Sie sind auf Seite 1von 25

Journal of Business Logistics, 2019, 1–25 doi: 10.1111/jbl.

12231
© 2019 Council of Supply Chain Management Professionals

Blockchain in Additive Manufacturing and its Impact on Supply


Chains
Stefan Kurpjuweit, Christoph G. Schmidt, Maximilian Kl€ockner, and Stephan M. Wagner
Swiss Federal Institute of Technology Zurich

A dditive manufacturing (AM) appears to be a particularly attractive use case for blockchain. This research combines inductive in-depth
interviews with the Delphi method to explore what potentials blockchain technology in AM creates, which adoption barriers firms need to
overcome, and how supply chains will be affected by the integration of these two potentially disruptive technologies. The results suggest oppor-
tunities that are related to intellectual property (IP) rights management, the monitoring of printed parts throughout their lifecycle, process
improvements, and data security. The most important barriers for blockchain adoption in AM are an absence of blockchain-skilled specialists
on the labor market, missing governance mechanisms, and a lack of firm-internal technical expertise. By addressing important limitations of
AM, blockchain is expected to improve the competitiveness of AM in parts’ production, catalyzing the trend toward more decentralized manu-
facturing resulting in more agile, resilient, and flexible supply chains and reduced logistics costs. Beyond that, blockchain-based AM platforms
are expected to enhance supply chain visibility, drive supply chain digitalization, support supply chain finance, and contribute to the emergence
of shared factory systems.
Keywords: blockchain; DLT; additive manufacturing; 3D printing; supply chain; digitalization; IP rights; Delphi study

INTRODUCTION propeller was hacked and manipulated, causing the drone to


crash a few minutes after takeoff (Belikovetsky et al. 2016). This
Additive manufacturing (AM) is the creation of complex compo- example highlights the vulnerability of digital manufacturing
nents by successively adding layers of material (Gibson et al. ecosystems. Thus, several studies call for research on AM and
2015). The technology has drawn increasing attention from both specifically the challenges associated with its digital nature
practitioners and academics, as it is undergoing a transformation (Holmstr€ om et al. 2016; Rogers et al. 2016). Overcoming these
from a rapid prototyping tool to an integral part of many indus- challenges could unlock the strengths of AM and affect supply
trial production systems (Durach et al. 2017; D’Aveni 2018). chains and logistics across many industries.
Particularly, logistics and supply chain settings are expected to Blockchain, the most prominent form of distributed ledger
be significantly affected by the implementation of AM processes technology (DLT), promises to ensure the authenticity of digital
in manufacturing networks. Given its unique one-step production information, allowing firms to manage the ownership of data
characteristic, various studies anticipate that the widespread (Christidis and Devetsikiotis 2016). Essentially, blockchain is a
adoption of AM will lead to the emergence of more decentral- distributed architecture based on cryptographically linked blocks
ized manufacturing facilities, shorter lead times, more agile sup- that contain consensus-validated data records (Benos et al. 2017).
ply chains, lower logistics costs, and higher levels of product Within manufacturing and supply chain management, several use
customization (Rogers et al. 2016; Durach et al. 2017). cases for the novel technology have been identified (Schmidt and
However, the interorganizational adoption of AM processes Wagner 2019; Wang et al. 2019). However, the combination of
does not come without dedicated risks. The AM supply chain blockchain and AM might be particularly promising as block-
evolves around transacting digital information, such as construc- chain can potentially address the challenges associated with the
tion plans, CAD files, or material specifications to manufacture digital nature of AM. Four blockchain characteristics substantiate
physical goods. In this process, firms frequently exchange sensi- the proposed fit. First, the AM process chain is primarily con-
tive information. Their value can significantly exceed the value cerned with digital design files, printing parameters, and process
of the printed parts; the digital construction plans embody the configurations. In contrast to the vast majority of traditional ana-
actual development efforts and costs. Due to the digital nature of log manufacturing techniques, the digital nature of the AM pro-
AM supply chains, major challenges to the adoption of AM cess, in theory, allows for the seamless and efficient adoption of
include increased vulnerabilities to data theft, potential IP a digital infrastructure, such as blockchain. Second, every digital
infringements, and an increased risk for counterfeit products process chain is associated with data security risks (Benos et al.
(Kurfess and Cass 2014; Gao et al. 2015; Yampolskiy et al. 2017). Blockchain provides immutability of records which could
2018). As an extreme illustration of the relevance of data secu- safeguard sensitive manufacturing information against data
rity in AM, one study shows how the CAD file of a drone manipulation and theft (Crosby et al. 2016; Casey and Wong
2017). Third, AM is also exposed to various IP-related risks and
constraints (Gao et al. 2015; Yampolskiy et al. 2018). A dis-
Corresponding author: tributed ledger could help originators of design files to protect
Stephan M. Wagner, Department of Management, Technology, and their IP rights, as a blockchain-based platform validates owner-
Economics, Swiss Federal Institute of Technology Zurich, Wein- ship of data and controls access to protected files. For example,
bergstrasse 56/58, Z€urich 8092, Switzerland; firms can retrace the number of copies a value chain partner has
E-mail: stwagner@ethz.ch already printed. Finally, the unique characteristics of the
2 S Kurpjuweit et al.

blockchain infrastructure can foster supply chain visibility by While prior studies have made valuable contributions to the
ensuring that all network participants have access to valid high- understanding of how blockchain might affect supply chains on
quality information (Wang et al. 2019). This is particularly cru- a general level, not tailored to AM specifically, they leave us
cial in distributed manufacturing networks, which can consist of with few empirical insights. In particular, since large-scale adop-
hundreds of AM service providers and OEMs, where a single tion examples beyond pilot projects are still missing, the initial
source of valid information could reduce complexity and facili- studies derive their findings mainly conceptually. At this stage,
tate coordination and collaboration among the various transaction we believe that the literature will benefit from adding empirical
partners without the need to build trustful relationships. research to this predominantly theoretical discussion. Therefore,
In practice, companies already recognize this potential and we followed the research design of a Delphi study, but replaced
thus join forces in large consortia projects to explore the com- the first Delphi round with an inductive, qualitative approach
bination of both technologies. For instance, the objective of the based on 12 in-depth expert interviews. The interviews allowed
Genesis of Things (GoT) project is to develop a blockchain- us to explore and identify potentials, barriers, and supply chain
based AM exchange platform, connecting many independent implications related to the integration of blockchain and AM in
printing facilities and allowing participants to store and transfer more detail. Based on the results of the expert interviews, a Del-
AM construction data securely on the blockchain. The platform phi panel of 53 experts rated the importance of the previously
will enable firms to select the most suitable printing provider identified factors in two rounds in order to achieve consensus on
for their production needs, place orders, and settle payments their importance and possible impact on supply chains. Subpanel
automatically (Blechschmidt and St€ ocker 2016). The project comparisons, such as between managers, consultants, and aca-
envisions the shared factory enabling companies to shift pro- demics, ensure the robustness of our Delphi study and generate
duction capacity flexibly according to actual demand. As additional findings.
another prominent example, the SAMPL (secure additive manu- The results provide insights into the emerging field of block-
facturing platform) project pursues a similar approach to estab- chain technology in operations and supply chain management
lish a chain of trust by developing a comprehensive security and add to the larger discussion on emerging technologies in
solution for AM using a blockchain infrastructure (SAMPL supply chains (Wu et al. 2013; Kache and Seuring 2017; Lee
2017). 2018) with special attention to managing the confidentiality of
During the implementation phase, the project partners face digital assets in supply chains (Massimino et al. 2018). More-
substantial barriers. To date, only few studies have explored the over, our study contributes to literature streams investigating AM
contingencies of blockchain adoption (Swan 2015; Tapscott and adoption, especially the digital nature and vulnerabilities of AM
Tapscott 2016). However, they address the topic very broadly (Kurfess and Cass 2014; Chen et al. 2015; Holmstr€ om et al.
and only conceptually. We build on these studies and provide 2016), as well as AM’s role in transforming supply chains
further empirical evidence on the factors hindering blockchain (Fawcett and Waller 2014; Durach et al. 2017; Ryan et al.
implementation. Specifically, the AM setting adds practical value 2017).
by offering managers, who consider blockchain adoption valu-
able, novel insights that could potentially reduce costs and
improve the implementation quality of their projects. Understand- RELATED LITERATURE
ing potentials and barriers, as well as their relative importance,
can also provide scholars with a starting point for future Additive manufacturing in supply chain management
research. This leads to our first research question: What are the
potentials and barriers of blockchain adoption in AM? In contrast to conventional subtractive manufacturing processes
The adoption of new technologies can create significant impli- (e.g. drilling, turning, or milling), additive manufacturing (AM),
cations for supply chains. Regarding AM, scholars predict the also known as 3D printing, is the umbrella term for a set of man-
structural decentralization of production sites, enhanced product ufacturing processes that build physical 3D objects by adding
customization opportunities, lower logistics costs, and improved successive layers of material (Huang et al. 2013). The premise of
supply chain capabilities (Rogers et al. 2016; Durach et al. AM is that “a model, initially generated using a three-dimen-
2017). Regarding blockchain, early studies identify the potential sional Computer Aided Design (3D CAD) system, can be fabri-
to improve supply chain visibility, decrease opportunistic behav- cated directly without the need for process planning” (Gibson
ior, and make supply chains more sustainable and resilient et al. 2015, p. 2).
(Kshetri 2018; Min 2019; Saberi et al. 2019; Schmidt and Wag- Its flexible application and the opportunity to quickly print
ner 2019). However, the potential reciprocal effects of both tech- and evaluate new component designs have made AM a popular
nologies remain unexplored, even though these insights would prototyping tool for new product development (Chen et al.
be valuable for both academia and practice. In addition, the syn- 2015). Technological advancements have triggered a paradigm
ergies of both technologies may create unique implications that shift from prototyping toward a legitimate manufacturing
go beyond the individual effects. Thus, exploring if and how approach (Huang et al. 2013; Calignano et al. 2017). AM pro-
blockchain might be the catalyst for AM promises important cesses are particularly competitive in low-volume and high-vari-
insights. Therefore, we specifically investigate how blockchain ety contexts (Holmstr€ om et al. 2016) such as spare parts’
might affect supply chains that are characterized by the interorga- production, tool production, or in industries like aerospace that
nizational application of AM processes (i.e. AM supply chains). are characterized by high-value parts (Wagner and Walton 2016).
Consequently, this leads to our second research question: How The application of AM in large-scale production is expected
does blockchain impact AM supply chains? to have implications for global supply chains in the areas of (1)
Blockchain in Additive Manufacturing 3

supply chain structure, (2) logistics performance, (3) supply complexity could challenge the overall visibility of supply
chain capabilities, and (4) customer-centric business models chains. Moreover, AM not only facilitates the outsourcing of
(Durach et al. 2017). From a structural perspective, most studies production to local service providers but it may also allow OEMs
suggest the emergence of more distributed or decentralized to involve customers or external designers to customize or com-
manufacturing (Wagner and Walton 2016; Durach et al. 2017; pletely design the products according to their own preferences
Oettmeier and Hofmann 2017; Ryan et al. 2017), associated (Ryan et al. 2017). In this way, multiple new sources of informa-
with shorter lead times and more agile supply chains (Fawcett tion emerge that need to be connected with the focal firm in a
and Waller 2014; Kothman and Faber 2016; Muir and Haddud manner that allows supply chain partners to exchange relevant
2018). Decentralized on-demand AM facilities further drive sup- part data or process-related information instantaneously. In an
ply chain capabilities such as flexibility (Rogers et al. 2016; AM context, it is thus crucial to gather all relevant information
Sasson and Johnson 2016; Ryan et al. 2017). Both agility and in order to create visibility along the supply chain. Centralized
flexibility facilitate supply chain responsiveness and resilience databases that are commonly deployed in contemporary supply
(Ghadge et al. 2018), since AM adopters are able to react faster chain settings, however, often inefficiently gather and authorize
to changing customer needs in volatile environments. Moreover, supply chain information (Wang et al. 2019). Common databases
production decentralization is often associated with decreasing further may not be able to address the data security needs of the
logistics costs in form of a reduction of safety and average AM environment and to create sufficient visibility to enable effi-
stock (Liu et al. 2014; Wagner and Walton 2016; Ghadge et al. cient AM networks.
2018), as well as reduced transportation costs (Huang et al.
2013; Ben-Ner and Siemsen 2017). In the long run, AM might Blockchain in the context of additive manufacturing
also support the direct integration of customers in the produc-
tion and design process and enable new business models, ulti- Blockchain, the most prominent form of distributed ledger tech-
mately transforming the supply chain into a demand chain nology (DLT), is in essence a distributed database of transac-
(Williams et al. 2002; Berman 2012; Christopher and Ryals tional records with a set of unique features. Cryptographic hash
2014). algorithms chronologically link blocks, enabling the storage of
transaction data, to the previous one by considering the previous
Supply chain visibility and additive manufacturing block’s hash as an input for the hash function (Christidis and
Devetsikiotis 2016). New blocks are added to the emerging
Supply chain visibility can be understood as having “access to chain, when transactions have been validated by the majority of
high quality information that describes various factors of demand nodes within the network, utilizing a consensus mechanism
and supply” (Williams et al. 2013; p. 545). This definition is (Swanson 2015; Brenig et al. 2016). Consequently, Benos et al.
often complemented by the ability to identify and verify impor- 2017, p. 1) define the blockchain as “a database architecture
tant information (e.g. identity, location, status) of a product tran- which enables the keeping and sharing of records in a distributed
siting along the supply chain (Francis 2008). Others refer to the and decentralized way, while ensuring its integrity through the
capability of knowing which path a product has taken as trace- use of consensus-based validation protocols and cryptographic
ability (Wowak et al. 2016). Despite the partially fuzzy terminol- signatures.” The resulting blockchain is not stored on a central
ogy, researchers agree that supply chain visibility is associated server, but on every node, forming a distributed database net-
with various positive operational and financial outcomes, includ- work (Cachin 2016).
ing reduced uncertainties and disruption risks, lower inventories, Blockchain provides an inherent degree of security, immutabil-
and improved responsiveness (Christopher and Lee 2004; Caridi ity of records, transaction transparency, authenticity of digital
et al. 2014). Overall, achieving supply chain visibility is not only information, ownership of data, and the possibility to replace
a major objective of supply chain management and logistics but trust from intermediating institutions by trust through crypto-
also an important outcome of key supply chain activities, such as graphic consensus (Beck et al. 2016; Christidis and Devetsikiotis
external integration and knowledge exchange (Barratt and Oke 2016). In addition, blockchain enables secure smart contracting.
2007; Williams et al. 2013). Smart contracts can be considered as programming code, repre-
Over the past decades, technological innovations have senting business logic, stored on the blockchain like transactional
improved access to supply chain information. In particular, glo- data (Christidis and Devetsikiotis 2016; Lu and Weng 2018).
bal and complex supply chains depend upon information technol- These codes make it possible to run arbitrarily customizable
ogy to collect, manage, and analyze data from various sources in applications, such as automated purchasing and payment func-
order to support supply chain managers in their decision making. tions (Swan 2015).
While research on technology adoption (e.g. EDI, ERP, and In a supply chain context, blockchain is expected to serve as a
RFID) has always been prominent in supply chain management, seamless peer-to-peer network infrastructure, overcoming the dis-
the recent wave of new digital technologies has led to a resur- connection of different information sources in distributed supply
gence of the topic (Feng and Shanthikumar 2018; Miller et al. chains. Information would no longer reside in private information
2018). However, firms need to fully trust that the information silos, but each supply chain partner obtains access to a valid source
provided is accurate, complete, and has not been altered unilater- of information in form of distributed copies of the blockchain
ally by any supply chain partner. transactions. This single source of truth would increase the reliabil-
As AM is expected to amplify the decentralization of manu- ity of information and prevent information asymmetries, which
facturing toward more local production, the increasing structural would ultimately result in increasing supply chain visibility.
4 S Kurpjuweit et al.

Beyond driving supply chain visibility in general (Kshetri investigates potentials, barriers, and supply chain implications
2018; Treiblmaier 2018), blockchain might hold the potential to related to the integration of blockchain and AM.
resolve AM issues in two major areas. First, the AM supply
chain hinges on the creation and exchange of digital assets, such
as CAD files (Chen et al. 2015). For instance, the construction METHODOLOGY
models have to be developed, iteratively optimized, transformed,
and tested, to enable the physical fabrication of the 3D model. We apply the Delphi method to explore the potentials and barri-
The AM process creates sensitive manufacturing and processing ers of blockchain adoption in AM and to examine the implica-
data, generating a digital thread, the “information path that is tions for supply chain management. The Delphi method,
gathered and stored when manufacturing a single part” (Kim developed by Dalkey and Helmer (1963), is used to collect
et al. 2015, p. 2). While data manipulation, theft, and infringe- expert-based empirical data and is understood as an “iterative
ment are considered serious challenges for today’s AM supply multistage process, designed to transform opinion into group
chains (Gao et al. 2015), blockchain’s record-keeping capabilities consensus” (Hasson et al. 2000, p. 1008). In the first round, par-
might be able to innately protect the sensitive AM designs, miti- ticipants traditionally answer open-ended questions that elicit
gating these risks. Since the risk of losing IP is a major obstacle long, explanatory, independent, and associative responses and
for information exchange and collaboration in interorganizational facilitate the generation of new concepts. Subsequent rounds
relationships, the inability of existing technologies to protect IP might repeat the open-ended approach or have respondents rank
rights in AM supply chains could inhibit the development of or rate previously derived constructs according to several dimen-
decentralized AM production networks for established products sions (Hsu and Sandford 2007). As a consensus-based approach
(Kurfess and Cass 2014; Holmstr€ om et al. 2016). is particularly suitable to study the future evolution and potential
Second, going beyond the data created and used during the implications of new phenomena (Pare et al. 2013), we deem the
AM process, the printed part itself could become an important method appropriate to explore the emerging blockchain technol-
data source for organizations. Certifications, quality assurances, ogy in AM. In addition, the Delphi method has been widely
status and condition, performance measurements, input materials, applied in operations and supply chain management, specifically
and ownership transfers could be collected throughout the lifecy- to study technology adoption (Kache and Seuring 2017;
cle of a printed part. The digital representation of a physical Roßmann et al. 2018).
good and its historic and current properties is known as digital To answer our research questions, we follow a two-step pro-
twin (Schleich et al. 2017). Against this background, blockchain cess, depicted in Figure 1. In the first step, to mitigate the con-
and its ability to manage ownership of data and execute data cerns with many Delphi studies regarding oversimplified and too
rights might be a solution to safeguard relevant digital twin data general initial responses (Sackman 1975; Hsu and Sandford
of printed parts. For example, the startup Spherity has developed 2007), we replaced the initial open-ended Delphi round with in-
a blockchain-based platform on which digital twins of physical depth expert interviews (von der Gracht 2012; Jiang et al. 2017).
objects are stored (St€ocker et al. 2018). These objects are tagged Expert interviews can generate qualitative, context-enriched, and
with an identifier, creating a unique link to the digital twin for more in-depth input data. The approach is considered more accu-
each object. This immutable digital memory could in turn be fed rate and rigorous, especially when dealing with highly novel phe-
with all relevant information throughout the object’s lifecycle nomena (Edmondson and McManus 2007), such as blockchain.
(e.g. information on ownership, GPS position). In a second step, based on the potentials, barriers, and supply
Analogous to the adoption of past emergent technologies that chain implications identified in the interviews, we conducted two
promise substantial opportunities for SCM and business logistics Delphi rounds with closed-ended questions to reach consensus
(Cao et al. 2014; Schoenherr and Speier-Pero 2015; Kache and on their importance and the strength of their potential impact.
Seuring 2017), blockchain adoption is likely to evoke a range of
barriers adopting organizations will need to overcome. Evidence Step 1: Expert interviews
from prior conceptual studies suggests that blockchain adoption
might be challenging due to missing standards, complex and To interview a heterogeneous sample of interviewees and capture
conflicting regulatory conditions, privacy concerns, or the lack of a broad array of perspectives on the novel phenomenon, we
coordination among project partners (Swan 2015). Most concep- recruited experts in three ways. First, in an initial desk research
tual works emphasize technical challenges that relate to the phase, we identified two major European consortia projects,
immaturity of the technology such as performance, scalability, or seeking to design and implement blockchain-based AM
security concerns (e.g. Yli-Huumo et al. 2016; Tapscott and Tap-
scott 2017). However, there is neither empirical evidence con-
firming these barriers nor a systematic approach for Figure 1: The two-step Delphi research process.
categorization.
Step 1 Step 2
In conclusion, the literature and first practical examples indi-
cate considerable opportunities for combining both technologies, Delphi Delphi survey
while also indicating the existence of critical adoption barriers expert interviews

firms need to overcome. Moreover, the potential reciprocal (Identification PBI list
Round 1
Ranked
Round 2
(Rating (Adjustment
effects of both technologies on supply chains remain unexplored, of PBI)
of PBI)
PBI list
of PBI rating)
even though an investigation would create valuable insights for
PBI: Potentials, barriers, (supply chain) implications of blockchain adoption in AM
academia and business practice. Our study, therefore, empirically
Blockchain in Additive Manufacturing 5

platforms. In either case, the project partners have successfully concepts related to the interviewees’ experience and terminology
developed proof of concepts and are currently running pilot pro- (Gioia et al. 2013). For example, from the following statement of
jects with industrial partners to adopt their blockchain platforms E1, we generated the code audit trail: “In order to trust the
in real-world settings. For instance, one project aims to safeguard printed product, it is necessary to have the audit trail including
AM spare parts’ production in the aviation industry. Both pro- all printing parameters and the CAD design.”
jects deal with a heterogeneous set of stakeholders, such as Throughout the process of data collection and analysis, we
research institutes, large companies from various industries, and contrasted identified first-order concepts, outlined similarities and
small startup companies. Thus, the consortia executives were differences by taking the specific interview context into consider-
prime interview candidates due to their wide range of back- ation, merged codes, and adjusted terminology to ensure consis-
grounds, experience, and domain knowledge. Second, we tency. Then, we aggregated the identified concepts to second-
attended workshops and special interest conferences dedicated to order themes (Gioia et al. 2013). For instance, the concepts audit
blockchain technology and its application in (additive) manufac- trail, transparency, and traceability build the second-order theme
turing, to get in touch with experts in this field. Third, utilizing a lifecycle monitoring. We introduce these concepts and themes in
snowball sampling approach, we asked all experts to recommend more detail in our Results section.
additional interviewees (Denzin and Lincoln 2000). We stopped
sampling potential interview partners when the same experts Step 2: Delphi survey
were suggested repeatedly, indicative of a very shallow pool of
experts in the area. Our final sample consisted of 22 potential Based on the interviews, we developed a survey-based Delphi
interview partners. instrument to evaluate and rank the identified factors, namely (1)
We limited our sample to professionals or academics who the potentials of blockchain adoption in AM, (2) the barriers to
were either involved in an ongoing blockchain adoption project implementing blockchain in AM, and (3) the supply chain impli-
or conducting research on the topic. Potential interviewees also cations of integrating both technologies. Consistent with recent
had to have been engaged with blockchain or AM for at least ranking-type Delphi studies (Richardson et al. 2016; Kache and
12 months and hold leading positions in their respective project. Seuring 2017), we used five-point Likert scales. Blockchain
When contacting the potential experts, we briefly explained our potentials and barriers in AM were rated in terms of importance
research approach, our objectives, and the topics to be covered and criticality. For the supply chain implications, we formulated
in the interview, which allowed us to check the suitability of the statements for each (presumed) effect and asked the panelists to
pre-selected experts for our study. Following several reminder express their degree of agreement. For every item, panelists
letters and phone calls, 12 experts participated in the interviews, could also answer “I have no opinion.” To ensure the complete-
yielding a response rate of 55%. Table 1 provides an overview ness of the factors, we asked one expert from the interviews for
of our interview partners. feedback on the survey. Following Hasson et al. (2000) and Pare
We used a semistructured interview protocol with open-ended et al. (2013), eight independent researchers, who are familiar
questions that was structured in accordance with the research with Delphi and survey studies, pretested the survey. They pro-
questions (see Appendix A). While the core themes of the inter- vided valuable feedback to improve comprehensibility and con-
view remained consistent, we adopted and extended the interview struct reliability.
guide over time (Patton 2015). The participants’ active involve- To build a panel of knowledgeable professionals and aca-
ment allowed us to explore specific potentials and barriers that demics, we relied on known experts and used the professional
they had observed in practice. The 12 interviews ranged from 40 network platform LinkedIn to source further panelists. Specifi-
to 131 min, with the average interview lasting 56 min. While cally, we searched for related keywords (“blockchain” AND
most interviews were conducted face-to-face, four interviews (“additive manufacturing” OR “3D printing”)) to identify poten-
were conducted via phone due to experts’ preferences and sched- tial panelists. We screened each profile individually to ensure
ules. Most of the project partners are based in German-speaking that individuals had a professional connection either to block-
countries, which significantly increased the likelihood for face- chain or to AM. The initial sampling process resulted in a list of
to-face interviews. We recorded the interviews, took notes, and 210 experts consisting of three equally sized subgroups: acade-
created memos to capture our first impressions immediately after mia, practice, and consulting.
each interview (Miles and Huberman 1994). Subsequently, we
summarized the main learnings from each interview and sent it Round 1
to the respective expert to avoid potential misinterpretation and We sent an invitation mail with a detailed overview of our study
to ask for additional insights. To clarify any remaining issues, and a brief explanation of the Delphi method to all 210 potential
we had additional follow-up conversations with one third of the experts. The letter also assured participants of their anonymity
experts. We also asked each interviewee for additional material and content confidentiality. We received 54 complete responses,
that could be valuable for our research (e.g. press releases, pro- resulting in a response rate of 26%. To improve rigor and ensure
ject whitepapers, project flyers, or online articles). We used the data quality, we included a screening question regarding the pan-
obtained 80 pages from five experts to strengthen our under- elist’s knowledge of blockchain, AM, and supply chain manage-
standing about the projects’ status and the discussed concepts. ment on a scale from (1) “Never heard of it” to (5) “I am an
The transcribed interviews were coded inductively, without a expert in this field.” We excluded one participant with below-av-
predefined coding structure. Immediately after the first interview, erage experience scores for both blockchain and AM, leading to
we began generating codes (first-order concepts) reflecting the a final panel of 53 experts for our first Delphi round. Most of
6 S Kurpjuweit et al.

Table 1: Overview of interviewed experts

Institution/
Expert Perspective industry Position Employees Country Blockchain in AM expertise

E1 Practice Energy company, Senior 40,000 Germany Cofounder of consortia project targeting
blockchain manager blockchain in AM
startup
E2 Practice Software company Managing 250–500 Germany Cofounder of consortia project targeting
director blockchain in AM
E3 Academia University Researcher – Germany Technical advisor for blockchain in AM
consortia project
E4 Academia University Researcher – Germany Blockchain in AM consortia project member
E5 Practice Blockchain startup Managing <10 Germany Cofounder of consortia project targeting
director blockchain in AM
E6 Academia, AM startup, Founder, <20 Italy Founder of blockchain in AM startup
practice university CEO,
Professor
E7 Practice Financial services Corporate 50,000 Germany Blockchain in AM consortia project member
strategy
manager
E8 Academia Blockchain Researcher 50–100 Germany Blockchain in AM prototype development,
research center execution of several blockchain industry
projects
E9 Academia Blockchain Researcher 50–100 Germany Execution of several blockchain industry
research center projects, conceptual work on AM case
E10 Academia University Researcher – Germany Blockchain in AM prototype development
E11 Practice IT law firm Partner, 1,500 Germany Blockchain in AM consortia project member
Lawyer
E12 Practice AM startup Founder, <10 Switzerland AM startup using blockchain technology
CEO

our panelists are located in Germany (46%), the United States Analysis
(17%), and the Netherlands (7%). Other experts come from Fin- Subsequent to the Delphi study, we analyzed our data to answer
land, Spain, Italy, Switzerland, France, Austria, India, Russia, our research questions. First, we calculated descriptive statistics,
Turkey, Canada, and New Zealand. such as mean or standard deviation, to rank the factors and to
evaluate the spread of opinions. Second, based on item score dis-
Round 2 tributions across the two rounds, we evaluated consensus and sta-
We invited all 53 experts for a second Delphi round. In total, 41 bility measures. Stability is the consistency of responses between
panelists completed the survey, yielding a dropout rate of 23%. consecutive Delphi rounds; consensus measures the agreement of
In the second round, we presented a ranked list of all factors and experts within each round (von der Gracht 2012). In this study,
included the group average for each item, allowing participants we consider consensus to have been reached when the coefficient
to objectively benchmark and possibly revise their opinion. We of variation (CV) is below 0.5 (English and Kernan 1976), or
also added free-text fields for the participants to explain opposing the interquartile range (IQR) is 1 or below (von der Gracht
views and provide arguments for their potential deviation from 2012). We achieved consensus for all items according to the CV,
the group consensus. Round-two panel demographics (Table 2)
reveal that experts are equally distributed among academics
(na = 14; 34%), practitioners (np = 14; 34%), and consultants Table 2: Round-two panel demographics
(nc = 13; 32%). For the purpose of subgroup analyses, we built
subgroups based on the experts’ key knowledge area. In total, 19
Academia Practice Consulting Total
experts (46%) assessed their blockchain expertise higher than
their AM expertise (BC background), while 11 (27%) rated their BC background 11 6 2 19
AM knowledge higher (AM background). Another 11 panelists AM background 1 3 7 11
(27%) rated their competences equally (mixed background). Both Mixed background 2 5 4 11
subpanel approaches are consistent with Okoli and Pawlowski Total 14 14 13 41
(2004), who suggest a subpanel size of 10–18.
Blockchain in Additive Manufacturing 7

while for 93% of all items both criteria indicate consensus. Like- According to our interview partners, blockchain technology
wise, decreasing standard deviations (i.e. negative convergence could serve as a safeguarding mechanism to protect the IP rights
rates) and nonsignificant mean differences between both Delphi (average panel rating = 4.10) of the originator in case of interor-
rounds indicate stability of opinions (von der Gracht 2012; ganizational data transfer, for example of construction plans, pro-
Roßmann et al. 2018). On average, standard deviations decreased duct specifications, or CAD files. Complementary to the
by 21% between rounds, signifying a strong convergence of protection of IP, a blockchain infrastructure enables the smart
expert opinions. For all items, we achieved stability for at least contract-based distribution of IP rights (3.98). For example,
one of the two criteria, while for 89% of all items both criteria selected AM service providers could be authorized to print a cer-
indicate stability. Overall, our Delphi data reveal convincing tain number of components from a copyright-protected CAD file.
results for both consensus and stability. Therefore, we did not Beyond the proactive safeguarding mechanisms for IP rights,
conduct a third Delphi round of rating. blockchain ensures the reactive enforcement of IP rights (3.98)
Finally, we explored the differences between the round-two which panelists rated just as important as the distribution of
subpanel groups and tested the expert agreement scores on sug- rights. In this light, based on the immutable record of all transac-
gested supply chain implications for significance. Since our Lik- tions and smart contracts, the originator can prove legal owner-
ert data do not necessarily follow a normal distribution, we used ship of the digital assets.
nonparametric approaches to test for agreement significance and Consider another practical example. Experts E2, E4, E5, and
possible significant differences between groups (Hair et al. E11 described their current projects developing a blockchain
2010). Specifically, we conducted Kruskal–Wallis tests (Kruskal platform that supports a secure AM-related IP transfer between
and Wallis 1952) which are an appropriate nonparametric proce- organizations. Due to the size of common 3D construction files,
dure for three or more independent samples. However, the they are transferred on conventional data exchange solutions,
Kruskal–Wallis approach indicates only the presence or absence referred to as “offchain.” However, an encrypted hash value,
of significant differences between tested subgroups. Therefore, uniquely representing the file, is stored on the blockchain. The
we used post hoc Wilcoxon–Mann–Whitney tests (Wilcoxon receiving node, for example the AM machine, can access the file
1945; Mann and Whitney 1947) to identify potential pairwise and verify its authenticity. Organizations might also use block-
subgroup differences (see Appendix B for subgroup test results). chain-backed smart contracts for licensing and authentication
One-tailed Wilcoxon tests indicate the significance of experts’ through digital signatures. Smart contracts ensure that the IP
agreement on formulated supply chain implications. receiver can only act according to the licensing contract, which
for example limits the number of product copies allowed.
Potentials and barriers for blockchain in additive Taken together, these potentials suggest that a blockchain-
manufacturing based solution could mitigate the production and distribution of
counterfeit products (4.37) or pirated copies, both posing a sub-
Potentials stantial problem to any digital good (Crosby et al. 2016; Mas-
The analysis of the expert interviews reveals 11 potentials that simino et al. 2018). In order to distinguish originals from
we cluster into four categories: (1) IP rights management, (2) unauthorized copies, the AM component is tagged, using for
lifecycle monitoring, (3) processes improvement, and (4) data example RFID chips, to ensure the distinctive and secure connec-
security. Table 3 shows our coding structure and provides illus- tion between the physical good and its respective blockchain
trative quotes from the interviews. Table 4 summarizes the record representation. The tag would then allow access to the
results of our quantitative Delphi study. digital record of all license information, printing details, dates,
IP rights management. Our study reveals that blockchain tech- and the transaction signatures of involved actors. Based on this
nology creates substantial potentials for the management of IP immutable information record, the user is empowered to identify
rights in AM. Blockchain provides a protective infrastructure that counterfeit parts in retrospect. Panelists rated this possibility as
enables both the secure transfer of sensitive construction data the most important IP management potential.
and IP rights authorization, and facilitates their legal enforce- Lifecycle monitoring. Our findings also indicate that block-
ment. These characteristics can also help to tackle the chronic chain can securely store all part-related lifecycle data, including
problem of counterfeit products, which is endemic in AM. One the raw material supply, various production stages, and product
of the experts emphasized the importance of securely transferring usage or consumption. The historic and present part-specific data
IP in a practical example: create a virtual mirror of the part, the digital twin, uniquely asso-
ciated with its physical counterpart. The digital twin of an AM
One of my co-workers develops motorcycle tuning parts part substantially increases product transparency (4.24). Stake-
that can be 3D-printed. Recently, he told me that he prints holders could obtain access to the consolidated and immutable
the parts at home and then sends them by mail. I couldn’t historic record of all imaginable part-related data, documents,
believe that as it contradicts the benefits of 3D printing. and transfers. Because of this part-specific record, stakeholders
When I asked him why he doesn’t just send the file, he are able to trace AM components through upstream production
responded, “if I send the file, it will be in all forums stages. Our experts rated traceability (4.41) as the third most
online and I will never sell it again.” [. . .] We have so important potential. In this sense, our panelists were confident
many technological possibilities today, but are increas- that blockchain holds the potential to significantly increase the
ingly limited in the use of these possibilities due to the overall visibility of respective AM supply chains. One expert
lack of a safeguarding mechanism. (E5) stressed the issue of transparency and traceability:
8 S Kurpjuweit et al.

Table 3: Results of expert interviews on “potentials”

Category Potentials Representative Quotes

IP rights IP rights protection “The designer gets proof that his file has been printed exactly 10 times and not 11 times –
management depending on the license. IP protection is a central issue of the platform.” (E1)
“You are able to protect intellectual property rights [. . .]” (E11)
“Within our project, we follow the idea of securing the IP rights of the originator.” (E5)
IP rights distribution “I am able to assign rights to it [the printer], for example, the right to print a part.” (E5)
“Our core aspect is the digital rights management approach, that is to say, the
implementation of a license management for 3D printing files.” (E4)
“Our fundamental idea is the licensing of printing processes. To enable a certain printer
through a secure element only printing when it is allowed to and when it has consumed a
license within the blockchain.” (E2)
IP rights enforcement “The originator is better in enforcing his IP protection.” (E2)
“Our solution results in huge advantages because it makes rights enforceable.” (E5)
“Beyond copyright protection, there is even law enforcement for the designer of a
product.” (E5)
Anti-counterfeit “I want to be able to distinguish: Am I holding an original part or a pirated copy in my
hands?” (E2)
“We want to avoid pirated copies.” (E2)
“[. . .] I do have a guarantee that this part has been printed legally – including a legitimate
license.” (E11)
Lifecycle Transparency “Our project objectives are to build a platform that [. . .] guarantees transparency [. . .]
monitoring along the product lifecycle.” (E4)
“I can map all this information across the whole product lifecycle within the blockchain.
And when holding the part in my hands, I know who printed it, at what time and with
which parameters.” (E2)
Traceability “The blockchain [. . .] simplifies the opportunity to give those objects an identity across
several systems – contributing to their traceability.” (E1)
“The originator is able to retrace who did what, when, and where.” (E2)
“The printing service provider [. . .] can now say: “I am able to produce components for
you that are secure and traceable.” (E5)
Audit trail “In a product liability case, I want to know who printed what.” (E2)
“In order to trust the printed product, it is necessary to have the audit trail including all
printing parameters and the CAD design.” (E1)
“Within the auditing process, one can easily check: When was it printed? From which
printer? Did a license exist?” (E4)
Process Process efficiency “The manufacturer will benefit from streamlining all these processes.” (E10)
improvement “Digitalization would result in significant process simplification – also for the issuer of
those certificates.” (E10)
“Of course, blockchain-based payments will come because you are able to substitute a
whole, separate process.” (E5)
Product development “I could take the data and the printed parts and use algorithms to learn how much I could
optimize my manufacturing data to improve the production process. In other words, I do
not use the digital twin just for the audit trail, but also as a kind of input to learn and
optimize the input parameters.” (E1)
“If I am able to collect lifecycle data of a product in this way, I might be able to use AI to
optimize this data in the future. So, I would have feedback for engineering without a
substantial degree of human interaction but rather an autonomous machine optimization.
And thus, a changing product lifecycle at the time.” (E5)
Data security Protection against data “In all probability, these data cannot be subsequently manipulated.” (E11)
manipulation “Project goals are [. . .] and of course the protection against points of attack such as
manipulation of CAD files.” (E4)
“I generate a hash value for this data and thus, I have a kind of fingerprint to validate that
it has not been manipulated.” (E5)
Detection of data theft “Today, a postman who is carrying a USB stick can just steal the data.” (E3)
“At the fair [. . .] people came to us and said: “We want to set up 3D printing but our
management asked about IT security and data protection.” (E5)
Blockchain in Additive Manufacturing 9

Table 4: Round-two Delphi results on “potentials”

Potentials Category RankR2 MeanR2 Δmean (mR2–mR1) SDR2 CR IQRR2 CVR2

Protection against data manipulation DS 1 4.51 0.24 0.64 29.9% 1 0.14


Audit trail LM 2 4.41 0.03 0.71 6.8% 1 0.16
Traceability LM 3 4.41 0.06 0.74 8.6% 1 0.17
Anti-counterfeit IP 4 4.37 0.03 0.83 16.5% 1 0.19
Transparency LM 5 4.24 0.15 0.73 8.1% 1 0.17
IP rights protection IP 6 4.10 0.11 0.86 20.4% 1 0.21
IP rights enforcement IP 7 3.98 0.23* 0.69 28.8% 0 0.17
IP rights distribution IP 8 3.98 0.16 0.83 16.1% 1.25 0.21
Detection of data theft DS 9 3.85 0.26 0.92 16.5% 1.25 0.24
Process efficiency PI 10 3.68 0.18 0.82 0.3% 1 0.22
Product development PI 11 3.32 0.24 0.85 11.3% 1 0.26

Notes: DS = data security, LM = lifecycle monitoring, IP = IP rights management, PI = process improvement, index R1 = Delphi round 1, index
R2 = Delphi round 2, m = mean, SD = standard deviation, CR = convergence rate (percentage change of SD between both rounds), IQR = interquartile
range, CV = coefficient of variation (SD/m).
*indicates statistical significance of mean difference between both Delphi rounds at the .05 level (p-value < .05).
Scale: (1) “Not important at all” to (5) “Extremely important”

For example in aviation, if there are any safety critical The continuous data recording at all stages of the product lays
problems then, of course, the OEM wants to know every- a foundation for (potentially) automated and iterative product
thing about this problem: Where do the individual blades development (3.32). As an example, consider the data collected
of our turbine come from? How were they produced? during the use of a printed part could improve design and func-
When were they installed? Of course, this also helps the tionality aspects in future product generations. To seize this
OEM to find out why something did not work. (E8) potential, however, the printed parts would need to be equipped
with sensors throughout the usage phase, raising privacy con-
Blockchain technology fosters transparency and traceability in cerns. It might take a while to settle the accompanying legal
an AM context and facilitates an exhaustive and sound audit issues, so this potential might not materialize in the near future.
trail (4.41), allowing stakeholders, especially AM service provi- Perhaps, due to the long-term character of this potential, our pan-
ders, to prove their compliance with the manufacturing process. elists rated it as the least important. Based on subgroup compar-
Data records might include construction files, simulations, isons, we find that practitioners (3.69) consider this potential
machine parameters, quality tests, certificates, and manufacturing significantly more important than panelists with an academic
data, such as temperature, energy, or pressure. Analogous to the background do (3.00).
discussion on IP rights management, our interviewees noted that Data security. The immutable nature of blockchain in combi-
an immutable record could serve as an important piece of evi- nation with the cryptographic hash function validation of the
dence in compliance and related liability claims: transaction creates a particularly high level of security, especially
protecting data against manipulation and detecting data theft.
We can have legal proof in form of a record at which According to our panelists, protection against data manipulation
point we have done what. And therefore, it’s not a protec- (4.51) is the most important potential. Within our subgroups,
tion, it’s more a defense in case we have legal issues or consultants in particular consider this potential as extremely
problems with someone. (E12) important (4.85).

Process improvement. Digitally mapping all part records can If I hash a data set and store the hash on the blockchain,
speed up or even replace paper-based verification and authentica- it is stored there for all eternity. Nobody, no third party, is
tion procedures. The removal of procedural bottlenecks might able to manipulate it. [. . .] That’s the immutability of the
result in more efficient processes (3.68). For instance, regulators blockchain. I store the hash and can be sure that nobody
and auditors could become part of the blockchain network, con- is going to change it. (E1)
siderably reducing manual effort, as one expert points out:
Although the protection of highly sensitive AM construction
files or printing parameters is considered especially relevant, our
The advantage is that the regulator is able to see every panel also recognizes the detection of data theft (3.85) as valu-
blockchain transaction in real-time. [. . .] Today, all these able. Interestingly, panel members with a stronger AM back-
processes are very paper-intensive and manual. Thus, we ground (4.50) rated data theft detection significantly higher than
conduct transactions, consolidate them and subsequently experts with a particular blockchain background (3.58). Recent
send them to the regulator. The regulator then needs to demonstrations of digital AM vulnerability (Belikovetsky et al.
review them. That’s very time-consuming. (E7) 2016) led the AM community to shift its focus from the
10 S Kurpjuweit et al.

technical limitations of 3D printers to data security. As another (3.71). The blockchain protocol itself requires a bundle of related
expert reported: technologies, including supplementary offchain storage, transfer
solutions, sensor networks, as well as other hardware and network
Two or three years ago, the AM community was very con- components in order to build a sufficient solution for AM.
cerned with technical issues, such as process stability, costs In the same context, experts frequently mentioned missing
[. . .] of course, these are still important issues but especially standardization (3.88) for the blockchain technology as an expla-
IT security concerns have become increasingly important. nation for integration issues in the firm’s IT infrastructure and
[IT security] already plays a big role in other industries, but within the blockchain system itself. However, standards for pro-
it was not so relevant for 3D printing a few years ago. I tocols and interfaces are required to facilitate adoption of block-
think the situation has changed since then. (E5) chain in AM and to gain operational benefits. The following
statement illustrates this diverse technological landscape:

Barriers Ideally, every transaction, its modeling, and distribution


In the interviews, we asked our experts about the challenges they across the network should be standardized. Ethereum and
have faced in their blockchain projects, resulting in a list of 24 bar- Hyperledger built their own standards and basically, every
riers. To categorize the barriers, we drew from the TOE frame- blockchain has its own. The question is: Will there ever be
work, developed by Tornatzky and Fleischer (1990) and widely a universal blockchain? (E2)
used to explain the adoption of new information technologies.
Consistent with the framework, we adopt the three main dimen- In addition, moving beyond issues regarding the integration of
sions—technology, organization, and environment—to categorize blockchain technology, a set of barriers is linked to the setup and
our barriers. However, we followed Chan and Chong (2013) and configuration of the blockchain (3.27). The solution design and
extended the TOE structure by a relational dimension, as we exam- intended use influence the technical parameters of the implementa-
ine interorganizational technology adoption. Table 5 summarizes tion. Configuration refers not only to the decision among public,
and clusters the identified barriers, with representative quotes from private, permissioned, or permissionless blockchains; adopters also
interviews. Descriptive Delphi results can be found in Table 6. need to take several blockchains (e.g. Ethereum, Hyperledger,
Technological barriers. We find several technological barriers Multichain) and consensus algorithms (e.g. proof of work and
ranging from compatibility-driven integration problems for the proof of stake) into consideration. To operate a blockchain, firms
firm to technical challenges when setting up a productive block- have to collect and prepare all the required input data (3.37).
chain instance to security issues related to interfaces to other While sensor networks generate and transfer data automatically,
devices and systems. other input devices require recording effort from the companies.
Four of the five most critical technological barriers are the in- Subsequent to its collection, the data need to be prepared in a way
tegration in IT landscape, interoperability, the blockchain archi- that is suitable for its blockchain input and storage. Our experts
tecture, and standardization, all of which relate to challenges of evaluated this barrier as slightly more challenging than the config-
developing and integrating a compatible blockchain solution. uration of a blockchain solution for the AM ecosystem. In particu-
Blockchain, as a technology in a very early stage, might explain lar, the privacy (3.66) and performance (3.54) are barriers that are
the panel’s emphasis on more general integration issues. One of closely linked to configuration decisions. Especially, public block-
the key challenges is the integration of the blockchain into an chains are frequently associated with inadequate performance,
existing organizational IT landscape (3.76). Processes have to be specifically regarding scalability, transaction speed, and through-
established or aligned and interfaces created to communicate put. They often pose a concerning lack of privacy, due to their
with internal systems, such as ERP or PLM. Many of our panel inherent public information access structure:
experts stress such integration issues:
There are many companies focusing on blockchain tech-
The first challenge for me was to be technically able to nology, but due to privacy issues, public blockchains are
connect our firm to this blockchain platform. So, it was out of the question. [. . .] Privacy is the make or break cri-
not compatible with any types of explorers, machines or terion. (E4)
computers we are using. [. . .] It is a very operational
problem, but I think it is a major problem for big corpora- Although the blockchain validation and consensus mechanisms
tions. (E12) and all information stored on the blockchain are designed to be
highly secure, the input data and crucial interfaces to other
The greatest barrier, interoperability (3.95), refers to the missing devices and systems create potential vulnerabilities of the system,
compatibility among blockchain systems. A broad variety of block- making the lack of security a formidable barrier (3.80). One pan-
chain protocols, for example using different validation or consensus elist expressed his concerns pointedly:
mechanisms, emerges frequently and in parallel. The lack of inter-
operability between blockchain instances might prevent firms from The validity of the collected data is important and hard to
participating in multiple blockchain networks, for example repre- verify. [. . .] Blockchain can make the data immutable but
senting different business units of the company, or migrating from garbage in is garbage out. (Delphi panelist)
one chain to another. Considering another level below the IT land-
scape and the different blockchain systems, we find issues within Focusing on the interfaces, the panel considers primarily pro-
surrounding technologies that create the blockchain architecture duct tagging, offchain data storage and transfer, and the
Blockchain in Additive Manufacturing 11

Table 5: Results of expert interviews on “barriers”

Category Barriers Representative Quotes

Technological Blockchain performance “At the moment, there are too many disadvantages of public blockchains: scalability,
transaction speed, transaction volume.” (E5)
“Scalability is a big issue. For public blockchains, it may is the core problem.” (E4)
Blockchain “Of course, a big challenge is the technology selection. In other words, on which
configuration blockchain do we agree?” (E4)
“A crucial question is: Which blockchain?” (E2)
Blockchain architecture “From my perspective, the biggest technical challenge is the need of integrating so many
different components when developing a blockchain solution – such as modern
cryptography, p2p communication, file storage, IoT, and others.” (E1)
Integration in IT “The effort to use our platform strongly depends on individual requisites, framework
landscape conditions, processes and IT systems. We will need good interfaces to ensure simple
functionality and a reduced implementation effort.” (E5)
“When digging deeper into the process, you will notice that there are several file
conversions, changing formats, questions of documentation. [. . .] The whole process is
incredibly complex.” (E11)
Interoperability “A further issue is interoperability. You want to increase the benefits of blockchains by
improving cooperation.” (E4)
“Hopefully, somebody will find a solution for interoperability until then. But that’s not too
easy.” (E5)
Standardization “It’s all about standardization. [. . .] That’s the challenge. [. . .] What do we need to do? A
lot of things. But first of all, this [standardization]. It’s not useful to reinvent the internet a
seventh time and in five different shades. Initially, it’s important to say: ‘The thing is
called blockchain.’ Then, we might have different forms. But first, we need to define the
core of all these forms properly and collectively.” (E3)
Privacy “Privacy is the sticking point. [. . .] A big challenge [. . .] also for consortia blockchains.”
(E4)
“Everybody is concerned about privacy rights now.” (E6)
Security “[. . .] Then I have a high-secure blockchain but the connection [to other systems and
hardware] builds the gateway for attacks. [. . .] Then you have an immutable information
in your blockchain, but may it’s incorrect.” (E5)
“We often call this [problem] a media interruption: Everything that has been stored on the
blockchain can be seen as securely recorded. But, obviously, the blockchain does not
safeguard the way data has been fed in.” (E11)
Handling of further IP “[. . .] of course, the printing service provider’s human interaction with the [originator’s]
processing file is an attack vector. But in most cases, this is necessary because the printing service
provider needs to optimize the file to guarantee its printability.” (E5)
Offchain data storage “The printing files are huge. You cannot store them on the blockchain. You have to ensure
and transfer the connection between the files and the blockchain-based license via a hash value.” (E2)
“You could also imagine a central cloud. However, this would be a tradeoff related to the
decentralized nature of blockchain.” (E4)
Product tagging “For tagging, there is a broad range of solutions, for example, RFID, barcodes, or sputter
coating [. . .] I want to have a simple solution to access the information and to identify the
part.” (E5)
“From a technical perspective, lots of effort is necessary. Also, the imprinting of RFID
which has its own risks.” (E4)
Data collection and “Usually, the main difficulty is to gather the data and prepare it to be fed into the
preparation blockchain. Let’s not deceive ourselves: The printer needs sensors, [. . .] the truck that
transports components needs to feed data into the blockchain. [. . .] The presumed process
completeness is nice, but I have to consider the technical perspective: Is all of this
feasible? How do we get all the relevant data?” (E8)

Continued.
12 S Kurpjuweit et al.

Table 5: (Continued)

Category Barriers Representative Quotes

Organizational Implementation costs “Not every company or small consortium wants to build its own blockchain and deal with
technology overhead.” (E4)
“Probably, it’s just a question of time and a financial issue. It’s not a cheap technology.”
(E11)
User resistance “Most businesses don’t like Ether and Bitcoin. They are not able to map it within their
accounting structures and the volatility is too high.” (E1)
“Then we have user acceptance. Last year, we ran a survey at the fair. Especially for
medium-sized companies the topic is far away.” (E4)
Lack of top-management “It’s not possible to provide a precise business case. And for most decision makers that’s
support really hard.” (E1)
“If it grows, you will need the top management support. [. . .] But the top management
always wants to have a business case. Only very few make it without a business case.”
(E1)
Lack of technical “Especially when starting it, the problem in a project like this, also in our company,
expertise honestly, is that there will be questions like: ‘Blockchain? Can you eat it? What is this?’”
(E2)
Lack of supportive “The next thing is ecosystem innovation. You have to share things and ideas. That’s in
culture total contrast to a top manager’s behavior that is based on working with NDAs, benefiting
from information asymmetries [. . .] and patents. So, distributed ownership and ecosystem
innovation find themselves in complete contradiction to every type of leadership culture
and mindset. [. . .] Not openly addressing these topics [blockchain] is a substantial
hurdle.” (E1)
Environmental Labor market “Basically, the blockchain developers are in incredibly high demand.” (E1)
availability “Specifically, smart contract developers are very rare.” (E4)
Speed of technological “It is very difficult. As soon as you have selected a technology, there could be a new,
developments better approach.” (E5)
“This dynamic environment is very challenging. There are other words than ‘challenging,’
but ‘challenging’ may be the most neutral one.” (E3)
Regulatory conditions “The legal conditions are not suitable. We are facing this currently.” (E2)
“Generally speaking, I think that regulatory aspects and associated uncertainty are big
hurdles for companies.” (E5)
Relational Cooperation “You have to involve different stakeholders. Generally speaking, different stakeholders
commitment worldwide, in different legal areas, in different cultures. We have all aspects that make it
difficult in every way.” (E9)
Coordination complexity “Based on a broad range of partners, there is a certain coordination complexity.” (E5)
“We really have a substantial variety of different stakeholders.” (E10)
Conflicts of interest “Apparently, we have very different interests.” (E2)
“Do we manage to find a collective approach? Or will individual interests dominate?” (E5)
Governance “Obviously, governance is a challenge. Especially when the consortium grows.” (E4)
“It will be the task of a dedicated team to develop a governance system that defines
participation and influencing rights. [. . .] That could be very complicated.” (E5)

handling of further IP processing as critical for data security. basic problem is that it [product tagging] depends on the
Product tagging is the application of tags for the unique identifi- individual case. A standardized solution cannot exist. (E5)
cation of parts and components. Tags include RFID chips, near-
field communication (NFC), or barcodes. Bridging the physical As CAD construction files can exceed the storage capacity of
and digital worlds, they distinctively link physical objects with almost all blockchain solutions, most of the data need to be
their digital twin record on the blockchain. However, our panel stored and transferred offchain (3.32). In most cases, only the
considers secure and cost-efficient product tagging (3.61) as a file’s hash value is stored on the blockchain to link the file.
challenge, since no standardized tagging solution is available: However, selection, alignment, and integration of suitable off-
chain data exchange systems remain critical. A related critical
In my eyes, the biggest challenge is the connection interface raising security concerns is the handling of further IP
between the physical and the digital worlds. [. . .] The processing (3.70). Since CAD files frequently require
Blockchain in Additive Manufacturing 13

Table 6: Round-two Delphi results on “barriers”

Category Barrier RankR2 MeanR2 Δmean (mR2–mR1) SDR2 CR IQRR2 CVR2

Technological Interoperability 1 3.95 0.01 0.63 32.1% 0 0.16


Standardization 2 3.88 0.07 0.71 29.3% 1 0.18
Security 3 3.80 0.17 0.71 37.2% 1 0.19
Integration in IT landscape 4 3.76 0.08 0.66 32.1% 1 0.18
Blockchain architecture 5 3.71 0.07 0.72 24.7% 1 0.19
Handling of further IP processing 6 3.70 0.13 0.69 37.6% 1 0.19
Privacy 7 3.66 0.14 0.73 36.4% 1 0.20
Product tagging 8 3.61 0.02 0.80 27.3% 1 0.22
Blockchain performance 9 3.54 0.09 0.87 24.3% 1 0.25
Data collection and preparation 10 3.37 0.05 0.80 12.0% 1 0.24
Offchain data storage and transfer 11 3.32 0.18 0.88 5.5% 1 0.26
Blockchain configuration 12 3.27 0.14 0.63 37.6% 1 0.19
Organizational Lack of technical expertise 1 4.05 0.12 0.84 10.9% 1 0.21
Lack of supportive culture 2 3.83 0.25 1.02 17.7% 2 0.27
Lack of top management support 3 3.67 0.12 0.98 12.2% 1 0.27
User resistance 4 3.58 0.17 0.90 5.5% 1 0.25
Implementation costs 5 3.48 0.05 0.68 22.1% 1 0.20
Environmental Labor market availability 1 4.12 0.22 0.71 19.1% 1 0.17
Regulatory conditions 2 4.02 0.08 0.69 33.6% 0 0.17
Speed of tech. developments 3 3.88 0.03 0.64 32.3% 0 0.17
Relational Governance 1 4.10 0.20 0.58 32.4% 0 0.14
Coordination complexity 2 3.90 0.08 0.66 31.9% 1 0.17
Cooperation commitment 3 3.66 0.01 0.73 20.2% 1 0.20
Conflict of interests 4 3.59 0.03 0.97 0.3% 1 0.27

Notes: R1 = Delphi round 1, index R2 = Delphi round 2, m = mean, SD = standard deviation, CR = convergence rate (percentage change of SD
between both rounds), IQR = interquartile range, CV = coefficient of variation (SD/m).
Scale: (1) “Not challenging at all” to (5) “Extremely challenging”.

postmodeling optimization before printing, external printing ser- deliberately pushing the implementation process, is key for a
vice providers need to access the construction file in order to successful implementation. However, the lack of a supportive
optimize it for subsequent manufacturing. Even though block- organizational culture can hinder the interorganizational develop-
chain technology and sufficient offchain storage systems might ment of AM designs:
be able to resolve or at least mitigate this issue, our experts are
worried that there is no longer end-to-end protection of the origi- From my experience cultural change is the most difficult
nator’s IP. task to accomplish in this type of projects. Many people will
Organizational barriers. In contrast to the blockchain-specific be reluctant on sharing their 3D models. (Delphi panelist)
technological barriers, we further identify more general adoption
barriers on an organizational level. Among these barriers are im- The lack of general knowledge also leads to user resistance
plementation costs, technical expertise, organizational culture, (3.58). The negative connotation of cryptocurrencies as well as
user resistance, and top management support. According to our the intangibility of major potentials such as effective IP manage-
panel, blockchain implementation is dependent on and partially ment and improved data security might raise skepticism. At the
constrained by organizational processes, structures, responsibili- bottom line, the adoption depends on top managers’ financial
ties, and interdependencies. Organizational barriers generate im- investment decisions. They require a business case, indicating
plementation costs (3.48) which, to our surprise, our experts expected returns. However, the novelty of the technology,
rated as the least important. Blockchain implementation requires volatility, and uncertainty of environmental conditions complicate
substantial technical expertise (4.05) of a variety of blockchain the development of accurate business plans. The associated risks
elements and their applicability. Very few members of an organi- highlight the importance of sufficient top management support
zation possess the expertise to understand and communicate the (3.67) for successful technology adoption. As one of our experts
benefits of blockchain in AM. Underpinning the importance of emphasized:
this know-how, our Delphi panelists consider a lack of technical
expertise as the most formidable organizational barrier. Technol- They [the managers] need to be willing to push a project
ogy adoption initiatives usually also face challenges on a more and to invest – even without a business case. All parame-
personal level. A supportive organizational culture (3.83), char- ters hinder the business case calculation. There are too
acterized by employees and managers taking risks and many unknown factors. (E1)
14 S Kurpjuweit et al.

Environmental barriers. In extension to the organization and engineers, CAD designers, logistics service providers, OEMs,
its internal processes and structures, the business environment raw material and component suppliers, and customers. The panel
also evokes particular barriers to new technology adoption. Our stressed that cooperation commitment (3.66) to initiate platform
experts find the general speed of technological developments, development, maintenance, and operations is a critical but over-
regulatory conditions, and the availability of skilled workers on looked challenge, especially in projects without an external plat-
the labor market, to constrain the development of blockchain- form provider. Relational issues among stakeholders are closely
based AM platforms. The blockchain technology is subject to related to project governance (4.10) decisions. Governance is
continuous change, and rapid technical developments and disrup- about establishing the right structures to formally manage a
tions. New blockchain protocols, governance structures, consen- diverse set of stakeholders. Important governance decisions
sus mechanisms, or use cases emerge almost every day. Firms evolve around responsibilities to operate, maintain, and refine the
struggle to keep up with the speed of technological developments platform, access and designing rights, authorization procedures
(3.88), leading to uncertainty and delayed adoption decisions. of new participants, exclusion of existing users, cost structures,
One expert summarizes the problem: and distribution. Overall, the panel considered the development
of appropriate interorganizational governance structures as the
All these technological pillars that one needs to integrate. most challenging relational barrier. Several experts highlighted
There is a high level of dynamics [. . .] the technology is the difficulty of establishing efficient governance mechanisms:
developing rapidly. You have to be able to integrate sev-
eral technologies, to conduct research, to understand Governance is a big challenge. Nobody can provide an
recent developments and you have to be open for substitut- answer to that. There are existing models [. . .] but since
ing already adopted technologies because there is so much there is no active multi-stakeholder platform on the mar-
going on right now. It’s extremely complicated to screen ket, nobody has been able to answer this question. (E1)
the flood of new developments and innovations. (E1)
The high degree of heterogeneity among the AM ecosystem
Regulatory authorities also seem to have trouble keeping up stakeholders can result in conflicting interests (3.59), amplifying
with technological developments. Regulatory uncertainty can result the importance of governance. For instance, while an external
in delayed or revoked adoption. Important issues yet to be covered designer may worry about IP protection, buyers emphasize poten-
by regulatory bodies include digital signatures, blockchain records tials related to traceability of parts. Especially, experts with partic-
as legal evidence, and smart contracts. Consequently, our panel ularly strong blockchain knowledge among panelists consider the
considers regulatory conditions (4.02) as one out of only four bar- coordination of these interests as an important challenge (4.00).
riers with an average rating exceeding four. Nevertheless, every The network effect (Shapiro and Varian 1998) also applies to
blockchain solution mandates compliance with the current national AM platforms. An increasing number of firms using the platform
law. Firms, especially their technical developers, are cautioned to increase the direct value of IP management, lifecycle monitoring,
cooperate closely with legal advisors and public authorities to sort improvement of processes, and data security for participating
out individual issues regarding data protection, privacy, IP law, or organizations. Hence, blockchain platforms face a trade-off
capital market law in case of financial platform services. While between network effects and coordination complexity (3.90). In
regulatory conditions heavily depend on national legislation, AM the words of one expert:
supply chains can spread across many countries. Therefore, dis-
crepancies in national standards might impede international collab- Blockchain is a network infrastructure. What is interesting
oration and lead to national AM blockchain solutions, about networks? The beginning is hard. If nobody uses the
considerably limiting the technology’s potential: network, it’s worthless. (E10)

Of course, regulatory issues could become a risk. [. . .]


Obviously, it’s bad to have this uncertainty. (E4) Supply chain implications of blockchain in additive
manufacturing
In line with the importance of technical expertise, the recruit-
ment of these experts is the highest barrier due to the limited According to the expert interviews and the two-round Delphi rat-
availability of qualified workers on the labor market (4.12). ing procedure, the adoption of blockchain in AM affects several
Interestingly, consultants (3.71) rated this barrier significantly supply chain areas (Tables 7 and 8). Specifically, we derive two
less important than academics (4.36) and practitioners (4.31) did. distinct effect paths. First, we identify direct blockchain implica-
Skilled blockchain and smart contract developers seem to be tions for supply chain management and logistics in an AM con-
extremely rare and in high demand, as one expert pointed out: text (see Figure 2, effect (a)). These are created primarily by
integrating blockchain technology. In this sense, the affected AM
[. . .] there are some companies working on blockchain supply chain would, for example, become more transparent as all
use cases and they are desperately looking for people who transactions are traceable on the blockchain. Second, building on
are familiar with these topics – from a technical perspec- previous studies investigating logistics and supply chain manage-
tive. [. . .] Everyone is searching. (E4) ment implications from AM integration (see Figure 2, dashed
arrow), we present impact areas that are initially enabled by AM
Relational barriers. Blockchain adoption in AM involves mul- but potentially catalyzed by blockchain (Figure 2, effect (b)). For
tiple stakeholders, including AM service providers, construction instance, by addressing typical interorganizational AM limitations
Blockchain in Additive Manufacturing 15

Table 7: Results of expert interviews on “supply chain implications”

Supply chain
Categories implications Representative Quotes

Direct Supply chain “I think that supply chains become more transparent.” (E2)
blockchain visibility “The big advantages are [. . .] and supply chain transparency.” (E1)
effect on “[. . .] to trace the lifecycle of a product.” (E4)
AM Supply chain “When we are able to preserve privacy, this [blockchain in AM] will have a big impact on
supply digitalization supply chain digitalization and optimization.” (E1)
chains “I think that the supply chain will become more seamless and transparent. [. . .] Seamless in the
sense of having more data digitalized [. . .] and transparent and available.” (E2)
Supply chain “Our objective is to use blockchain technology to connect all these actors. To ensure fully
finance automated payments between these actors.” (E7)
“As a bank, we focus particularly on financing the supply chain. [. . .] Therefore, we are really
curious about having the blockchain solution to ensure supply chain financing better and at lower
risk.” (E7)
Shared factories “Our project focuses on the aspect of a shared factory.” (E1)
“Ideally, I want to reach a state where I develop a product and am able to say: ‘I need five of
these at a specific place.’ Then, the system will find the optimal solution based on my target
setting. [. . .] The system organizes itself.” (E5)
Indirect Logistics “[. . .] and so, saving logistics costs and time.” (E1)
blockchain performance “The OEM who wants to [. . .] save warehousing and transportation costs, have a reduced capital
effect on commitment and so on.” (E2)
AM Decentralized “Blockchain can serve as an enabler. It can lead decentralized manufacturing to industrial scale.”
supply manufacturing (E1)
chains “The product will be printed at the place of demand – without any transportation.” (E5)
Supply chain agility “Spare part information is generally available. I am faster.” (E3)
“Blockchain enables possibilities that AM brings to supply chains: We have increasing agility.”
(E10)
Supply chain “This will have a big impact on increasing supply chain flexibility.” (E1)
flexibility “In the future, supplier and supply chain flexibility are obvious potentials.” (E2)
Supply chain “The more possibilities you have and the less you depend on single sources, the greater the
resilience resilience and agility. You will be able to react better to problems.” (E5)
Shift of value “We have the customer, or the OEM, who wants to have a spare part. But he doesn’t want to
creation produce it himself, so an AM service provider will do it.” (E10)
“Finally, we will see a higher specialization of supply chain actors.” (E10)

(e.g. IP theft), blockchain simplifies secure outsourcing of AM In combination with AM, supply chain visibility is primarily
parts’ fabrication to AM service providers, fostering the AM-en- driven by the identified lifecycle potentials transparency, trace-
abled shift toward more decentralized manufacturing. To sum up ability, and audit trail, which have been rated comparatively
our findings, we develop a set of propositions for each path of high. Specifically, blockchain allows the creation of a part-
supply chain effects. specific record of product and production-related information
and grants access to all relevant stakeholders. Since each cre-
Direct blockchain effects ated block entails a timestamp, supply chain members see the
The first supply chain effect that emerged from our qualitative chronological sequence of events around a printed part. The
interview data suggests increasing supply chain visibility. Our immutable data are, however, not limited to informative pur-
Delphi panel considers visibility as the most important supply poses; it can also build the foundation for improving supply
chain implication of blockchain in AM (average panel rat- chain planning and decision making. As one of our interview
ing = 4.78). We follow broader definitions of supply chain visi- partner highlights:
bility, understood as a firm’s access to high-quality supply and
demand information including the possibility to trace products There is also potential for optimization. If there is supply
along the supply chain (Francis 2008). AM is not the only use chain transparency, I can see where the critical path lies,
case for which blockchain potentially increases supply chain visi- which supplier could cause problems. Then I recognize
bility. Several recent studies make similar predictions for interna- much earlier where I could get problems and where I have
tional trade, luxury goods, or medical products (Casey and to change and do something in my supply chain or my
Wong 2017; Toyoda et al. 2017). supply network. (E8)
16 S Kurpjuweit et al.

Table 8: Round-two Delphi results on “supply chain implications”

Category Supply chain implications RankR2 MeanR2 Δmean (mR2–mR1) SDR2 CR IQRR2 CVR2

Direct blockchain Supply chain visibility 1 4.78 0.09 0.42 40.2% 0 0.09
effect on AM Supply chain digitalization 2 4.37 0.07 0.62 22.8% 1 0.14
supply chains Supply chain finance 3 4.22 0.18 0.88 17.8% 1 0.21
Shared factories 4 4.08 0.14 0.66 29.9% 0.25 0.16
Indirect blockchain Decentralized manufacturing 1 4.44 0.42 0.71 42.1% 1 0.16
effect on AM Logistics performance 2 4.29 0.23 0.78 30.9% 1 0.18
supply chains Shift of value creation 3 4.13 0.23 0.76 29.5% 1 0.18
Supply chain agility 4 3.93 0.01 0.72 30.6% 0 0.18
Supply chain resilience 5 3.83 0.01 0.70 25.4% 0 0.18
Supply chain flexibility 6 3.76 0.06 0.80 27.4% 1 0.21

Notes: R1 = Delphi round 1, index R2 = Delphi round 2, m = mean, SD = standard deviation, CR = convergence rate (percentage change of SD
between both rounds), IQR = interquartile range, CV = coefficient of variation (SD/m).
Scale: (1) “Strongly disagree” to (5) “Strongly agree”.

Figure 2: Distinction between direct and indirect blockchain effects for the AM supply chain ecosystem.

Blockchain (BC)

RQ1: Potentials RQ2: RQ2:


and barriers (b) Indirect effect (a) Direct effect

Supply Chain Management

Decentralized manufacturing Supply chain visibility


Additive Logistics performance Supply chain digitalization
Manufacturing (AM) Supply chain capabilities Supply chain finance
Shift of value creation Shared factories
AM driven implications BC driven implications

Research context: AM supply chain networks

Our findings further suggest that blockchain in AM could All these certificates, for example, provenance-related cer-
accelerate supply chain digitalization (4.37). We understand sup- tificates, are mostly paper-based which are used by vari-
ply chain digitalization as the use of digital technologies to ous different people. Of course, digitizing the whole
improve supply chain structures, processes, and activities. A process would also mean a significant simplification of the
prominent illustrative example for the benefits of blockchain process for those issuing the certificates. (E10)
adoption in supply chains is the joint project of Maersk and
IBM. In international trade, many processes are still paper-based, Moreover, these digital processes and records could lay the
hindering the efficient flow of both information and goods groundwork for future big data analytics, business intelligence
(Wang et al. 2019). Before Maersk and IBM began tracing ship- tools, or new digital business models. Some of our experts envi-
ping and customs transactions using a blockchain infrastructure, sioned future blockchain-based 3D designer markets that would
they first needed to digitize all relevant supply chain information unite several AM supply chain actors by creating new digital
(del Castillo 2018). Since the platform infrastructure requires business models:
appropriate data input formats, the initial digitization of all rele-
vant documents (e.g. certificates, quality assurance results, and “The objective is to build a market that brings together
invoices) is a fundamental requisite for efficient platform opera- originators who can upload their designs and potential
tions. Beyond pure document digitization, the resulting digital customers [. . .] through this marketplace, the designer and
processes are increasingly automated, improving supply chain the factory can earn money.” (E1)
efficiency, reducing complexity, and thus driving operational
performance. Other blockchain use cases show similar benefits Blockchain-enabled potentials, focusing on the optimization
(Swan 2015; Tapscott and Tapscott 2017). Our interview part- and automation of process flows, also present a plethora of
ners emphasized the implied performance benefits of supply opportunities for supply chain finance (4.22). Supply chain
chain digitalization enabled by a blockchain environment: finance deals with “optimized planning, managing, and
Blockchain in Additive Manufacturing 17

controlling of supply chain cash flows to facilitate efficient sup- Our project focuses on the aspect of a shared factory.
ply chain material flows” (Wuttke et al. 2013, p. 773). Consider Objects could identify their manufacturing plant autono-
smart contracts as an example. They allow organizations to syn- mously, provide their own parameters and become pro-
chronize automated payments with the transaction of construction duced. [. . .] Another option would be to share capacities
plans and specifications between supply chain actors. Enabling in the case of gaps in your production schedule. These
peer-to-peer payments in an interorganizational AM setting, gaps are tokenized, offered and another firm could buy
blockchain-secured smart contracts would render financial inter- access to the machine. [. . .] customers can borrow a part
mediaries obsolete (Yuan and Wang 2016). Consistent with the of the plant. (E1)
basic principles of the Bitcoin blockchain, all interorganizational
monetary transactions would be validated by the network nodes, In conclusion, our findings indicate that blockchain in AM is
and potentially audited and verified by involved regulatory expected to directly affect supply chains in four major areas.
authorities. This could drastically reduce the duration of payment Based on our panel, the agreement mean for blockchain-enabled
processes and overall transaction costs. For instance, the British supply chain visibility, digitalization, finance, and shared factory
Barclays Bank used blockchain for trade transactions, conse- systems exceed the neutral scale point of three significantly
quently cutting the process time from 10 days down to only four (p < .001). In combination with our qualitative expert interview
hours (Kelly 2016). Similarly, in the AM supply chain, origina- results, we thus propose:
tors of 3D designs could receive payments immediately after
another party printed their design. New payment and business Proposition 1a: Blockchain in AM will increase supply
models, based on subscription fees or pay per print options, are chain visibility.
possible. Moreover, banks that are involved in an AM block-
chain network could finance assets such as AM machines or Proposition 1b: Blockchain in AM will drive supply
offer insurances. Since blockchain would immutably record all chain digitalization.
activities, risks decrease, as a blockchain expert from a leading
European bank highlighted:
Proposition 1c: Blockchain in AM will support supply
chain finance.
What is the classical supply chain case for us today? A
company comes to us and says, “I have ordered some
Proposition 1d: Blockchain in AM will enable shared
goods but I do not have the money yet. Can you finance
factory systems.
the whole thing for me?” Then he has to show us some-
thing, which proves that he has really ordered something
Indirect blockchain effects
and that the goods are on their way. Nowadays, we accept
Our results indicate that blockchain adoption not only affects
a copy of the order confirmation which is very risky for
supply chains in a direct way but also enables and catalyzes gen-
us, because the person could just Photoshop something.
eral effects of AM on supply chains. Blockchain helps to over-
The better we can estimate how risky a transaction is, the
come prominent barriers of AM, especially in the areas of IP
lower our risk costs are. Therefore, we want a transparent
rights management and data security (Kurfess and Cass 2014;
process [. . .]. Of course, the blockchain helps a lot
Gao et al. 2015; Yampolskiy et al. 2018). Consequently, AM
because I know for sure that the order was placed. (E7)
adoption may be facilitated with potential implications for sev-
eral supply chain areas. To structure our results, we use the
Looking even further into the future, blockchain could enable
framework of Durach et al. (2017), who argue that AM affects
shared factory systems (4.08). Interestingly, both practitioners
supply chains in four major areas: (1) supply chain structure, (2)
(4.38) and panelists with a particularly profound AM back-
logistics performance, (3) supply chain capabilities, and (4) cus-
ground (4.45) agreed on this supply chain effect significantly
tomer-centric business models. We found that our six inductively
more than other subgroups. We define a shared factory as a net-
identified effects match with these four clusters.
work of independent production facilities in which capacities
The most emphasized indirect effect of blockchain on supply
are exchanged flexibly on a monetary basis to maximize capac-
ity utilization of each factory. The idea of this concept origi- chains, decentralized manufacturing (4.44), relates to supply chain
nates from the sharing economy and is quite intuitive: 3D structure. Various studies argue that AM allows to locate produc-
tion facilities closer to the location of demand (Khajavi et al. 2014;
printers can have long idle times, which firms could use to print
Wagner and Walton 2016). As production becomes distributed,
parts for other organizations, eventually increasing the utiliza-
decentralized, and localized, the roles of IP originators (i.e. the
tion of the network and thus reducing the total amount of pro-
organization developing the AM part design) and physical manu-
duction assets needed. However, our experts mentioned that
facturers (i.e. the local AM part producer) diverge gradually, thus
companies are only willing to participate in such shared factory
making it necessary to exchange sensitive information frequently
networks if they can verify that the other party acted compli-
on an interorganizational level. In these settings, blockchain can
antly, for example by printing a part according to the right
remove corresponding uncertainties related to IP rights manage-
specifications, and that their designs are sufficiently protected.
ment and data security. Furthermore, the digital infrastructure may
These two aspects explain the value of blockchain for these
networks. The following statement illustrates a shared factory allow OEMs to monitor and verify the production process by creat-
scenario: ing an immutable audit trail. One expert puts it plain and simple:
18 S Kurpjuweit et al.

I think that blockchain can serve as an enabler, which Our findings indicate that OEMs could adapt their business
brings decentralized manufacturing to industrial scale. models by separating value creation from value delivery, result-
(E1) ing in a higher specialization of supply chain actors. Beyond out-
sourcing possibilities for production (e.g. to local AM service
To discuss the second effect, we subsume supply chain agility providers), the firms’ business models could become more cus-
(3.93), flexibility (3.76), and resilience (3.83) under the concept tomer-centric because blockchain facilitates the integration of
of supply chain capabilities. Prior work has found AM to posi- customers into the value creation process. The partial mitigation
tively affect supply chain capabilities (Rogers et al. 2016; Dur- of IP and data security threats enables the customer integration.
ach et al. 2017). For instance, several studies discuss that local Involving customers into the design or manufacturing process is
manufacturing and shorter lead times, in combination with AM’s a major potential of AM (Fawcett and Waller 2014; Bogers et al.
ability to produce smaller batch sizes, enable much faster 2016). However, this potential has only materialized for a small
responses to changing market demands or disruptions (Holm- range of products so far. According to our experts, design mar-
str€
om et al. 2010). Blockchain adoption for AM is expected to kets where consumers offer 3D designs are still in the niche and
strengthen the general positive influence of AM on supply chain remain unattractive for B2B purposes.
structures as, for instance, supplier switching in case of a disrup- In conclusion, our data indicate that blockchain catalyzes sev-
tion or of an extreme demand is facilitated. To illustrate, a block- eral supply chain implications of AM. Figure 2 illustrates our
chain-based consortia network, including several AM service results. Based on our panel, the agreement means for items
providers, would allow the seamless shift of orders from one related to structural supply chain effects, logistics performance,
AM supplier to another (supply chain flexibility). While we supply chain capabilities, and shift of value creation exceed the
understand supply chain resilience as the responsiveness to and neutral scale point of three significantly (p < .001). In combina-
the recovery from unexpected disruptions, the focus of agility tion with our qualitative interview data, we thus propose:
lies in the operational speed in form of short lead times, not lim-
ited to the case of disruptive incidents (Gligor et al. 2019). How- Proposition 2a: Blockchain will strengthen the impact of
ever, these capabilities show some conceptual overlap (Swafford AM on supply chain structure, resulting in more local
et al. 2006; Wieland and Wallenburg 2013). For instance, both and decentralized manufacturing.
flexibility and agility contribute to supply chain resilience
(Christopher and Lee 2004). While emphasizing the reaction to Proposition 2b: Blockchain will strengthen the impact of
problems in the regard of supply chain resilience, one intervie- AM on logistics performance, resulting in a reduction of
wee indicates these interlinked advantages: logistics costs.

The more possibilities you have and the less you depend Proposition 2c: Blockchain will strengthen the impact of
on single sources, the greater the resilience and agility. AM on supply chain capabilities, resulting in more
You will be able to better react to problems. (E5) flexible, resilient, and agile supply chains.

The third affected supply chain area is logistics performance Proposition 2d: Blockchain will strengthen the impact of
(4.29). AM facilitates local on-demand manufacturing, reduces AM on the supply chain value creation process, resulting
waste, and makes supply chains shorter which may translate into in a shift of value creation in terms of higher
improved logistics performance in terms of reduced transporta- specialization and business models that involve
tion and warehousing costs (Holmstr€ om et al. 2010; Rogers et al. customers and new actors in the value chain.
2016). Both the interviews and the Delphi results indicate that
these effects would be strengthened, as AM adoption would
become more attractive on an interorganizational level with DISCUSSION
blockchain removing important barriers.
Finally, our experts emphasize that blockchain provides the Theoretical implications
opportunity to rethink the roles of supply chain actors, initiating
a shift of value creation (4.13), and move toward more cus- Our study adds to the growing discourse on emerging technolo-
tomer-centric business models, as two experts explain: gies in operations and supply chain management. The topic has
been prominently discussed during the last decades; however, it
It’s incredibly hard to estimate how the roles are chang- experiences a resurgence especially in the wake of digitalization
ing. But blockchain provides the opportunity to completely and Industry 4.0 initiatives, and a growing number of technologi-
rethink these roles. Maybe the manufacturer’s reason for cal developments. Early works have already shown general posi-
existence will shift from production towards development tive effects of IT integration on supply chain capabilities (Rai
and designing. Like Apple. (E5) et al. 2006; Wu et al. 2006) and firm performance (Subramani
2004; Devaraj et al. 2007), while others have studied the adop-
The manufacturer is now able to say: “There is no need tion and supply chain implications of specific technologies, such
to produce the parts in-house any longer. I am now able as RFID, ERP systems, or e-procurement systems (Tajima 2007;
to outsource my production and focus on IP development.” Whang 2010; Ramkumar and Jenamani 2015). Scholars have
(E7)
Blockchain in Additive Manufacturing 19

lately turned their attention to new digital technologies, investi- most critical barriers in our empirical investigation, labor market
gating, for instance, big data analytics and business intelligence availability, missing governance, lack of technical expertise, and
(Waller and Fawcett 2013; Schoenherr and Speier-Pero 2015; (uncertain) regulatory conditions, relate to a firm’s internal capa-
Kache and Seuring 2017; Lee 2018). Among these digital tech- bilities, its environment, and its interorganizational relationships.
nologies, blockchain, serving as a complementary platform Despite their relevance, nontechnical barriers seem to have
infrastructure connecting the application domains of other novel received insufficient attention so far which is why we urge schol-
technologies, could play an important role. Our findings further ars to include those aspects in future studies. While we study
suggest that blockchain holds the potential to enable new digital blockchain in AM, we expect most of the barriers to be con-
business models in the near future. firmed in a variety of adoption environments (e.g. international
An important result of our study is that blockchain can trade and luxury goods), as many barriers do not seem to be
improve supply chain visibility in the context of AM. The identi- highly specific to the AM setting. However, the relative impor-
fied potentials audit trail, transparency, and traceability, which tance of some barriers is likely to depend upon contextual fac-
are among the five highest ranked potentials, facilitate the moni- tors. For instance, the blockchain configuration (e.g. private or
toring and controlling of printed parts throughout their lifecycle. public), the solution’s design, and some environmental barriers,
In combination with the most emphasized potential, protection such as regulatory conditions, might vary among industries or
against data manipulation, our work indicates that supply chain countries.
partners could obtain access to a secure chronological record of On a methodological note, we are convinced that the substitu-
AM production and product data. Correspondingly, the increased tion of the first open-ended question round with in-depth expert
supply chain visibility can build the foundation for improved interviews addresses prominent Delphi critique regarding overly
supply chain planning and decision making. Another important general, simplified first-round responses (Sackman 1975; Hsu
group of potentials relates to secure interorganizational IP man- and Sandford 2007). Especially, Delphi studies exploring new
agement. We identify the detection of counterfeit products as phenomena should consider this adjustment in research design.
particularly crucial. Resolving the risk of manifold IP rights vio- We also recommend a more rigorous approach to selecting,
lations can foster information exchange and collaboration in sup- defining, analyzing, and interpreting consensus and stability met-
ply chains (Kurpjuweit et al. 2018; Massimino et al. 2018), rics (von der Gracht 2012). Since the Delphi method relies on
possibly even enabling the future scenario of shared factory sys- the consensus-based expert agreement and the convergence of
tems where firms can flexibly access and offer production capaci- opinions (Dalkey and Helmer 1963; Bolger and Wright 2011),
ties worldwide. the use of appropriate measures for consensus and stability
This study further adds to the literature on AM and its impli- improves the validity of the method.
cations for supply chain management and logistics. The AM pro-
cess chain is characterized by sensitive digital assets in form of Managerial implications
CAD files and construction IP, which seem to be vulnerable to
manipulation, theft, and infringement (Kurfess and Cass 2014; Utilizing our ranking-type Delphi study to “guide future manage-
Belikovetsky et al. 2016), and thus require special safeguarding ment action” (Pare et al. 2013, p. 208), we provide a consoli-
mechanisms (Rogers et al. 2016; Yampolskiy et al. 2018). In dated set of prioritized potentials and barriers. Our findings can
this context, we provide empirical data supporting our sugges- serve as a reference point for practitioners and guide managers
tion that blockchain addresses these limitations through its considering blockchain adoption in AM. While the benefits of
inherent IP rights management and data security capabilities. combining both technologies are straightforward, we encourage
Thus, this study responds to Holmstr€ om et al. (2016) and potential blockchain adopters to focus on the identified barriers
Rogers et al. (2016), both calling for future work investigating and to evaluate these against anticipated potentials. Managers
potential opportunities for IP and data management within need to consider contextual factors, such as the respective indus-
interorganizational AM settings. Conclusively, we propose that try, the supply network, or the company itself, when contemplat-
blockchain adoption catalyzes AM effects on supply chains. ing implementation initiatives.
More specifically, the integration of blockchain and AM is While many technological barriers, including blockchain per-
expected to accelerate the structural supply chain shift toward formance, interoperability, and standardization activities, can be
decentralized and local manufacturing (Durach et al. 2017), explained by the technology’s immaturity, both the blockchain
increasing logistics performance (Wagner and Walton 2016; architecture and the integration into companies’ IT landscapes
Ryan et al. 2017), customer-centric business models (Bogers depend on firm-internal structures and the preferred solution
et al. 2016), and enhanced supply chain capabilities (Huang design. Consequently, we advise managers to drive internal IT
et al. 2013). process alignment, simplification, and digitization of processes
Finally, a major contribution of this study is the identification before adopting blockchain technology.
and ranking of 24 barriers to blockchain adoption. Even though Practitioners should build blockchain capabilities internally,
many of the barriers have been proposed anecdotally or concep- as both the sourcing of capable blockchain engineers from
tually, such as interoperability, security, performance, or missing the labor market (environmental barrier) and the lack of firm-
standards (Swan 2015; Tapscott and Tapscott 2017; Saberi et al. internal technical expertise (organizational barrier) seem essen-
2019), we substantiate those with empirical evidence, add new tial impediments to successful implementation. The people
barriers, and present a prioritized list of challenges based on dimension has often been emphasized for technology adoption
expert consensus. It is also important to note that prior studies in operations and supply chain management (Schoenherr and
are mainly concerned with technical barriers; however, the four Speier-Pero 2015; Kache and Seuring 2017) and also seems
20 S Kurpjuweit et al.

to be a crucial success factor in our setting. Since blockchain scholars investigating blockchain use cases to apply the extended
in AM requires stakeholders’ involvement and commitment, version of the TOE framework.
managers should not stick to an intra-organizational perspec-
tive, but rather take on a supply network perspective. This
requires a shift of mindset and corporate culture toward col- REFERENCES
laborative ecosystem development. Only adopters who initiate
this shift, take risks, and find the right partners will be able Barratt, M., and Oke, A. 2007. “Antecedents of Supply Chain
to build effective governance structures and align stakeholder Visibility in Retail Supply Chains: A Resource-Based Theory
interests. Perspective.” Journal of Operations Management 25(6):1217–
Finally, managers need to be aware that blockchain could 33.
transform AM supply chains. Blockchain adoption in AM most Beck, R., Czepluch, J.S., Lollike, N., and Malone, S. 2016.
likely has structural implications, such as shifts in value creation, “Blockchain – the Gateway to Trust-Free Cryptographic
or intensified decentralized manufacturing. Therefore, we urge Transactions”. 24th European Conference on Information
companies to examine outsourcing opportunities for printing Systems (ECIS), Istanbul, Turkey.
activities, and pay attention to the possibly changing roles and Belikovetsky, S., Yampolskiy, M., Toh, J., and Elovici, Y. 2016.
power constellations within their supply chains. “dr0wned – Cyber-Physical Attack with Additive
Manufacturing”. arXiv 1609.00133: 1–15.
Limitations and future research Ben-Ner, A., and Siemsen, E. 2017. “Decentralization and
Localization of Production: The Organizational and Economic
Despite the applied methodological rigor, this study has some Consequences of Additive Manufacturing (3D printing).”
limitations that need to be taken into account when interpreting California Management Review 59(2):5–23.
our findings and conducting future research. First, it is worth not- Benos, E., Garratt, R., and Gurrola-Perez, P. 2017. The
ing that our results reflect the respondents’ views. It is conceiv- Economics of Distributed Ledger Technology for Securities
able that some of them might have an economic interest in Settlement, Staff Working Paper 670. London: Bank of
presenting an overly optimistic picture of the reality (e.g. consul- England.
tants and vendors of blockchain applications). We carefully con- Berman, B. 2012. “3-D printing: The New Industrial
ducted the interviews to distill relevant information for our Revolution.” Business Horizons 55(2):155–62.
study. Second, given the early stage of the investigated projects Blechschmidt, B., and St€ ocker, C. 2016. How Blockchain Can
as well as blockchain’s low maturity and industrial penetration, Slash the Manufacturing “Trust Tax”. Teaneck, NJ:
our forward-looking propositions reflect the panelists’ expecta- Cognizant.
tions on how the future could look like. However, we mitigate Bogers, M., Hadar, R., and Bilberg, A. 2016. “Additive
this potential concern through a carefully composed panel of Manufacturing for Consumer-Centric Business Models:
experts and two subsequent survey rounds reaching consensus of Implications for Supply Chains in Consumer Goods
opinions. Manufacturing.” Technological Forecasting and Social
Third, our propositions on the effects of blockchain on supply Change 102:225–39.
chain management are on a more general level. While the Bolger, F., and Wright, G. 2011. “Improving the Delphi Process:
broader scope reflects the heterogeneity of our experts and the Lessons from Social Psychological Research.” Technological
explorative nature of our research design, we encourage future Forecasting and Social Change 78(9):1500–13.
research to examine specific impact dimensions, such as struc- Brenig, C., Schwarz, J., and R€ uckesh€auser, N. 2016. “Value of
tural or performance-related supply chain implications. On that Decentralized Consensus Systems-Evaluation Framework”.
note, future studies could advance our understanding of how 24th European Conference on Information Systems (ECIS),
and under which circumstances blockchain affects AM and sup- Istanbul, Turkey.
ply chain management. As a first step, future work could deduc- Cachin, C. 2016. Architecture of the Hyperledger Blockchain
tively test our inductively derived set of propositions. In Fabric, Working Paper July 2016. R€ uschlikon: IBM Research
addition, researchers might explore the effects of blockchain – Zurich.
adoption in logistics and supply chain management conducting Calignano, F., Manfredi, D., Ambrosio, E.P., Biamino, S.,
longitudinal studies, challenging our propositions. More specifi- Lombardi, M., Atzeni, E., Salmi, A., Minetola, P., Iuliano, L.,
cally, scholars are encouraged to investigate the contextual fac- and Fino, P. 2017. “Overview on Additive Manufacturing
tors affecting blockchain adoption in AM, including particular Technologies.” Proceedings of the IEEE 105(4):593–12.
industries, relational settings, or organizational factors. Examin- Cao, Q., Jones, D.R., and Sheng, H. 2014. “Contained Nomadic
ing specific technological implementations, meaning specific Information Environments: Technology, Organization, and
blockchain types or configurations, or specific AM processes Environment Influences on Adoption of Hospital RFID
might yield valuable insights to advance the field and increase Patient Tracking.” Information and Management 51(2):225–
practical relevance. 39.
Our findings suggest that blockchain as a technology that is Caridi, M., Moretto, A., Perego, A., and Tumino, A. 2014. “The
potentially most effective in interorganizational settings is likely Benefits of Supply Chain Visibility: A Value Assessment
to be affected by relational barriers. The establishment of strong Model.” International Journal of Production Economics
governance structures seems relevant. Therefore, we recommend 151:1–19.
Blockchain in Additive Manufacturing 21

Casey, M.J., and Wong, P. 2017. “Global Supply Chains Are Francis, V. 2008. “Supply Chain Visibility: Lost in Translation?”
About to Get Better, Thanks to Blockchain.” Harvard Supply Chain Management: An International Journal 13
Business Review Digital Articles 2017:2–6. (3):180–84.
Chan, F.T.S., and Chong, A.Y.L. 2013. “Determinants of Mobile Gao, W., Zhang, Y., Ramanujan, D., Ramani, K., Chen, Y.,
Supply Chain Management System Diffusion: A Structural Williams, C.B., Wang, C.C.L., Shin, Y.C., Zhang, S., and
Equation Analysis of Manufacturing Firms.” International Zavattieri, P.D. 2015. “The Status, Challenges, and Future of
Journal of Production Research 51(4):1196–213. Additive Manufacturing in Engineering.” Computer-Aided
Chen, D., Heyer, S., Ibbotson, S., Salonitis, K., Steingrımsson, Design 69:65–89.
J.G., and Thiede, S. 2015. “Direct Digital Manufacturing: Ghadge, A., Karantoni, G., Chaudhuri, A., and Srinivasan, A.
Definition, Evolution, and Sustainability Implications.” 2018. “Impact of Additive Manufacturing on Aircraft Supply
Journal of Cleaner Production 107:615–25. Chain Performance: A System Dynamics Approach.” Journal
Christidis, K., and Devetsikiotis, M. 2016. “Blockchains and of Manufacturing Technology Management 29(5):846–65.
Smart Contracts for the Internet of Things.” IEEE Access Gibson, I., Rosen, D., and Stucker, B. 2015. Additive
4:2292–303. Manufacturing Technologies: 3D Printing, Rapid Prototyping,
Christopher, M., and Lee, H. 2004. “Mitigating Supply Chain and Direct Digital Manufacturing. 2nd ed. New York:
Risk through Improved Confidence.” International Journal of Springer.
Physical Distribution and Logistics Management 34(5):388– Gioia, D.A., Corley, K.G., and Hamilton, A.L. 2013. “Seeking
96. Qualitative Rigor in Inductive Research: Notes on the Gioia
Christopher, M., and Ryals, L.J. 2014. “The Supply Chain Methodology.” Organizational Research Methods 16(1):15–31.
Becomes the Demand Chain.” Journal of Business Logistics Gligor, D., Gligor, N., Holcomb, M., and Bozkurt, S. 2019.
35(1):29–35. “Distinguishing between the Concepts of Supply Chain
Crosby, M., Nachiappan, P., Verma, S., and Kalyanaraman, V. Agility and Resilience: A Multidisciplinary Literature
2016. “Blockchain Technology: Beyond Bitcoin.” Applied Review.” International Journal of Logistics Management 30
Innovation Review 2:6–19. (2):467–87.
D’Aveni, R.A. 2018. “The 3D-Printing Playbook.” Harvard Hair, J.F. Jr, Black, W.C., Babin, B.J., Anderson, R.E., and
Business Review 96(4):106–13. Tatham, R.L. 2010. Multivariate Data Analysis. Upper Saddle
Dalkey, N., and Helmer, O. 1963. “An Experimental Application River, NJ: Prentice Hall.
of the Delphi Method to the Use of Experts.” Management Hasson, F., Keeney, S., and McKenna, H. 2000. “Research
Science 9(3):458–67. Guidelines for the Delphi Survey Technique.” Journal of
del Castillo, M. 2018. “IBM-Maersk Blockchain Platform Adds Advanced Nursing 32(4):1008–15.
92 Clients as Part of Global Launch”. Forbes, 2018. https:// Holmstr€om, J., Holweg, M., Khajavi, S.H., and Partanen, J.
www.forbes.com/sites/michaeldelcastillo/2018/08/09/ibm-mae 2016. “The Direct Digital Manufacturing (r)evolution:
rsk-blockchain-platform-adds-92-clients-as-part-of-global-launc Definition of a Research Agenda.” Operations Management
h-1/#4c19cd9068a4 Research 9(1–2):1–10.
Denzin, N.K., and Lincoln, Y.S. 2000. Handbook of Qualitative Holmstr€om, J., Partanen, J., Tuomi, J., and Walter, M. 2010.
Research. Thousand Oaks, CA: Sage. “Rapid Manufacturing in the Spare Parts Supply Chain:
Devaraj, S., Krajewski, L., and Wei, J.C. 2007. “Impact of Alternative Approaches to Capacity Deployment.” Journal of
eBusiness Technologies on Operational Performance: The Manufacturing Technology Management 21(6):687–97.
Role of Production Information Integration in the Supply Hsu, C., and Sandford, B.A. 2007. “The Delphi Technique:
Chain.” Journal of Operations Management 25(6):1199– Making Sense of Consensus.” Practical Assessment, Research
216. and Evaluation 12(10):1–8.
Durach, C.F., Kurpjuweit, S., and Wagner, S.M. 2017. “The Huang, S.H., Liu, P., Mokasdar, A., and Hou, L. 2013.
Impact of Additive Manufacturing on Supply Chains.” “Additive Manufacturing and Its Societal Impact: A Literature
International Journal of Physical Distribution and Logistics Review.” International Journal of Advanced Manufacturing
Management 47(10):954–71. Technology 67(5–8):1191–203.
Edmondson, A.C., and McManus, S.E. 2007. “Methodological Jiang, R., Kleer, R., and Piller, F.T. 2017. “Predicting the Future
Fit in Management Field Research.” Academy of Management of Additive Manufacturing: A Delphi Study on Economic and
Review 32(4):1246–64. Societal Implications of 3D Printing for 2030.” Technological
English, J.M., and Kernan, G.L. 1976. “The Prediction of Air Forecasting and Social Change 117:84–97.
Travel and Aircraft Technology to the Year 2000 Using the Kache, F., and Seuring, S. 2017. “Challenges and Opportunities
Delphi Method.” Transportation Research 10(1):1–8. of Digital Information at the Intersection of Big Data
Fawcett, S.E., and Waller, M.A. 2014. “Supply Chain Game Analytics and Supply Chain Management.” International
Changers-Mega, Nano, and Virtual Trends-and Forces that Journal of Operations and Production Management 37(1):10–
Impede Supply Chain Design (i.e., building a winning team).” 36.
Journal of Business Logistics 35(3):157–64. Kelly, J. 2016. “Barclays Says Conducts First Blockchain-Based
Feng, Q., and Shanthikumar, J.G. 2018. “How Research in Trade-Finance Deal.” Reuters Technology News. https://www.
Production and Operations Management May Evolve in the reuters.com/article/us-banks-barclays-blockchain/barclays-says-
Era of Big Data.” Production and Operations Management conducts-first-blockchain-based-trade-finance-deal-id
27(9):1670–84. USKCN11D23B
22 S Kurpjuweit et al.

Khajavi, S.H., Partanen, J., and Holmstr€om, J. 2014. “Additive Okoli, C., and Pawlowski, S.D. 2004. “The Delphi Method as a
Manufacturing in the Spare Parts Supply Chain.” Computers Research Tool: An Example, Design Considerations and
in Industry 65(1):50–63. Applications.” Information and Management 42(1):15–29.
Kim, D.B., Witherell, P., Lipman, R., and Feng, S.C. 2015. Pare, G., Cameron, A.F., Poba-Nzaou, P., and Templier, M.
“Streamlining the Additive Manufacturing Digital Spectrum: 2013. “A Systematic Assessment of Rigor in Information
A Systems Approach.” Additive Manufacturing 5:20–30. Systems Ranking-Type Delphi Studies.” Information and
Kothman, I., and Faber, N. 2016. “How 3D Printing Technology Management 50(5):207–17.
Changes the Rules of the Game Insights from the Patton, M.Q. 2015. Qualitative Research and Evaluation Methods:
Construction Sector.” Journal of Manufacturing Technology Integrating Theory and Practice. 4th ed. London: Sage.
Management 27(7):932–43. Rai, A., Ravi, P., and Nainika, S. 2006. “Firm Performance
Kruskal, W.H., and Wallis, W.A. 1952. “Use of Ranks in One- Impacts of Digitally Enabled Supply Chain Integration
Criterion Variance Analysis.” Journal of the American Capabilities.” MIS Quarterly 30(2):225–46.
Statistical Association 47(260):583–621. Ramkumar, M., and Jenamani, M. 2015. “Organizational Buyers’
Kshetri, N. 2018. “1 Blockchain’s Roles in Meeting Key Supply Acceptance of Electronic Procurement Services – an
Chain Management Objectives.” International Journal of Empirical Investigation in Indian Firms.” Service Science 7
Information Management 39:80–89. (4):272–93.
Kurfess, T., and Cass, W.J. 2014. “Rethinking Additive Richardson, D.A., de Leeuw, S., and Dullaert, W. 2016. “Factors
Manufacturing and Intellectual Property Protection.” Affecting Global Inventory Prepositioning Locations in
Research-Technology Management 57(5):35–42. Humanitarian Operations – a Delphi Study.” Journal of
Kurpjuweit, S., Reinerth, D., and Wagner, S.M. 2018. “Supplier Business Logistics 37(1):59–74.
Innovation Push: Timing Strategies and Best Practices.” Rogers, H., Baricz, N., and Pawar, K.S. 2016. “3D Printing
Research-Technology Management 61(2):47–55. Services: Classification, Supply Chain Implications and
Lee, H.L. 2018. “Big Data and the Innovation Cycle.” Research Agenda.” International Journal of Physical
Production and Operations Management 27(9):1642–46. Distribution and Logistics Management 46(10):886–907.
Liu, P., Huang, S.H., Mokasdar, A., Zhou, H., and Hou, L. Roßmann, B., Canzaniello, A., von der Gracht, H., and
2014. “The Impact of Additive Manufacturing in the Aircraft Hartmann, E. 2018. “The Future and Social Impact of Big
Spare Parts Supply Chain: Supply Chain Operation Reference Data Analytics in Supply Chain Management: Results from a
(SCOR) Model Based Analysis.” Production Planning and Delphi Study.” Technological Forecasting and Social Change
Control 25(13–14):1169–81. 130:135–49.
Lu, H.P., and Weng, C.I. 2018. “Smart Manufacturing Ryan, M.J., Eyers, D.R., Potter, A.T., Purvis, L., and Gosling, J.
Technology, Market Maturity Analysis and Technology 2017. “3D Printing the Future: Scenarios for Supply Chains
Roadmap in the Computer and Electronic Product Reviewed.” International Journal of Physical Distribution
Manufacturing Industry.” Technological Forecasting and and Logistics Management 47(10):992–1014.
Social Change 133:85–94. Saberi, S., Kouhizadeh, M., Sarkis, J., and Shen, L. 2019.
Mann, H.B., and Whitney, D.R. 1947. “On a Test of “Blockchain Technology and Its Relationships to Sustainable
Whether One of Two Random Variables Is Stochastically Supply Chain Management.” International Journal of
Larger than the Other.” Annals of Mathematical Statistics Production Research 57(7):2117–35.
18(1):50–60. Sackman, H. 1975. Delphi Critique. Boston, MA: Lexington
Massimino, B., Gray, J.V., and Lan, Y. 2018. “On the Books.
Inattention to Digital Confidentiality in Operations and SAMPL. 2017. “SAMPL Secure Additive Manufacturing
Supply Chain Research.” Production and Operations Platform”. https://sampl.fks.tuhh.de/fileadmin/user_upload/
Management 27(8):1492–515. projects/sampl/publications/SAMPL_Project_Overview_engl._
Miles, M.B., and Huberman, A.M. 1994. Qualitative Data 5.2017.pdf
Analysis. 2nd ed. Newbury Park, CA: Sage. Sasson, A., and Johnson, J.C. 2016. “The 3D Printing Order:
Miller, J.W., Ganster, D.C., and Griffis, S.E. 2018. “Leveraging Variability, Supercenters and Supply Chain Reconfigurations.”
Big Data to Develop Supply Chain Management Theory: The International Journal of Physical Distribution and Logistics
Case of Panel Data.” Journal of Business Logistics 39 Management 46(1):82–94.
(3):182–202. Schleich, B., Anwer, N., Mathieu, L., and Wartzack, S. 2017.
Min, H. 2019. “Blockchain Technology for Enhancing Supply “Shaping the Digital Twin for Design and Production
Chain Resilience.” Business Horizons 62(1):35–45. Engineering.” CIRP Annals 66(1):141–44.
Muir, M., and Haddud, A. 2018. “Additive Manufacturing in the Schmidt, C.G., and Wagner, S.M. 2019. “Blockchain and Supply
Mechanical Engineering and Medical Industries Spare Parts Chain Relations: A Transaction Cost Theory Perspective.”
Supply Chain.” Journal of Manufacturing Technology Journal of Purchasing and Supply Management 25
Management 29(2):372–97. (4):100552. https://doi.org/10.1016/j.pursup.2019.100552
Oettmeier, K., and Hofmann, E. 2017. “Additive Manufacturing Schoenherr, T., and Speier-Pero, C. 2015. “Data Science,
Technology Adoption: An Empirical Analysis of General and Predictive Analytics, and Big Data in Supply Chain
Supply Chain-Related Determinants.” Journal of Business Management: Current State and Future Potential.” Journal of
Economics 87(1):97–124. Business Logistics 36(1):120–32.
Blockchain in Additive Manufacturing 23

Shapiro, C., and Varian, H.R. 1998. Information Rules: A Supply Chain Design and Management.” Journal of Business
Strategic Guide to the Network Economy. Boston, MA: Logistics 34(2):77–84.
Harvard Business Press. Wang, Y., Han, J.H., and Beynon-Davies, P. 2019.
St€
ocker, C., de Jonghe, D., and R€ uther, M. 2018. “Blockchain “Understanding Blockchain Technology for Future Supply
2.0: Exchanging Value in the Physical and Digital World, Chains: A Systematic Literature Review and Research
Continuously and Simultaneously”. International Business Agenda.” Supply Chain Management: An International
Times. https://www.ibtimes.co.uk/blockchain-2-0-exchanging- Journal 24(1):62–84.
value-physical-digital-world-continuously-simultaneously- Whang, S. 2010. “Timing of RFID Adoption in a Supply
1656257 Chain.” Management Science 56(2):343–55.
Subramani, M. 2004. “How Do Suppliers Benefit from Wieland, A., and Wallenburg, C.M. 2013. “The Influence of
Information Technology Use in Supply Chain Relationships?” Relational Competencies on Supply Chain Resilience: A
MIS Quarterly 28(1):45–73. Relational View.” International Journal of Physical
Swafford, P.M., Ghosh, S., and Murthy, N. 2006. “The Distribution and Logistics Management 43(4):300–20.
Antecedents of Supply Chain Agility of a Firm: Scale Wilcoxon, F. 1945. “Individual Comparisons by Ranking
Development and Model Testing.” Journal of Operations Methods.” Biometrics Bulletin 1(6):80–83.
Management 24(2):170–88. Williams, B.D., Roh, J., Tokar, T., and Swink, M. 2013.
Swan, M. 2015. Blockchain: Blueprint for a New Economy. “Leveraging Supply Chain Visibility for Responsiveness: The
Sebastopol, CA: O'Reilly. Moderating Role of Internal Integration.” Journal of
Swanson, T. 2015. “Consensus-As-A-Service: A Brief Report on Operations Management 31(7–8):543–54.
the Emergence of Permissioned, Distributed Ledger Systems”. Williams, T., Maull, R., and Ellis, B. 2002. “Demand Chain
https://allquantor.at/blockchainbib/pdf/swanson2015consensus. Management Theory: Constraints and Development from
pdf Global Aerospace Supply Webs.” Journal of Operations
Tajima, M. 2007. “Strategic Value of RFID in Supply Chain Management 20(6):691–706.
Management.” Journal of Purchasing and Supply Wowak, K.D., Craighead, C.W., and Ketchen, D.J. Jr. 2016.
Management 13(4):261–73. “Tracing Bad Products in Supply Chains: The Roles of
Tapscott, D., and Tapscott, A. 2016. “The Impact of the Temporality, Supply Chain Permeation, and Product
Blockchain Goes Beyond Financial Services.” Harvard Information Ambiguity.” Journal of Business Logistics 37
Business Review Digital Articles 2016:2–5. (2):132–51.
Tapscott, D., and Tapscott, A. 2017. “How Blockchain Will Wu, F., Yeniyurt, S., Kim, D., and Cavusgil, S.T. 2006. “The
Change Organizations.” MIT Sloan Management Review 58 Impact of Information Technology on Supply Chain
(2):10–13. Capabilities and Firm Performance: A Resource-Based View.”
Tornatzky, L.G., and Fleischer, M. 1990. Processes of Industrial Marketing Management 35(4):493–504.
Technological Innovation: Reviewing the Literature. Wu, Y., Cegielski, C.G., Hazen, B.T., and Hall, D.J. 2013.
Washington, DC: National Science Foundation. “Cloud Computing in Support of Supply Chain Information
Toyoda, K., Mathiopoulos, P.T., Sasase, I., and Ohtsuki, T. System Infrastructure: Understanding When to Go to the
2017. “A Novel Blockchain-Based Product Ownership Cloud.” Journal of Supply Chain Management 49(3):25–41.
Management System (POMS) for Anti-counterfeits in the Post Wuttke, D.A., Blome, C., and Henke, M. 2013. “Focusing the
Supply Chain.” IEEE Access 5:17465–77. Financial Flow of Supply Chains: An Empirical Investigation
Treiblmaier, H. 2018. “The Impact of the Blockchain on the of Financial Supply Chain Management.” International
Supply Chain: A Theory-Based Research Framework and a Journal of Production Economics 145(2):773–89.
Call for Action.” Supply Chain Management: An Yampolskiy, M., King, W.E., Gatlin, J., Belikovetsky, S.,
International Journal 23(6):545–59. Brown, A., Skjellum, A., and Elovici, Y. 2018. “Security of
von der Gracht, H.A. 2012. “Consensus Measurement in Delphi Additive Manufacturing: Attack Taxonomy and Survey.”
Studies: Review and Implications for Future Quality Additive Manufacturing 21:431–57.
Assurance.” Technological Forecasting and Social Change 79 Yli-Huumo, J., Ko, D., Choi, S., Park, S., and Smolander, K.
(8):1525–1536. 2016. “Where Is Current Research on Blockchain Technology?
Wagner, S.M., and Walton, R.O. 2016. “Additive – A Systematic Review.” PLoS ONE 11(10):1–27.
Manufacturing’s Impact and Future in the Aviation Industry.” Yuan, Y., and Wang, F.Y. 2016. “Towards Blockchain-Based
Production Planning and Control 27(13):1124–30. Intelligent Transportation Systems”. 2016 IEEE 19th
Waller, M.A., and Fawcett, S.E. 2013. “Data Science, Predictive International Conference on Intelligent Transportation
Analytics, and Big Data: A Revolution that Will Transform Systems (ITSC), IEEE, 2016:2663–68.
24 S Kurpjuweit et al.

APPENDIX A

INTERVIEW INSTRUMENT

I. Introduction
▪ Please briefly introduce yourself and your current position.
▪ What topics are you focusing on in the context of blockchain and AM?
▪ Please tell me about your project addressing blockchain in AM.
▪ What are the objectives of this project?
II. Blockchain potentials in AM
▪ Based on these objectives: What are the advantages of blockchain in AM (compared to other technological solutions)?
▪ Why do you apply blockchain technologies (and not more conventional approaches)?
▪ Which stakeholders could benefit from a blockchain-based platform in AM?
▪ How much would these stakeholders benefit? What are their advantages?
▪ Which business models could arise in the context of blockchain in AM?
III. Blockchain challenges in AM
▪ What specific challenges did you face during your project?
▪ What are anticipated barriers for blockchain adoption within the AM ecosystem?
▪ What are the main technological (organizational, environmental, relational) challenges when integrating blockchain in AM?
▪ How would you estimate the relevance of these challenges?
IV. Supply chain implications of blockchain in AM
▪ How much could a blockchain-based platform influence supply chains?
▪ What implications for supply chain management do you see in the context of blockchain in AM?
▪ What are the implications for involved stakeholders in the AM supply chain?
▪ How much could blockchain facilitate industrial AM and its supposed supply chain impacts?
▪ How do companies react when illustrating the benefits of secure, decentralized manufacturing possibilities?
V. Conclusion and outlook
▪ Please draw a brief conclusion. What are the most important potentials of and the most challenging barriers to blockchain adoption
in AM?
▪ What were the major lessons for your firm from this project?
▪ What is already possible (i.e. ready for the market) today? What are dreams of the future?
▪ How would you estimate the future development of blockchain in AM?
▪ Do you want to add something to the answers?

APPENDIX B

SUBGROUP COMPARISONS

Table A1: Results of subgroup comparison “perspective”

Item MA (n = 14) MC (n = 14) MP (n = 13) Kruskal–Wallis H Significant Pairs

Potentials
Product development 3.00 3.29 3.69 6.65* AP*
Barriers
Labor market availability 4.36 3.71 4.31 8.48* AC*, CP*
Supply chain implications
Shared factory system 3.69 4.14 4.38 7.23* AP*

Notes: m = mean, H = Kruskal–Wallis test statistic (chi-squared distributed); Pairwise post hoc Wilcoxon–Mann–Whitney tests indicate significant
pairs: A = academic, C = consultant, P = practitioner.
*indicates statistical significance at the .05 level (p-value < .05), only significant items and pairs are included.
Blockchain in Additive Manufacturing 25

Table A2: Results of subgroup comparison “background”

Item mBC (n = 19) mAM (n = 11) mMB (n = 11) Kruskal–Wallis H Significant Pairs

Potentials
Detection of data theft 3.58 4.50 3.73 7.18* BCAM*, AMMB*
Barriers
Conflict of interests 4.00 3.45 3.00 8.32* BCMB*

Notes: m = mean, H = Kruskal–Wallis test statistic (chi-squared distributed); Pairwise post hoc Wilcoxon–Mann–Whitney tests indicate significant
pairs: BC = blockchain background, AM = AM background, MB = mixed background
*indicates statistical significance at the .05 level (p-value < .05), only significant items and pairs are included.

SHORT BIOGRAPHIES

Stefan Kurpjuweit (M.Sc., Karlsruhe Institute of Technology) worked as a research associate and PhD candidate at the Chair of
Logistics Management at the Swiss Federal Institute of Technology Zurich (ETH Zurich), Switzerland, and is now a Business Develop-
ment Manager at ABB. His research interests relate to open innovation, startup suppliers, and the impact of innovative technologies on
supply chains. He has published in scholarly journals such as International Journal of Physical Distribution & Logistics Management,
International Journal of Production Research, and Journal of Business & Industrial Marketing, as well as managerial journals such
Business Horizons, Research-Technology Management, and Supply Chain Management Review.

Christoph G. Schmidt (Ph.D., EBS University) is a postdoctoral researcher at the Chair of Logistics Management at the Swiss Fed-
eral Institute of Technology Zurich (ETH Zurich), Switzerland. His prime research interest relates to the use and effects of new tech-
nologies for supply chain management, and his work has appeared in Journal of Business Logistics, Journal of Supply Chain
Management, International Journal of Production Research, and other journals.

Maximilian Kl€ockner (M.Sc., Darmstadt University of Technology) is a research associate and PhD candidate at the Chair of Logis-
tics Management at the Swiss Federal Institute of Technology Zurich (ETH Zurich), Switzerland. He currently studies the adoption of
emerging technologies in operations and supply chain management, and implications of digitalization in logistics.

Stephan M. Wagner (Ph.D., University of St. Gallen) is a professor and holds the Chair of Logistics Management at the Swiss Fed-
eral Institute of Technology Zurich (ETH Zurich). Recent projects investigated the management of startups as suppliers, supplier innova-
tion, digitalization, as well as sustainability and ethics in operations and supply chain management. His research has been published in
management journals, such as the Academy of Management Journal or Journal of Management; operations management journals, such
as Journal of Operations Management or Journal of Supply Chain Management; methods journals, such as Organizational Research
Methods or Sociological Methods and Research; and managerial journals such as Interfaces or California Management Review.

Das könnte Ihnen auch gefallen