 What Can Be the Main Challenges of Applying Blockchain in a Supply Chain? . . . . . . . . . . . . . . . . . 37 4.1.5 What Processes in a Supply Chain Are Suitable for Blockchain Application? . . . . . . . . . . . . . . . . . . . . . 39 4.1.6 How Could the Adoption of Blockchain Develop in the Future? 40 4.2 Examples of Blockchain Projects . . . . . . . . . . . . . . . . . . . . . 41 4.2.1 Blockchain in Digitization Projects . . . . . . . . . . . . . . . 41 4.2.2 Blockchain in Traceability Projects . . . . . . . . . . . . . . . 42 4.3 SKF - A Case Study . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 4.3.1 SKF’s Supply Chain and Sales Channels . . . . . . . . . . . . 44 4.3.2 The Problem of Counterfeiting for SKF . . . . . . . . . . . . . 46 4.3.3 SKF Brand Protection Activities . . . . . . . . . . . . . . . . 47 5 Analysis and Discussion 49 5.1 A Summary of SKF’s Counterfeiting Problem . . . . . . . . . . . . . 49 5.2 Potential Benefits of Blockchain and How It Could Prevent Counter- feiting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 5.3 Blockchain Design in a Supply Chain Context . . . . . . . . . . . . . 51 5.4 Difficulties with Using Blockchain to Prevent Counterfeiting . . . . . 54 5.4.1 Establishing Unique Identities . . . . . . . . . . . . . . . . . . 54 5.4.2 Entering and Storing Information on a Blockchain . . . . . . . 56 5.4.3 Blockchain and Supply Chain Size . . . . . . . . . . . . . . . . 56 6 Conclusion 59 6.1 Conclusion of Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 6.2 Future Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 6.3 Practical Implications for SKF . . . . . . . . . . . . . . . . . . . . . . 61 References 63 vii List of Figures 2.1 Abductive research with iteration between literature and data . . . . . 8 2.2 The analytical framework . . . . . . . . . . . . . . . . . . . . . . . . . 13 3.1 A conceptual model of the theoretical framework . . . . . . . . . . . . 16 3.2 The Bitcoin network and the transmission of a transaction . . . . . . 18 3.3 The functionality of private and public key cryptography . . . . . . . . 19 3.4 An illustration of a blockchain and how blocks are linked together . . . 20 3.5 Digitization and Digitalization . . . . . . . . . . . . . . . . . . . . . . 26 3.6 Demands on Supply Chain 4.0 . . . . . . . . . . . . . . . . . . . . . . 29 4.1 SKF’s Downstream Supply Chain . . . . . . . . . . . . . . . . . . . . 45 4.2 The problem of counterfeiting exists mainly on the aftermarket . . . . 47 viii Terminology Bitcoin - A digital cryptocurrency that was launched in 2009 and that is based on blockchain technology. CPU - Central Processing Unit. Electronic circuit in a computer that carries out the instructions of a computer program. Double-spending - When a digital currency is spent twice. EDI - Electronic Data Interchange. EDI means that there is an integration be- tween business systems of the seller and buyer, where orders can be automatically generated without any human intervention. GBP - Group Brand Protection. The unit responsible for SKF’s trademarks, and that fights counterfeiting of SKF bearings. ICT- Information and Communications Technology. OEM - Original Equipment Manufacturer. For example a car manufacturing com- pany. SKU - Stock Keeping Unit. A unique code that is given to every individual product item that a manufacturer produce, in order to be able to identify that specific product. ix 1 Introduction This chapter gives an introduction to this thesis and the problems it aims to tackle. The problem of counterfeiting is presented from a holistic perspective and then some existing ways to work against it are briefly discussed. The potential for blockchain technology to make it harder for counterfeiters is then broadly introduced. Thereafter, the background to the thesis is presented, before defining the purpose and research questions along with delimitations of the study. 1.1 What Is Counterfeiting? Counterfeiting is one of the problems connected to global trade and globalization, and the World Trade Organization (WTO) define it as "unauthorized representation of a registered trademark carried on goods identical or similar to goods for which the trademark is registered, with a view to deceiving the purchaser into believing that he/she is buying the original goods" (WTO, 2019). According to the Organiza- tion for Economic Co-Operation and Development (OECD) (2007), counterfeiting is growing in scope, scale and threat. This since fake products are being produced and consumed in virtually all economies, with Asia emerging as the single largest producing region. The types of products being counterfeit has, during the recent years, expanded from luxury items to products that have an impact on personal health and safety. OECD (2007) further argue that counterfeit products are of concern to governments, businesses and consumers. To governments they are of concern because of the negative impact they might have on innovation, the threat they pose to the welfare of consumers and the substantial resources that they chan- nel to criminal networks, organized crime and other groups that disrupt and corrupt society. To businesses, counterfeit products are of concern because of the impact they might have on sales and licensing, brand value and firm reputation as well as the ability of firms to benefit from the breakthroughs they make in developing new products. Since substandard counterfeit and pirated products could pose vital safety and health risks, they are of concern to those who consume them as well. Berman (2008) classifies counterfeit products into four different types: (1) knockoffs 1 1. Introduction - a duplicate of the original that bear a different name and customers are aware that they are purchasing an inexpensive copy; (2) true counterfeit products - products that are very similar to the original and use the same brand name; (3) products manufactured by an outsourced supplier by using a "third shift" that the original manufacturer is unaware of and (4) products manufactured by an outsourced sup- plier that do not meet the required standard but are not properly labeled as defect or destroyed. Thus, there are two different types of counterfeiters; the ones who sell the product for a similar price as for the original product and the one who sell the product for a significant lower price. That leads to two different types of consumers of counterfeit products; the ones who think they are buying the original product and the ones that know they are buying a counterfeit. According to a study conducted by OECD in collaboration with the European Union Intellectual Property Office (EUIPO) (2016), the international trade of counterfeit products in 2013 corresponded to 2,5% of world trade, which is equivalent to USD 461 billion. In the European Union, these products amounted up to 5,1% of imports the same year, comparable to USD 116 billion. A previous study made by the same organizations in 2008 estimates that the market for these products was 1,9% of world imports, implying that the threat of counterfeiting has increased during the past ten years. These numbers also imply that the impact of counterfeit products is twice as high for countries within the EU than it is for the world as a whole. The increasing use of e-commerce provides the counterfeiters with a platform to cost efficiently capture a large number of potential customers (OECD/EUIPO, 2016). E- commerce enables counterfeiters to get access to areas that were previously beyond their scope. Counterfeiters are also able to avoid being caught by functioning across multiple jurisdictions as well as closing down and setting up websites overnight without decreasing their customer base. Counterfeiting is therefore a profitable market since by imitating a product there are no costs associated with research and development, advertising, quality control standards or regulated labor (Harvey, 1987). Hence, counterfeiters have none of the traditional costs related to introducing or selling a specific product which enables them to sell the product at a lower price while still generating profit. Thereof, it is important for companies producing the original product to fight counterfeiting and piracy to both maintain and regain sales but also to protect their brand. 1.2 Counter the Counterfeiters Since a company’s brand is one of the most valuable intangible assets they posses and brand success raise counterfeiters, it is crucial to protect it as much as possible (Green and Smith, 2002). WTO developed the Agreement on Trade-Related As- pects of Intellectual Property Rights (TRIPS) as an effort to control counterfeiting. TRIPS is an international legal agreement, that all the member nations of WTO has to be a part of, that sets minimum standards of regulations by national governments regarding multiple forms of intellectual property. This means that individual gov- 2 1. Introduction ernments can put pressure on countries that lacks laws and rules about intellectual property, and try to influence their activities. Other courses of action to address counterfeiting are to educate stakeholders at the source by convincing governments that intellectual property protection is in their best long-term interest, and use advertising to inform customers about coun- terfeiting to eliminate the market for such products (Shultz and Saporito, 1996). Companies exposed to counterfeiting can also use different techniques to tag their products in way that is difficult to duplicate, which is further described in section 3.3 Techniques to Create Unique Identities. They can also create coalitions with other industry members to battle the counterfeiting together. Further, examples of raids of manufacturers that sell and produce counterfeit products have shown to be an effective way to reduce counterfeiting and regain sales (Green and Smith, 2002). 1.3 Blockchain and Counterfeiting In later years, blockchain technology, further referred to as simply blockchain, have rapidly gained a strong interest in society. During the end of 2017, a large surge was created around the technology and businesses in several industries tried to jump on the bandwagon and drive business with what they claimed to be blockchain-backed products and services. Web searches on the term “Blockchain” exploded and reached its peak a few days before Christmas that year (Google, 2019). The sudden increase in interest largely affected the world’s biggest cryptocurrency Bitcoin (which is based on blockchain), that reached its all-time high closing price of $19 345 on the 16th of December 2017 (Yahoo, 2019). The price then plummeted almost as fast as it had reached its peak during the beginning of 2018, before stabilizing a little bit and continue in a modest decrease. Today, May 17th 2019, the value for one Bitcoin is approximately $7 200. Bitcoin’s price development seem to closely correlate with the hype around blockchain, and web searches for “Blockchain” today is around 28% of the amount they were at the height of the peak (Google, 2019). Before, during and after the blockchain and Bitcoin hype, there have been many examples of projects that try to use blockchain for other applications than for a cryptocurrency. Examples include IBM and Maersk’s joint venture TradeLens from 2018, a shipping digitization project to improve visibility and decrease costs, and the company Everledger’s Ecosystems of trust, a transparent platform in the diamond industry to increase trust in the value chain. According to Chang, Iakovou, and Shi (2019), the application of blockchain in supply chain related projects are expected to have a compounded annual growth rate of 87% from 2018 to 2023. Blockchain will there provide better traceability, automate processes and secure chain-of-custody, among other things. Given these and other blockchain capabilities, hope exists that blockchain can be an efficient tool in the fight against counterfeiters. Several question marks does however still exist. Is it possible to implement a blockchain solution in the network of actors that constitutes a supply chain? What potential 3 1. Introduction benefits exists and do they outweigh the costs of the implementation itself? What challenges exists when trying to implement blockchain in such a network? This study therefore intended to explore the area of blockchain application in supply chain, to better understand its suitability. 1.4 Background The initiative to this master thesis was taken by SKF AB, from here on only SKF. Today, SKF is a global company that manufactures bearings to be used in all dif- ferent kinds of machines that covers most industries in the world, where mechanical tools and machines are required for operating (SKF, 2019). Currently, the company is present in approximately 130 countries and have about 45 000 employees world- wide. They have production factories in 24 countries around the world, and have around 400 000 active SKUs in their assortment. SKF’s bearings are renown for their superior quality, which is why their customers trust them to be the supplier of bearings to machines which operate in the most exposed and extreme conditions. It also enables SKF to charge a premium price for their products as compared to their competitors. In many parts of the world, SKF have a problem with counterfeiting as other man- ufacturers produce bearings and then brand them as genuine SKF products. These counterfeits do not have the same quality as the SKF products, but can for an untrained eye look very similar. This leads to that some customers purchase fake products which can potentially lead to severe consequences. To counter the counter- feiters, SKF have a department called Group Brand Protection (GBP) that solely work with activities to counteract fraudulent bearings and the people creating and selling them. They are responsible for managing and protecting SKF’s trademarks. The group’s objective is to ensure that SKF’s customers receive genuine products and are not cheated by counterfeiters, something that could lead to severe implica- tions. They mainly want to achieve that by increasing awareness in the market, but also by making it more costly to produce and sell fake products, thus minimizing the amount of fake products being sold. Today, some of the progress is measured in estimations of sales recovery, i.e. increase in sales due to brand protection ac- tivities, which is a clear indicator for the effect of their work. The unit consists of 13 people worldwide, of which most are based in Gothenburg. Compared to other similar companies, this is a relatively large department to tackle issues of coun- terfeiting and trademark infringements. They also have a different approach than many, where they try to win back business rather the punish the counterfeiters. When GBP started their operations in 2009, it was easier to verify that a product was fake based on a few parameters. In later years, counterfeiters have however become better at copying the genuine products, making it harder for the group to determine if they are fake or not. Given this development, GBP are constantly trying to find new ways to make verification easier and more secure, and to make copying 4 1. Introduction harder. They have heard about blockchain projects from contacts in the market, and therefore wanted to investigate the possibility to use blockchain for product authentication and secure traceability in their supply chain. If this thesis would present potential to use blockchain in the authentication of products in supply chain networks, SKF might pursuit investing in the technology in order to establish a more reliable distribution and recover sales that were previously held by counterfeiters. A full description of SKF’s supply chain together with a deeper description of the counterfeiting problem and GBP’s operations is found in section 4.3 SKF - A Case Study. 1.5 Purpose The purpose of this study is to examine the applicability of blockchain technology in industrial supply chains to assess its suitability to prevent counterfeiting of physical products. 1.6 Research Questions To serve the purpose of the study, the following general research question and cor- responding sub-questions have been formulated: Is blockchain technology suitable to use in an extensive industrial supply chain in order to prevent counterfeiting of physical products? (a) What is blockchain technology and what are the applications used today? (b) Can a physical product and the digital blockchain be irrevocably linked together? (c) What can be the main benefits of applying blockchain technology in a supply chain? (d) What can be the main challenges of applying blockchain technology in a supply chain? (e) What processes in a supply chain are suitable for blockchain application? (f) How could the adoption of blockchain technology develop in the future? 5 1. Introduction 1.7 Delimitations The study will exclusively research blockchain’s possible applicability for tracing physical products, hence intangible services will not be examined further. Regarding the different classifications of counterfeit products, the study will solely focus on true counterfeit products, i.e. products that are very similar to the original and use the same brand name. Furthermore, no alternative methods other than the use of blockchain to avoid counterfeiting will be investigated during the study, however some will be presented to describe the current situation. The costs associated with further investigations and implementation of blockchain will not be examined in detail. In the case study on SKF, only the downstream supply chain from SKF to end customers will be considered. 6 2 Method This chapter aims to thoroughly describe the method used to conduct this study to give the reader a deep understanding of how the authors proceeded to find answers to the existing research questions previously presented. 2.1 Research Strategy and Design Bryman and Bell (2015) distinguish between two different research strategies that a study can employ. These are quantitative and qualitative research, and differs in the way they aspire to conduct research. While the first distinction on these two is made on the collection and analysis of data, where the quantitative approach em- phasize quantification and the qualitative emphasize words, there is more separating the approaches. On a general level, qualitative research is often associated with the development of new theories, rather than testing existing theories which is more associated with the quantitative approach. For the purpose of this study, extensive theory does not exist on blockchain application in supply chains to prevent counter- feiting, making a quantitative study hard to conduct. To address this problem, the study took an exploratory stance, making the qualitative strategy with interviews and descriptive data most suitable. There are three main perspectives of looking at the role of theory in research (Bry- man and Bell, 2015). These different views are deductive, inductive and abductive, and they lead to different processes of working in a research project. In the deductive approach, the researcher starts with theory and the knowledge of a specific domain, to form a hypothesis that puts the theory to the test and is either confirmed or rejected. However, in the inductive approach, the result of the research is theory. Based on observations and findings, inductive research often aim to draw general conclusions that builds new theory. Deduction is often associated with quantita- tive research, whereas induction is associated with qualitative research. A more pragmatic view is the abductive approach, which can be seen as a combination of the previous ones. Abduction begins with a puzzling phenomenon that cannot be fully explained by existing theory. Through back-and-forth iteration between the 7 2. Method social observations and literature, abductive reasoning seeks to find the conditions for which the phenomenon can be explained. This approach was applied in the study, as the capability of blockchain application to prevent counterfeiting could be seen as a puzzling phenomenon. It was necessary to iterate between literature and collection and analysis of data (Figure 2.1), as neither could help to explain the research questions solely. Figure 2.1: Abductive research with iteration between literature and data When conducting a research study, it is important to decide how to collect and an- alyze data. According to Bryman and Bell (2015), this procedure is called research design and there exists five different types: experimental, cross-sectional, longitudi- nal, case study and comparative. Case study was deemed most appropriate since it entails the detailed and intensive analysis of a single case. Comparisons with other blockchain projects were also made to get a better understanding of the blockchain technology and its application areas. A further motivation for using a case study is explained below. 2.1.1 Case Study It is important for empirical research to have a strong foundation in related litera- ture, in order to identify a gap in the existing literature that the study intends to fill (Eisenhardt and Graebner, 2007). However, if the proposed research strategy is to use cases to build theory that could help to answer the existing research questions, such an approach must be well motivated. For the purpose of this study, it was important to get a good understanding of a global supply chain to evaluate what processes are ready for blockchain implementation along with difficulties in doing so. Given that SKF has such an extensive supply chain, and that their products are being counterfeit, they seemed suitable for a case study. Building theory from case studies is an approach that means to use one or several 8 2. Method cases describing a certain phenomenon in order to create theoretical constructs or propositions based on empirical evidence (Eisenhardt and Graebner, 2007). A case study was deemed appropriate as a part of the study to use as a comparison with other cases and theory in order to draw conclusions about blockchain’s suitability in supply chains. To be able to build new theory about the phenomenon, the sampling of studies should not be random like many may argue (Eisenhardt and Graebner, 2007). Instead, SKF was chosen specifically because it present a case that offer insight to the studied context. The data from case studies can be rich in variety, ranging from interviews and obser- vations to archival and survey data (Eisenhardt and Graebner, 2007). Interviews are especially good to gather rich data, but at the same time present a risk of bias among the respondents (Eisenhardt and Graebner, 2007). Aside from interviews with em- ployees, work on the project was mainly conducted on SKF premises. This allowed the authors to observe how GBP conducted the daily work and better understand the context of counterfeiting in the SKF supply chain, as well as an understanding of the supply chain itself. Observational notes were here made if relevant to the case, to be used as data in the study. The case is presented in full under 4.3 SKF - A Case Study. 2.2 Data Collection The data collection was constituted by two different methods, namely conducting interviews and by the search and compilation of secondary data. The data was gathered as a basis for analysis that intended to bring clarity to the purpose of the study. 2.2.1 Interviews The interviews held during the course of the study were conducted in a semi- structured way, as this format was deemed most suitable to collect rich, qualitative data. A semi-structured interview is a flexible process where researchers start with a quite specific set of areas to be covered but allow for wide variations in the re- spondent’s answers (Bryman and Bell, 2015). Questions are prepared beforehand but might not be asked in the predetermined order, and deviations due to follow- up questions are allowed. Some main things that are important to consider when preparing a semi-structured interview guide are to create a certain level of order on the question topics, formulate questions so that they help to answer the research question(s) of the study and to not ask leading questions. This is the structure that was followed during the interviews. Before interviews, interview guides were prepared, containing questions that were developed based on the impression of what the person could give insight to and contribute with. Since it was not always possi- 9 2. Method ble to determine this beforehand, the interviews had to be semi-structured to allow for collecting deviating answers that were richer than anticipated, or gave insights in different areas than expected. When sampling informants, there are many different methods to employ. In qual- itative research, snowball sampling is the most used method and occurs when the informants provide the researchers with contact information to other informants (Noy, 2008). The process is as follows: the informant refers to other informants, who are then contacted by the researchers and then in turn refer to additional informants et cetera. For this study, a few persons with great knowledge about blockchain was contacted who then further referred to other people with knowledge about the technology within their networks. According to Noy (2008), this is an effective tool to gather information and knowledge about areas where the experts are hard to distinguish from an outside perspective. If the contacted persons had possibility to participate with their knowledge, an interview was scheduled. During the study, several interviews were held and the participants were separated into three different categories and are presented in Table 2.1 below. These categories consist of (1) blockchain technology experts, that gave insight to the technology’s capabilities and limitations, (2) SKF employees, that helped to build an under- standing of the case of SKF’s supply chain and (3) employees at other companies trying out blockchain in their supply chain. Six blockchain experts were interviewed and some questions about blockchain were only asked to them in the beginning of the study, to give the authors a clarifying understanding of the technology due to its complex nature. A total of five interviews were held to study the case of SKF and build an understanding of their supply chain and anti-counterfeit operations. Mainly, employees were interviewed based on their role in the company. To miti- gate the risk of bias, however, some questions were asked to several respondents to present different perspectives. An interview was also carried out with an industry expert to gain knowledge about a specific blockchain project. Overall, interviews were conducted until the data started to converge into similar types of answer, and additional interviews did not add much extra value. The data was later codified and is presented under 4.1 Expert Interview Data. 10 2. Method Name Company Position Date Blockchain Experts: Oliver Oram Chainvine CEO 2019-02-25 Ludvig Öberg Genesis Block Consultancy Founder 2019-02-26 Frida Höjvall RISE Project Leader 2019-02-27 Peter Altmann RISE PhD. Senior Researcher 2019-03-11 Mats Snäll Lantmäteriet Chief Innovation Officer 2019-03-14 Sukesh Kumar Tedla Swedish Blockchain Association Chairman 2019-03-26 SKF Employees: Johan Bravert Petter Rönnborg SKF Director GBP & Brand Protection Manager 2019-02-07 Ulf J Andersson SKF L&DC Planning Sales & Operations 2019-03-14 Hans Sjöström SKF P&ICR Automotive & Aerospace & Global 2019-03-25 Ulrica Nilsson SKF Brand Protection Manager 2019-04-15 Ketil Eliassen SKF Brand Protection Manager 2019-04-16 Industry Experts: Hans Svensson Stena Steel Vice CEO & Market Director 2019-04-10 Table 2.1: Interview objects 2.2.2 Secondary Data The need for secondary data from other blockchain project was motivated by the fact that blockchain is a relatively new field of technology where there is a lack of theory on its application in supply chains. Since there are many examples of companies using blockchain in some sort of supply chain, the projects presented in 4.2 Examples of Blockchain Projects were specifically chosen because they presented a different set of application areas for blockchain. To find these projects, the authors used several databases, such as Google, Google Scholar and Chalmers’ library’s digital platform, together with asking blockchain experts for examples during the interviews. 11 2. Method 2.3 Data Analysis The data gathered consisted as previously mentioned of expert interviews on the topic of blockchain, secondary data on other use cases for blockchain application and case study interviews with SKF representatives. The thorough analysis of this data together with theory was then the foundation for answering the study’s research questions. The first step of the data analysis was thus to collect the data. Even if collecting data per se is not analysis, it gradually builds up an understanding of the studied subject. Second, to get a better overview of what the interviews had covered up to that point, the data from the expert interviews was mapped against the research questions. This activity was done together by the authors in an unstructured way, meaning that both read through each interview and mapped data under the research question it was most relevant to answer against. The activity was valuable to make sure that both authors got a deep understanding of the existing data, but also en- abled an understanding of how well the accumulated data answered the research questions and what gaps still existed. The third step of analysis was to codify data in a more detailed manner. Domains were identified under each research question and then built out with data and examples of expert quotes, to show similarities and differences between the answers of the experts. Fourth, the data from interviews with SKF employees were used to map up the existing SKF supply chain and how GBP conduct their work. The blockchain projects were used to compare similarities with the supply chain at SKF as well as identify general challenges with blockchain implementation. By connecting the codified data with the understanding of SKF’s supply chain, examples of other blockchain projects and theory, insights were either gained and discussed under 5 Analysis and Discussion, or gaps were identified. If gaps could not be filled by generating new constructs from the existing data, new interviews were needed that addressed the identified the gaps. The data from the second round of interviews was then directly codified into the already developed constructs since there was no need to map it against a specific research question. This iterative analysis processes is presented in Figure 2.2 below. 12 2. Method Figure 2.2: The analytical framework 2.4 Research Quality This section presents how the authors ensured a good thesis quality as well as some ethical considerations that were taken into account. 2.4.1 Reliability and Validity There are three main criteria for assessing the quality of a research; reliability, replication and validity (Bryman and Bell, 2015). Reliability is concerned with the quality of measures, if knowledge is gathered in a consistent and trustworthy way. This is important to enable the result of a research to be repeated but with other objects. The research must also be capable of replication to determine the reliability of a measure. Bryman and Bell (2015) divide reliability into internal and external reliability. The former entails whether or not the researchers agree about what they are observing, while the latter is the degree to which the research can be replicated. For qualitative research, external reliability is difficult to meet since a social setting is easily changed, making it hard to replicate the initial research. This study has tried to maintain reliability and trustworthiness by being transparent about the different interview objects, how they were contacted and what questions were asked during these meetings. The authors have also described in detail how the study was conducted to enable replication and reliability. Validity is seen as the most essential criterion of research and can in general be defined as the relevance of the collected data, i. e. if the researchers are observing or measuring what they are supposed to (Bryman and Bell, 2015). A valid measure needs to be reliable and measure what it is intended to measure. Similar to reliability, 13 2. Method validity is also divided into internal and external validity. Here, internal validity entails how well the observations match the theoretical ideas, while external validity involves the findings ability to be generalized. To obtain high validity throughout the study, the authors have used triangulation, which according to Bryman and Bell (2015) involves using more than one method to collect data on the same topic. As mentioned above, the data concerning blockchain technology was collected through interviews, a case study and secondary data. Regarding high quality of theory, the literature mostly consists of academic research papers and books, but also extend to announcements made by legit actors in the blockchain community and educationally written articles from online forums. The latter sources were only used on a few occasions, and were then critically evaluated before used as they are not to the same extent reviewed before published. This was done by confirming the information with other, more formal sources, but the sources were still kept as they provided a simpler way of explaining the theory in those cases. 2.4.2 Ethics Bryman and Bell (2015) present ethical considerations that needs to be taken into account when conducting a research: data management, copyright, reciprocity and trust along with affiliation and conflicts of interest. During this study, these con- sideration have been managed by informing interviewees about the purpose of the study and how the information given by them would be used. The authors have also asked for the interviewees permission to use that information in this report. Furthermore, contracts with SKF was signed to clarify conditions about sensitive information. A final version of this report was presented to SKF before publishing to ensure their consent. 14 3 Theoretical framework An important part when exploring a new topic is to create a theoretical framework of literature in relevant domains. The first purpose of this activity was to bring an understanding of the individual building blocks of theory that were relevant to comprehend the setting of the study. Another purpose was to iterate between collected data and theory, to see if what was observed and analyzed could find support in existing theory. A third purpose was to use theory to find or develop frameworks that helped to better analyze collected data. 3.1 Introduction to the Theoretical Framework In order to give an answer to the general research question, i.e. if blockchain is suitable to use in a supply chain to prevent counterfeiting of physical products, sev- eral theoretical areas need to be addressed. These areas are: blockchain technology, tags that can create unique identities, global supply chains, digitalization of supply chains and chain of custody. A section explaining the blockchain technology, and its different components, is needed to give the reader a brief understanding of what blockchain actually is. This is then followed by a section presenting a few different techniques to uniquely tag a product, to show which methods are used today. To further investigate whether blockchain is suitable to use in a supply chain of phys- ical products, global supply chains are addressed to demonstrate how the different actors in a supply chain are connected to each other, as well as problems and risks associated with them. This is then naturally followed by a section explaining how digitalization has changed supply chains along with demands, benefits and risks of digital supply chains, to give the reader an understanding of what blockchains role can be in a supply chain. Finally, four chain of custody models describing how the ownership of a product changes within a supply chain are presented to show how traceability is done today. How these theoretical areas relate to each other is visualized in Figure 3.1 below. 15 3. Theoretical framework Figure 3.1: A conceptual model of the theoretical framework 3.2 Blockchain Technology Blockchain can briefly be described as a decentralized and distributed database sys- tem for transactions of different types of assets, including currency, material and immaterial property (Swan, 2015; Appelbaum and Smith, 2018). Swan (2015) de- scribes blockchain as a big spreadsheet containing all assets in a global network that can be seen by all participants in the network. Originally, blockchain was created to enable safe transactions between two parties without the need for a third party or the element of trust, as it enables transactions directly between two parties and is visible to and verified by the whole network (Nakamoto, 2008). This lead to the creation of the cryptocurrency Bitcoin, which is the first application of blockchain. In Bitcoin, the blockchain is the public record of all Bitcoin transactions that have ever occurred and the technology builds up a safe payment system (Swan, 2015). Given that the intention of this section is to provide the reader with an overview of blockchain that enables an understanding of the thesis, blockchain is not presented in full depth. Therefore, the authors do not claim that this is a complete description of what blockchain is, but sufficient enough for the purpose of reading this thesis. 3.2.1 Describing Blockchain as Applied in Bitcoin Explaining blockchain in a straight line from start to finish is not simple. There are several different parts that together constitutes blockchain and are circularly dependent on each other. To understand blockchain is a matter of understanding 16 3. Theoretical framework the individual parts and then connecting them together. Therefore, it could be suitable to explain these parts individually at first, to establish the domain in which blockchain exists. The parts considered here are the network of nodes, transactions and how they are conducted, blocks of transactions and how they are linked together in a chain, the consensus algorithm and the incentives that exist in a decentralized system like Bitcoin. Since Bitcoin is the first application of blockchain, and currency is an intuitive way to understand transactions, explaining blockchain through its application in Bitcoin seems suitable. The cryptocurrency Bitcoin was created by the person or the group of persons called Satoshi Nakamoto and was released on the 9th of January 2009 (Nakamoto, 2009). Before Bitcoin was released, Nakamoto wrote a paper called "Bit- coin: A Peer-to-Peer Electronic Cash System" which explained the problems with existing electronic payment systems and proposed a solution that removed the need for a trusted third party and solved the problem of double-spending (Nakamoto, 2008). Nakamoto further argues that in most transactions over the Internet, fi- nancial institutions such as banks are needed as a third party to process electronic payments and make these payments trusted by consumers and make the system work. As electronic payments are not visible and tangible, banks are needed to legitimate transactions. At the same time, they incur transaction costs and make transactions less efficient and flexible. By creating a technology that uses computers in a network to keep track of a timestamp server that is distributed peer-to-peer and requires proof to be computed, Nakamoto built an electronic payment system that removed the third party and still kept the transactions safe and currency impossible to double-spend. 3.2.1.1 Network The world of Bitcoin is built up by a network of nodes, where the nodes constitute every computer that is connected to the Bitcoin network (D’Aliessi, 2016). These nodes are interconnected to each other (Nakamoto, 2008). The interconnection between nodes and how a transaction is spread in the network is illustrated in Figure 3.2. Every node in the network has a copy of the public record, a so called ledger, of every transaction of Bitcoin that has ever been recorded (D’Aliessi, 2016). Thus, the system is distributed, meaning that the ledger is not stored in a central location, but separately within each node, so that no one has full authority. How to make sure that all nodes share the same ledger, and thus transactions history, is described in the following sections. 17 3. Theoretical framework Figure 3.2: The Bitcoin network and the transmission of a transaction 3.2.1.2 Transactions With digital currencies, how can it be made sure that each unit of that currency is only spent once by the owner that holds it? Since a digital currency is not a material product that can be at only one location at any given time, this can be more difficult than it seems. As described above, this problem has often been solved by having a third party, like a bank, verifying transactions by checking them for double-spending (Nakamoto, 2008). In such a system, privacy could be kept in between the sender, recipient and the third party, and these would be the only ones to know about the transaction. In Bitcoin however, to avoid the problem of double-spending everyone needs to know about the full transaction history to know that money being sent has not been spent before. Therefore, all transactions must be announced to the whole system and nodes in the system must agree on a single history of transactions. Otherwise, an owner would be able to digitally verify the same unit of currency multiple times and send it to several different recipients. If everyone knows about the transactions that has happened in the system, such double transactions will not be accepted. A node can therefore only create outgoing transactions by referring to previously incoming transactions to its address, which implies that the network is self-referring (Bitcoin.org, 2019). But why would anyone accept that their monetary transactions are publicly known in a global network? The solution to that problem in blockchain is that every node in the network has an address, to which transactions are directed (Nakamoto, 2008). The identity of the owner of the address is not stored anywhere, and is therefore anonymous. The network can thus see that transactions are directed to a certain address, but do not know who the person receiving the money is. So when transactions are anonymous, how can one be sure that the money sent in a transaction are coming from the address it says it is coming from? To solve this 18 3. Theoretical framework problem, Bitcoin use public-private key cryptography which means that every node has a pair of a private key and a public key (Massessi, 2018). The private key is a string of random data and the public key is then generated from that private key (Bitcoin.org, 2019). The public key is then cryptographically hashed, which means that it is shortened and altered so that it cannot be re-engineered back to the original public and private key. A hash function is a function that converts data of any size into data of fixed size and is cryptographically secure (Konstantopoulos, 2017). This makes manual transactions easier and provides security against unexpected problems. The hashed public key could then be seen as a Bitcoin address to where transactions are to be sent (Bitcoin.org, 2019). Briefly explained, the private key is used to sign transactions and broadcast mes- sages, while the public key is then used by others to verify that the message or transaction was sent from that specific private key (Massessi, 2018). For example, as shown in Figure 3.3 below, actor A uses its private key to digitally sign and en- crypt a transaction message for 5 Bitcoins (BTC) to actor B, and then broadcasts it to the network. Nodes receiving the transaction message then use the public key of actor A to verify that it was indeed actor A that broadcasted the transaction in the network, thus verifying that it is legit. This is possible because the digital signature generated when encrypting a transaction with the private key is a string of text depending on the transaction request together with the private key (D’Aliessi, 2016). Hence, if the transaction request is changed, the digital signature will change, making it difficult for someone else to alter it. Thus, the public key is the only thing that can decrypt an encrypted message from the corresponding private key and vice versa (Massessi, 2018). Figure 3.3: The functionality of private and public key cryptography 19 3. Theoretical framework The private key is kept secret to prevent others from using it but the hashed public key can be distributed to the network without any problems since it merely is a Bitcoin address as well as a verifying tool (Massessi, 2018). The keys are simply used to sign and verify all transactions made in the blockchain network. Since the public key is anonymous, a new key pair could be used for every new transaction to prevent them from being linked to a common owner (Nakamoto, 2008). 3.2.1.3 Blocks In order to verify that transactions are valid and not previously spent, they must be placed in a chronological order after each other (Nakamoto, 2008). Therefore, trans- actions must be collected in blocks where the hash of a block must be timestamped and transmitted widely in the network. Since each timestamp includes a link to the timestamp of the previous block, the blocks get ordered after each other in a chain of blocks: a blockchain. This is visualized in Figure 3.4 below. As transactions within the same block are considered to have occurred at the same time, it is possible to say what transactions occurred first as new blocks are added to the chain and therefore make sure that Bitcoins are not being double-spent. From the moment a new block is created and accepted, nodes collect all new transactions that reach them into a new, private block of transactions (Nakamoto, 2008). While doing so, they work on solving a mathematical problem called proof-of-work, which is explained below. Figure 3.4: An illustration of a blockchain and how blocks are linked together 3.2.1.4 Proof-of-Work Consensus about which block is added next in the blockchain is needed to agree on the single, valid transaction history in the blockchain. The algorithm used to do so is a mathematical problem or puzzle, that needs to be solved before a new block can be added to the chain, called proof-of-work (Crosby, Pattanayak, Verma, and Kalyanaraman, 2016). The name refers to that the node creating the new block needs to prove that it has used enough CPU power to solve the mathematical problem for that specific block. Solving the mathematical problem involves adding a random number, called a nonce, to a given data so that the hash outcome is less than a predefined number (Martinez, 2018). Reversing the hash function is 20 3. Theoretical framework impossible, which means that trying all different combinations is the sole way to solve the problem (Konstantopoulos, 2017). The more computing power that a node put in to solve the problem, the faster it will solve it. Once a node finds the proof-of-work, i.e. discover the correct nonce, it earns the right to broadcast the block it has been working on, its private block, to the whole net- work (Nakamoto, 2008). Other nodes then verifies the validity of that specific block by controlling that the hash outcome is less than the predefined number (Konstan- topoulos, 2017). Certain nodes, so called full-nodes, also verify that the transactions in the block are valid and that those Bitcoins have not been spent in earlier trans- actions (Nakamoto, 2008). If two nodes solve the proof-of-work at the same time and broadcast their private block to the network, both blocks could be accepted and connected to the blockchain and create a so called fork (Bitcoin.org, 2019). Some nodes receive one of the blocks and some nodes receive the other. To show that they have accepted the new block, nodes start to work on adding the next block after the one just created. The consensus model in Bitcoin implies that the longest chain of blocks is the single valid history of transactions, because most work has gone into that chain (Nakamoto, 2008). So the first side of the fork that adds an additional block will be considered the correct one. Thus, the network has agreed on a single, valid history of transactions. Due to the large amount of work to be done before proposing a new block, there is little risk of a block containing fraudulent transactions since this block will be rejected by the other nodes and thus become very costly for the node proposing the block (Singhal, Dhameja, and Sekhar Panda, 2018). For an average computer, solving the proof-of-work could take approximately one year (D’Aliessi, 2016). Given the large number of computers (or nodes) in the network, the average time to add a new block to the blockchain is ten minutes. Since changing a transaction in a previous block requires the need to redo all proof-of-work for all subsequent blocks, each additional block reinforces the ones that came before it (Bitcoin.org, 2019). 3.2.1.5 Incentives in a Decentralized System The risk for fraudulent transactions and adding blocks containing them, logically decreases with the proof-of-work algorithm. But why would any node want to try to solve the mathematical problem if it costs a lot of electricity and CPU power? There are two main monetary reasons for that in the Bitcoin network. The first reason is that whoever solves the problem and gets to add the new block to the blockchain also gets to create a special transaction in the block that sends Bitcoin to their own address (Nakamoto, 2008). This will however not continue forever; the number of coins are predetermined and eventually all Bitcoins will be distributed in the network. The second reason is that transaction fees can be added to transactions broadcasted to the network, and these will belong to the creator of the new block. Once all Bitcoins have been distributed, the incentives could potentially transition to only be constituted by the fees added to transactions (Nakamoto, 2008). 21 3. Theoretical framework There is another major reason to play by the rules in the Bitcoin network. If an attacker would try to cheat the system and double-spend coins or try to overtake the system, the outcome would simply be that the system was undermined and not trusted. This leads to that the wealth of the attacker would lose its validity and value (Nakamoto, 2008). With this said, not all nodes see the need to try to solve the proof-of-work and add new blocks to the blockchain. Some just participate in the network to send and receive Bitcoins, using it as a payment system that is safe, trusted and flexible without having a central third party. The activity of running a software to try to solve the mathematical problem is called mining (D’Aliessi, 2016). Like gold miners invest their resources to mine gold, Bitcoin miners invest CPU power and electricity (Nakamoto, 2008). Since there are so many nodes in the network, and each individual node would take a long time to find the correct solution to the proof- of-work, nodes that conduct mining often arrange themselves in groups and split up the task between each other to speed up the process (D’Aliessi, 2016). When they find the solution, they split the rewards between themselves. 3.2.2 Private and Public Blockchains Since this field of study is relatively new, several definitions of blockchain exist and the community has not yet reached consensus on a definition of the technology. Potentially, one could make a distinction between two types of blockchains; permis- sionless and permissioned (Bussmann, 2017). For example Peck (2017) argues that a permissionless, also known as public, blockchain is open for anyone to enter and anyone who does is anonymous. Bitcoin is an example of a public blockchain were anyone could join and there is no central actor that is in charge of the network. Peck (2017) further argues that a permissioned blockchain instead contains actors with known identities, and only selected actors can view the data on the blockchain. In a permissioned blockchain, an actor needs to be approved by the authority or author- ities owning it (Peck, 2017). Cachin and Vukoli (2017) have a similar definition, but add that a permissioned blockchain is operated by a group of entities that create a blockchain for a certain context. Cachin and Vukoli (2017) further argue for a third type of blockchain called private, which is a special permissioned blockchain that is operated by one single entity. A permissioned blockchain is suitable when all actors in the network already have a small degree of trust among them and would like to replace the services of a neutral third party (Peck, 2017). The consensus problem regarding what block is added to the chain is, as previously mentioned, often solved by proof-of-work in a public blockchain, but in a permissioned one it is a bit different. Since the right to add new blocks is assigned by the owner of the private blockchain, there is no need for proof-of-work. Instead, there are several other ways to reach consensus. One common approach is that the nodes within the network vote for which block should 22 3. Theoretical framework be added with their digital signature, and the block is then added to the chain when the majority of the nodes have approved it (Cachin and Vukoli, 2017). 3.2.3 Smart Contracts In order to extend blockchains beyond monetary transactions, the ability to add conditions to a transaction is needed. For this purpose, something called a smart contract is used. These contracts allow for the possibility to exchange anything of value, not just money, without the need for a third party (Blockgeeks, 2018). The concept of smart contracts was introduced in 1994 by Nick Szabo who defines them as “a computerized transaction protocol that executes the terms of a contract” (Christidis and Devetsikiotis, 2016). Smart contracts are automated self-executing contracts that contains specific instructions which gets executed when certain con- ditions are met (Blockgeeks, 2018). The term self-executing refers to that when one set of instructions are done, the next set of instructions are executed and so on until the end of the contract. Sillaber and Waltl (2017) argue that smart contracts consists of three components: the contractual arrangements between the parties, governance and preconditions along with execution of the contract. The first com- ponent means that the involved parties agree on certain conditions which then are transcribed into executable program code that is stored in the blockchain. Each party is identified through their blockchain address. The second component means that all nodes evaluate whether the preconditions defined in the contract is met or not. The last component is the actual execution of the contract. If the preconditions have been met, the contract is executed and the transactions between the involved parties are performed. All transactions made through the smart contract is recorded and updated by the network, which makes each involved party accountable for their actions. Smart contracts are scripts stored on a blockchain that has a unique address and are triggered when a transaction is sent to that address (Christidis and Devetsikiotis, 2016). A smart contract needs to be deterministic, hence the same input must always produce the same output. Otherwise, when it is executed on every node, it might return different random results which would lead to difficulties in reaching consensus throughout the network. All interactions with a contract occur through signed messages that uses the same private and public key cryptography as explained above (Szabo, 1997). 3.3 Techniques to Create Unique Identities As counterfeiters become better at making copies of genuine products, it becomes harder for the manufacturers of the original goods to verify whether a product is genuine or not. Therefore, manufacturers try to find new ways of tagging or 23 3. Theoretical framework labeling their products that are difficult and expensive to copy, but easy to verify and preferably as cheap as possible. Below, a few of the methods existing today are presented. A very common way to identify different products is to stamp a barcode on the prod- uct or its packaging. A barcode is simply a visual representation of data describing the product and can be read by a machine. A QR-code could be said to be a type of barcode. Barcodes are an easy way to identify a product, but can however easily be copied and printed onto fake products. Therefore, it barely offers any protection against counterfeiters. One method to identify a living organism is by its DNA. Jeffreys, Wilson, and Thein (1985) early showed that it was possible to create “fingerprints” from DNA that was completely specific to a person, or an identical twin, from as little as one drop of blood. It was later shown that this method was suitable to use as a forensic approach to identify suspects (Gill, Jeffreys, and Werrett, 1985). Given these characteristics of DNA-tests, they can be made to identify for example chickens, to ensure that they come from the farm they are claimed to be from. However, most physical products do not have a DNA. Another way to create a unique identity for an item is by DNA marking it (Selec- tamark Security Systems, 2019). The marking liquid, or spray, contains a unique DNA combination that is applied to the object and can be registered in a database. Removing the mark is hard, and even if it could be removed, it cannot be reattached to another item (Jung, Hayward, Liang, and Berrada, 2016). According to Jung, Hayward, Liang, and Berrada (2016), DNA marking also have physical properties that cannot be replicated, creating a sufficient level of unique identification to en- able authentication and provide a track and trace system for marked items. DNA marking does however not apply to a large amount of identical physical products since it would require a new spray for each item. Quantum dots are semiconductors in the size of nanoparticles (Murphy, 2002). Ex- citation of quantum dots create an emission wavelength spectra when hit by a light source, and given their unique optical properties this makes them suitable for ap- plications like anti-fake labeling and security identification (Bai, Du, Zeng, and Yu, 2008). Chen, Lai, Marchewka, Berry, and Tam (2016) present a low-cost method for producing nano-thin films of quantum dots and cellulose nano-crystals that can be used for anti-counterfeiting purposes. Many different colors can be emissioned by adjusting the concentrations of the two building blocks of the film. Radio Frequency Identification (RFID) is a device with an antenna connected to a microchip that provide identification of different products or goods (Tuyls and Batina, 2006). The information can be accessed through a reader and used for au- thentication and tracing, possibly for anti-counterfeiting purposes. If the necessary information is there, the product is declared to be genuine. However, the informa- tion on the chip can be captured and copied onto a new chip, making it possible to 24 3. Theoretical framework clone it. 3.4 Global Supply Chains In the manufacturing sector, the supply chain can be viewed as a network of actors that are interdependent of each other and together physically transform materi- als, components and substances into finished products (Klötzer and Pflaum, 2017). These actors are together responsible for managing and improving the flow of ma- terial from suppliers to the end users. In the global economy, companies operating internationally will have to manage a global supply chain. Due to many different factors, such as different currencies, tariffs and taxation differences, an international supply chain becomes more difficult to manage than those contained within one single country (Vidal and Goetschalckx, 1997). In order to understand the com- plexities in the supply chain of a global company it is important to first build an understanding of the complexities within a single market. Mattsson (2003) has a markets-as-networks view, describing markets as "networks of multidimensional, dynamic exchange relationships between economic actors who control resources and carry out activities". Markets can be viewed in terms of connectivity, which de- scribes the amount of direct and indirect connections between actors on the market. Here, Mattsson (2003) makes the distinction between positive and negative connec- tions. A positive connection is based on cooperation between e.g. a supplier and a buyer, while a negative connection is based on competition for relationships third parties. To a large extent, processes in the market is endogenously created through the activities of the present actors. At the same time exogenous forces exists on a market to affect its structure and rules to play by, stemming from actors that are not seen as part of the market, such as governments. Since any firm’s supply chain can be seen as embedded in the markets-as-networks view, it is dependent on all the actions made by actors within the network (Mattsson, 2003). In the network view of markets, each actor can be seen to have one position in the market, which is built up by the connections they have in the market and the internal resources they control. But, as relationships between actors constantly changes, interconnection with other markets become less or more prominent. Since actors enter and exit the market, the market is highly dynamic. There are several problems associated with the global supply chain of a firm. (Matts- son, 2003). Internally the firm must organize to coordinate activities across nations and externally the firm is connected to other actors on the market in cooperative and competitive relationships. Thus, the firm is not only affected by the actions they do themselves, but also by those of other actors in the market network such as competitors etc. The risks associated with a global supply chain increases since it often have been extended, is more complex and harder to evaluate in terms of impact of potential 25 3. Theoretical framework disruptions (Manners-Bell, 2014). With extended supply chains, problems like e.g. more hand-offs to other parties, quality control challenges and corruption further down the chain can appear. 3.5 Digitalization of Supply Chains Under this section, the concept of digitalization is presented, and it is elaborated on how digitalization has changed the operations and management of supply chains in later years. The new set of demands on the Digital Supply Chain is also explained, along with the benefits and risk. Finally, a brief understanding of how blockchain could be applied in this environment is presented. 3.5.1 What Is Digitalization? The term digitalization is often thought of as the process of making data and pro- cesses digital. This is only partly true, which is why it is necessary to define some terms before continuing. Digitization is the process of converting analog data to digital data, which means to transform it into digital bits (Tilson, Lyytinen, and Sörensen, 2010; Gobble, 2018. This definition is confirmed by Gartner (2019), that describe digitization as “taking an analog process and change it to a digital form without any different-in-kind changes to the process itself”. Digitalization, on the the other hand, is something different. Tilson et al. (2010) describe it as applying digitization in a wider social and institutional context, where the digital technologies become infrastructural. Further, digitalization could be referred to as using digitization and the digital technology to create new ways to generate value. To make a distinction between the two, digitization does not change the underlying business model, while digitalization in the end does (Tilson et al., 2010; Gobble, 2018). A visualization of these two concepts can be viewed in Figure 3.5 below. Figure 3.5: Digitization and Digitalization 26 3. Theoretical framework 3.5.2 How Digitalization Has Changed Supply Chains The relatively recent development of digitalization of the economy has affected, and is continuously expected to affect, the development and management of global supply chains (Klötzer and Pflaum, 2017). In this era where companies become more digital, data and information is of critical importance. Gunasekaran and Ngai (2007) define a digital enterprise as “characterized by the application of Information and Communications Technology (ICTs) [...] for the integration of activities in different functional areas as well as the so-called extended enterprises or partnering firms in the supply chain”. Given the definition of supply chains, it follows that integration between actors in them require integration of processes and information (Korpela, Hallikas, and Dahlberg, 2017). Digital technology, e.g. Electronic Data Interchange (EDI), has enabled this integration between supply chain partners, to base actions on the same data (Christopher, 2016). This has in later years changed business models in the supply chain to be based on real data instead of forecasts. Christopher (2016) argues that this type of collaboration is becoming more common and is also needed as companies tend to focus more on their core competencies and outsource remaining activities. The term Extended Enterprise can be used to describe these collaborative relationships where buyers and sellers share a vision to create larger end-user value to beat competing supply chains (Spekman and Davis, 2016). In the Extended Enterprise environment, where big companies becomes hubs in their network, collaboration for integration is needed throughout the whole supply chain and not just between two actors (Korpela et al., 2017). While there are several methods to integrate information directly between actors in the supply chain, companies still use third parties like banks for transactions between each other. The last years’ digital development in supply chains have led to the term Supply Chain 4.0, which Alicke, Rachor, and Seyfert (2016) define in the following way: Supply Chain 4.0 - the application of the Internet of Things, the use of advanced robotics, and the application of advanced analytics of big data in supply chain management: place sensors in everything, create networks everywhere, automate anything, and analyze everything to sig- nificantly improve performance and customer satisfaction. Customer expectations are growing in the Supply Chain 4.0. The competition be- tween supply chains is driven by transparency and the access to a wide range of product and channel options enabled by online functionality (Alicke et al., 2016). Overall, the Supply Chain 4.0 will thus include features like complete transparency of data, knowing location and conditions of individual shipments, automated feed- back on operations from machines, shared capacities in logistics between supply chain partners and flexibility to re-route shipments last minute. 27 3. Theoretical framework 3.5.3 Demands, Benefits and Risks of Digital Supply Chains The digitalization means that a new set of demands are placed on the supply chain. Gunasekaran and Ngai’s (2007) digital enterprises will create a network of both in- ternal and external partners. To succeed in such an environment, a common vision, alignment of goals and trust will be the foundation. Participants must feel confident that the technology system is secured and that privacy is protected. When talking about the Extended Enterprise, things like sharing of risks, rewards, information and technology is essential to succeed with collaboration and joint strategies (Spekman and Davis, 2016). Klötzer and Pflaum (2017) also add to the Extended Enterprise concept that the data and information itself is of critical importance. In this new paradigm of shared information, trust and commitment, joint determination towards strategies and transparency of information is needed in order to succeed (Christo- pher, 2016). Spekman and Davis (2016) also agree that since open communication and the use of technology to transfer information is critical for collaboration success, trust is essential. Looking more upon the technical demands on digital supply chains, the new trends and demands in the economy will force them to improve in several ways in order to stay competitive. The Supply Chain 4.0 needs to be faster, which requires advanced forecasting approaches based on e.g. internal demand data (Alicke et al., 2016). It needs to be more flexible by being dynamic to changes in demand through using real-time planning to a further extent. It also needs to be more granular when prod- uct customization will continue to become more present, and microsegmentation of customers is needed to achieve this. Further, Supply Chain 4.0 will have to be more accurate as its performance management systems will have real-time transparency throughout the supply chain. To achieve this, a huge amount of information from all levels and integrated between actors will provide a shared information base upon which different stakeholders make their decisions on. Finally, Supply Chain 4.0 needs to be more efficient. This will be achieved through an increased intensity of automation, both in planning and in operations. Robots will handle material flows by themselves in the warehouse processes, through activities such as receiving and unloading as well as picking, packing and sending goods. 28 3. Theoretical framework Figure 3.6: Demands on Supply Chain 4.0 A few benefits of a digital supply chain are increased efficiency and minimized cost for governing relationships to other actors (Korpela et al., 2017). The digital supply chain also makes information entries more frequent and accurate when it is properly used, minimizing the need for manual entry of data and in extension the amount of errors. Other benefits include reduced lead times and increased flexibility in the supply chain, delivering better value to customers through better services, more accurate prediction of demand and visibility. Regarding risks, Spekman and Davis (2016) mention six risk areas for the Extended Enterprise, from which a few can be associated with the changes inherent from digi- talization. In the Extended Enterprise, there is a security risk regarding a company’s internal information systems, both formal IT systems but also informal channels of communication. The risk is not the information itself, but who has access to it. Another risk is related to interdependence with other actors in the supply chain and the relationships built therein. With many relationships and shared information, there is a risk for opportunistic behavior from dishonest partners to for example steal information and act in their best self-interest. Risks need to be managed care- fully, because in the end a higher degree of risk will result in lower supply chain performance. 3.5.4 Possible Role for Blockchain in Digital Supply Chains When using third parties for transactions between actors in a digital supply chain, there are some drawbacks from a collaboration perspective (Korpela et al., 2017). First, full automation of data transfer is only possible for payments, and to a smaller extent also invoices. Second, involving more parties for exchange of documents in the supply chain leads to transaction costs and reduced speed. Third, there is a 29 3. Theoretical framework risk regarding cybersecurity, that criminals will hack into the system, e.g. a bank’s system, to steal sensitive information. To conduct transactions and exchange documents in a digital supply chain, parties must beforehand agree on how this is to be done (Korpela et al., 2017). A blockchain feature that can achieve this, not only for monetary transactions, is smart contracts described above. With smart contracts, digital supply chain transactions can be automated at a very granular level without the need for a third party. Due to blockchain’s capabilities, the technology seems suitable to provide similar services as the integration processes existing, but in a more flexible way with lower transaction costs. Blockchain does not however meet the need for standardization of electronic documents in the supply chain. To ensure full automation in documentation transfer, international standards would have to be relied on, and those would have to be further developed to be compatible for blockchain implementation. Even though current intermediates, such as EDI, have been practiced in supply chains for over 20 years, they lack functionalities that digital supply chains require for integration (Korpela et al., 2017). The authors mention lack of standardization, timestamping of transactions, overviewing and tracking flows of information as well as secure delivery of information. Blockchain technology have the potential to fill some of these gaps, which is needed to build a digital supply chain that is flexible and cost-effective. This would accelerate the integration between companies in a digital supply chain more than other systems could. However, integration requires stan- dards for interoperability between systems, and that is something that blockchain in itself does not offer more than any other solution does. 3.6 Chain of Custody Chain of custody refers to the chronological sequence that occurs when ownership or control of an asset is transferred from one actor to another in the supply chain (ISEAL Alliance, 2016). ISEAL Alliance (2016) define four commonly used chain of custody models that describe the systems used to trace the assets in a supply chain: (1) identity preservation, (2) segregation, (3) mass balance and (4) certificate trading. Identity preservation enables certified products, i.e. products that meet a certain standard, to be traced back to the point of origin. Since each batch of certified physical products is treated separately, this model ensures that both the products and corresponding documentation are separated from other sources, making the traceability possible. The second model, segregation, separates certified products from non-certified products, but it is not possible to identify the exact point of origin since certified products from different sources are mixed. Mass balance is slightly different from the two models just described, since it involves balancing volumes in the supply chain. Here, certified and non-certified products can be mixed, but the volume of certified products that enters the operation is equivalent to the volume of products leaving the operation that can be sold as certified. For 30 3. Theoretical framework example, if 10 kg of certified products enter an operation, only 10 kg of the products leaving the operation can be sold as certified. The volumes regarded in this model can be balanced at batch, site or group level. The final model, certificate trading, is similar to mass balance and used when certified and non-certified products are mixed freely within the supply chain. Certificates are issued in the beginning of the supply chain and can then be bought by actors on the market through a certificate or credit trading platform. This model intends to reward the producers of a product when it is difficult to trace the physical product back to the point of origin in the supply chain. However, there is no guarantee in any of these models that the physical end product is certified just because it has a supplied certificate since the product and the certificate is often sold separately. 31 4 Empirical Data and Findings This chapter presents the empirical data that was gathered for this study. The data consists of three different parts, as mentioned in the method chapter. These parts are expert interview data, primary and secondary data from other blockchain appli- cation projects and the case study of SKF. First, the expert interview data is divided based on the different research questions and presented individually. Second, the blockchain projects are presented in a descriptive manner, based on data gathered through official sources and interviews. Third and final, the SKF case study is pre- sented, also that in a descriptive manner but here the data was also gathered through on-site observations. 4.1 Expert Interview Data The expert interview data is presented under six different subsections, where each of the six sections correlates to the underlying research questions, one at a time. Per research question, the data is mapped against the different domains that were found and results for these domains are presented in tables, with exemplifying quotes if possible. 4.1.1 What Is Blockchain and What Are the Applications Used Today? In Table 4.1.1 below, data that answers against this research question is compiled. A broad description of blockchain is already presented in the theory chapter, so this expert data is presented merely to complement that description and to summarize the main attributes and capabilities. A summary of blockchain attributes is that it as a technology is not yet defined. However, most agree that it is a distributed and decentralized technology that is transparent to all actors in the network. The governance in a blockchain network, and the incentives to govern it, exist on the data structure and the protocol itself. Blockchain has the capability to automate processes 32 4. Empirical Data and Findings such as transactions, and reduce the trust needed in a network. Today, it is mainly applied in cryptocurrencies like Bitcoin. Given that blockchain is not defined, the experts acknowledge that the discussion around different types of blockchains exist, even though some are of the opinion that only public blockchains are blockchains at all. In the interviews, three types have been discussed. The are public, public and permissioned and private blockchains, and differ on the level of transparency and authorization that different actors in and around the network have. 33 4. Empirical Data and Findings 4.1.2 Can a Physical Product and the Digital Blockchain Be Irrevocably Linked Together? Regarding this research question, there are two distinct opinions whether or not this is possible to do. Some of the interviewed experts claim that to link the product and the blockchain together, an IoT device, forensic approaches, a QR-code/Barcode or some sort of tag could be used. However, other experts argue that these techniques does not ensure that the physical product matches the digital blockchain since for example QR-codes can be copied onto fake products and tags can be removed. With current technologies, the product requires a unique DNA for it to be impossible to copy. Thus, today the links are often between the digital representation and some sort of abstraction instead of between the product itself and the blockchain. Demands on the link is that it needs to be impossible to copy, durable and have the possibility to create a 1-1 link between the product and the blockchain. It must also be cheap to ensure that the benefit of having the link exceeds the cost of producing it. The data gathered to answer this research question is summarized in Table 4.1.2 below. 34 4. Empirical Data and Findings 35 4. Empirical Data and Findings 4.1.3 What Can Be the Main Benefits of Applying Blockchain in a Supply Chain? As seen in Table 4.1.3, there are many benefits of using blockchain in a supply chain. A summary of these benefits is that blockchain enhance transparency of all transac- tions made within the supply chain, which enables traceability and strengthens the trust in the network. Due to the transparency, the information in the blockchain cannot be manipulated without it being seen by the whole network, double-spending is not possible and it makes the system secure. With a blockchain, it also becomes more difficult for counterfeiters to sell fake products. 36 4. Empirical Data and Findings 4.1.4 What Can Be the Main Challenges of Applying Blockchain in a Supply Chain? The challenges connected to using blockchain in a supply chain is presented in two domains: in general for all blockchains and in private blockchains. The data gathered is compiled in Table 4.1.4. One major challenge that applies to all blockchains is the point of entry for the information. Since the blockchain itself cannot communicate if the information in it is correct or not, there is no efficient way to say if the information is authentic or not. Therefore, a good expression is “garbage in, garbage out”. Here, it means that the information entering the blockchain will stay there regardless if it is false or not. In summary, other challenges that apply to all blockchains are long transaction times, coordination difficulties with many actors along with difficulties connected to the implementation part such as the cost, legal issues and the user experience. Further, it is a challenge to create incentives for people to participate and validate public blockchains for applied blockchains. Regarding private blockchains, all the above mentioned challenges exist together with a few others. In a private blockchain there can be difficulties managing gov- ernance since the whole point of a blockchain is to remove the central controlling actor. Furthermore, deciding on a consensus algorithm or what actors in the supply chain that should be nodes in the blockchain network, constitutes a challenge if not all actors can be trusted. Another major challenge for the private blockchains is that there can exist several different blockchains within the same industry and no company want to join someone else’s. 37 4. Empirical Data and Findings 38 4. Empirical Data and Findings 4.1.5 What Processes in a Supply Chain Are Suitable for Blockchain Application? In table 4.1.5 below, the data gathered associated with this research question is assembled. As shown, there are three outlined domains where blockchains are used today: financial services, physical products and supply chains along with other use cases. Blockchain emerged in the financial service sector and are mainly a auditing tool for the banking industry and is used for monetary transactions. Today, blockchain is used in several processes in supply chains, such as documentation and certificate trading to mention a few. Another applicable process is for example when the content of an item is the important part, and not the item itself. A contract is a good example of this where the information within the contract is more important than the paper itself. 39 4. Empirical Data and Findings 4.1.6 How Could the Adoption of Blockchain Develop in the Future? The data that answers against this research question is assembled in Table 4.1.6. Primarily, it is worth mentioning that the blockchain technology is still relatively new and that it is hard to outline any specific future application areas. Some experts see great potential in the near future, while others are more skeptical about how far the development of the technology has reached today and see the big breakthrough first in a few decades. It will most likely not change the world on its own but together with other technologies such as IoT and AI, it will probably have an important role in the future. Some future application areas that the interviewees believe in are documentation verification, the mobility sector, in different supply chains and identity verification both in the IoT society but also in governmental systems for civilians. 40 4. Empirical Data and Findings 4.2 Examples of Blockchain Projects Given the capabilities of blockchain to provide traceability, digitalization of pro- cesses, make processes more secure and the supply chain more transparent, there are many examples of how a blockchain could potentially be used beyond Bitcoin. This section presents a fraction of the many cases were solutions that claim to be based on blockchain are used. 4.2.1 Blockchain in Digitization Projects In January of 2018, IBM announced that they together with Maersk were working on establishing a new platform to reduce barriers in global trade and increase security and efficiency in supply chains (White, 2018). The platform would be based on blockchain technology. The two main capabilities of the platform at its launch were to provide visibility throughout the supply chain for all actors, as they in a secure way could register shipping events in real time, and digitization of the massive paperwork processes that for long have been a part of the shipping industry. In August 2018, IBM and Maersk presented the result of their collaboration: TradeLens (IBM, 2018). In TradeLens, more than 100 organizations participate, including operators in ports and terminals, shippers and shipping lines, customs authorities in several countries, transportation and logistics companies. These actors are all a part of the digital supply chain, and participate to create a single view of all transactions in the ecosystem, without compromising the integrity of the individual actors. In TradeLens, actors collaborate on information exchange, creating a secure and immutable record of transactions by using a permissioned blockchain. Smart contracts are used to establish cross-organizational business processes and to prevent any actor from changing the business logic (Tradelens.com, 2019). Every step of the journey is added to the blockchain, enabling a paperless, frictionless and trusted network of actors. In TradeLens, participants can subscribe to events happening within the blockchain network. Depending on what role the own organization has within the network, it can subscribe to events regarding for example a port or a whole country. Regarding a specific shipment, only the participating actors in that shipment are able to submit, alter and approve related data. If any party would want to change anything concerning the shipment, it would require approval by all affected actors. Aside from the legitimacy aspect of these capabilities, actors can also get real time information about their shipments. According to Bridget van Kralingen, senior vice president for IBM Global Industries, Solutions and Blockchain, TradeLens have huge potential, but realizing it depends on whether the global shipping industry can unite around TradeLens as a common approach (IBM, 2018). Today, only a few months after its release, over 10 million events are processed on the platform every week (Tradelens.com, 2019). The next example does not consist of a distinct supply chain, but could be seen as a 41 4. Empirical Data and Findings supply chain of information with different actors. Lantmäteriet (belonging under the Ministry of Industry and responsible for the real estate division in Sweden) inves- tigated whether blockchain could be used in the process of property transactions. The current process of buying and selling properties today is quite complicated and involves several different parties such as seller, buyer, broker, banks, the state, Lantmäteriet et cetera. One central part of the process is the creation of a purchase agreement which is currently stipulated by law to be in paper form. Therefore, all documentation is done via physical papers and takes a long time. Thus, one of the reasons Lantmäteriet started this project was to influence the legislators and show them that it is possible to make the process digital. Together with Kairos Future, Telia Sonera and ChromaWay, Lantmäteriet developed a blockchain solution that involved every step of the process from broker to Lantmäteriet where it was finally registered. All the transactions became a workflow of digital signatures and smart contracts that were stored on the blockchain to ensure nothing was altered or ma- nipulated afterwards. This could then be used as evidence if one of the parties later broke the contract. With this solution, the lead time of the process was reduced from four months to only a couple of days (Lantmäteriet, 2016). Regarding consen- sus, when a transaction or update was made, all the nodes voted and validated it if they thought it was legit. Since the law still requires purchase agreements to be done in paper form, this project was never implemented in reality. Another major obstacle for this project involved not finding a suitable way of managing governance. Since purchasing agreements require personal information about the buyer and the seller, the transparency of blockchain became problematic given the information’s sensitive nature. 4.2.2 Blockchain in Traceability Projects As mentioned, blockchain can provide traceability of assets. Therefore, another application area for this technology is within the chain of custody systems previously described in section 3.6 Chain of Custody where sustainability is the main focus, such as the food industry. For example, IBM together with Walmart successfully traced mangoes back to its point of origin within seconds by using IBM Blockchain Platform (IBM, 2017). Further, a company called Cargill managed to track turkeys back to the farm they were raised at using a blockchain-based solution (Cargill, 2017). Blockchain reinforces the chain of custody models by increasing the transparency and trust between the actors in a network. If all transactions within the network is added to the blockchain, it enables the possibility to trace a product back to the point of origin. To enable this traceability of products, the products are in need of some sort of information tag, such as a barcode, QR-code or RFID (Abeyratne and Monfared, 2016). This information tag is what gives the product its unique identity and links the physical product to its virtual identity in the blockchain network. Through the virtual identity, it is then possible to display product information such as description, location, certifications and so on. 42 4. Empirical Data and Findings Regarding certificate trading, blockchain can secure and strengthen the digital cer- tificates to make sure that they are not altered or manipulated before being trans- ferred to another actor. Provenance offers a digital platform where certifiers and licensees can meet to share verified product information stored on a blockchain (Provenance, 2019). This also enables the traceability and transparency in a supply chain. Another example of a company using blockchain to trace products is Everledger who offers so called ecosystems of trust by using smart contracts, machine vision and IoT together with a blockchain platform to trace for example diamonds (Everledger, 2019). The company argues that the trust is created by the ability to transparently trace and manage assets on their way in the ecosystem through a trusted data protocol. The physical asset are here instead given a unique identity by forensic approaches rather than a tag, that later can be tracked as a digital asset stored in the blockchain network. The diamond’s unique identity depends on its cut, clarity, color et cetera, which is almost impossible to copy. Before a transaction can be added to the chain, consensus across the network is needed, which counteracts fraud and error. The blockchain ledger