Conference PaperPDF Available

Abstract

Since the manufacturing industry is facing increasingly advancing digitalization, digital twins (DT) have become a popular means for integrating various actors' value creation using a smart product. DTs are information systems that connect the physical and virtual worlds. The design of DTs is time-consuming, expensive, and lacks appropriate prescriptive design knowledge for its development. Design principles (DP) represent a mechanism to codify design knowledge into prescriptive knowledge. However, the mostly abstract DPs are often difficult for practitioners to operationalize during software development projects, rendering the design knowledge difficult to access. The paper at hand addresses these issues by providing a reference model for DT development as a semi-abstract artifact. The model has been constructed by drawing on a literature review and empirical cases in the manufacturing industry. The reference model includes multiple adaptation mechanisms to ensure a flexible development of company-specific DTs.
COVER SHEET
This is the author’s version of the following source:
-Author’s Accepted Manuscript-
Users may only view, print, copy, download and text and data-mine the context, for the purposes of academic research. The
content may not be (re-)published verbatim in whole or in part or used for commercial purposes. Users must ensure that the
author’s moral rights as well as any third parties’ rights to the content or part of the content are not compromised.
Wache, Hendrik; Hübner, Eric Jan; Hönigsberg, Sarah; and Dinter, Barbara,
"Closing the Implementation Gap of Digital Twins" (2022). AMCIS 2022
Proceedings. 3.
https://aisel.aisnet.org/amcis2022/sig_sand/sig_sand/3
Closing the Implementation Gap of Digital Twins
Twenty-eighth Americas Conference on Information Systems, Minneapolis, 2022
1
CLOSING THE IMPLEMENTATION GAP OF
DIGITAL TWINS
Completed Research
Hendrik Wache
Chemnitz University of Technology
hendrik.wache
@wirtschaft.tu-chemnitz.de
Jan Eric Hübner
Chemnitz University of Technology
jan-eric.huebner
@wirtschaft.tu-chemnitz.de
Sarah Hönigsberg
Chemnitz University of Technology
sarah.hoenigsberg
@wirtschaft.tu-chemnitz.de
Barbara Dinter
Chemnitz University of Technology
barbara.dinter
@wirtschaft.tu-chemnitz.de
Abstract
Since the manufacturing industry is facing increasingly advancing digitalization, digital twins (DT) have
become a popular means for integrating various actors' value creation using a smart product. DTs are
information systems that connect the physical and virtual worlds. The design of DTs is time-consuming,
expensive, and lacks appropriate prescriptive design knowledge for its development. Design principles (DP)
represent a mechanism to codify design knowledge into prescriptive knowledge. However, the mostly
abstract DPs are often difficult for practitioners to operationalize during software development projects,
rendering the design knowledge difficult to access. The paper at hand addresses these issues by providing
a reference model for DT development as a semi-abstract artifact. The model has been constructed by
drawing on a literature review and empirical cases in the manufacturing industry. The reference model
includes multiple adaptation mechanisms to ensure a flexible development of company-specific DTs.
Keywords
Reference Model, Digital Twin, Information System Modeling, Design Principles.
Introduction
Today's world is continually shaped by advancing digitalization. Digitalization has a particular impact on
the manufacturing industry, to the extent that it is referred to as the fourth industrial revolution (industry
4.0, hereafter I4.0). It involves using advanced information technologies that connect industry and its
plants with the internet of things and services (Kagermann et al. 2013). I4.0 enables an increasing
convergence between the physical and the virtual space. Digital twins (DT) embody this convergence by
connecting physical assets (e.g., a production plant) and virtual components (e.g., a behavioral model of the
plant) in particular. DTs have a large amount of data, such as product specifications, production process
models, operational performance data, or even written knowledge. However, the numerous virtual models
that perform analyses and simulations, thereby generating descriptive, predictive, and prescriptive
knowledge, are the focus of DTs. DTs can be considered to be oriented towards collaborative value creation
in a network of actors surrounding a smart object/product, which enables the integration of (knowledge)
resources for mutual benefits (Beverungen et al. 2019).
A DT enables companies to achieve considerable operational added value and can become the central point
of contact for all data in the relevant plant by integrating it with existing operational systems such as
enterprise resource planning, customer relationship management, or product data management. However,
the added value of DTs is also accompanied by challenges. For example, high development costs and the
high maintenance effort make widespread implementation in organizations difficult. Developing a DT from
scratch is a time-consuming challenge, especially for smaller companies typical for the manufacturing
Closing the Implementation Gap of Digital Twins
Twenty-eighth Americas Conference on Information Systems, Minneapolis, 2022
industry. This challenge is complicated by company-specific requirements and a high level of development
complexity, often leading to development without sufficient conceptual planning. Against this background,
there is a particular lack of design-oriented studies for supporting the development of I4.0 systems in
general (Baiyere et al. 2020) and DTs, especially with guidance for developers in practice.
Although previous research described some aspects of DT development, there is a lack of synthesis of
findings to support development efforts. On the one hand, practical DT developments are taking place in
the industry. However, DTs are often considered too technical. Their integration into the business processes
and the interaction between DTs and the human workforce is not sufficiently considered, often leading to
merely prototypical implementations. On the other hand, design knowledge is emerging from the scientific
discourse on DTs. In the information systems (IS) discipline, this design knowledge is often captured in so-
called design principles (DP), which are an abstracted representation of the characteristics of an artifact
condensed into a prescriptive statement to be considered for the design. In recent years there has been
increasing discussion about the extent to which these DPs are practical (Chandra Kruse et al. 2016;
Schoormann et al. 2021), i.e., to what extent a system developer who wants to develop a DT, for example, is
properly supported by this abstract knowledge. Thus, a two-sided problem arises: In practice, there is a lack
of best practices to design DTs as an integral part of corporate value creation, and from the scientific side,
there is a lack of a knowledge transfer mechanism that makes the academic findings on DTs easy to
implement for practitioners.
In a multi-year design study on the development of a DT platform, we have shown that the developed DPs
were useful and helpful for knowledge transfer in the academic environment (Wache and Dinter 2021).
However, an "implementation gap" arose in using the DPs for practitioners in the project. This gap arose
because the DPs could not be operationalized by the various developers and were perceived purely as "food
for thought" during the implementation. Therefore, it is necessary to find a less abstract vehicle for the
design knowledge, which should be able to integrate the DPs, is universally valid, and yet is still close to
implementation to enable the translation of the knowledge by practitioners to the instance level. Therefore,
this work presents an artifact that compiles the best practices behind the development of DTs in such a way
that an application of the design knowledge can be made in different contexts by different developing
companies. For this purpose, reference modeling is suitable, as it simplifies the design process and the
actual development by bringing together best practices, recommendations, and knowledge from a research
field and making them usable (Becker et al. 2007). This paper synthesizes acquired knowledge from prior
DPs, a literature review, focus group interviews, and workshops to create a reusable conceptual artifact that
contains prescriptive knowledge about DT design and development. This leads to the following research
question: How should a DT reference model be built to make design knowledge available to IS scholars and
practitioners and close the implementation gap?
The remainder of the paper is organized as follows: After we described the conceptual foundations of our
research, the research design is outlined. We develop the reference model in the subsequent section,
followed by an evaluation. The paper concludes after a discussion of the results.
Conceptual Foundations
Digital Twins
The first notion of using a DT originated in product lifecycle management, which conceived this twin as a
virtual representation of a physical system. The DT is supposed to contain all data about and current status
data of its physical counterpart (Grieves 2014; Grieves and Vickers 2017). However, the concept of the DT
is now also used outside the field of product lifecycle management, for instance, in the manufacturing sector
(Grieves 2014; Shafto et al. 2012). According to Rosen et al. (2015), manufacturing systems will have to
operate more autonomously in the future to meet the emerging challenges. DTs are expected to help
overcome these challenges by bridging the gap between the real and the digital world and providing
comprehensive data across all stages (Glaessgen and Stargel 2012; Rosen et al. 2015). In this sense, DTs
can be understood either as enablers or subtypes of cyber-physical systems (Dietz and Pernul 2020). In
general, however, DTs are not just a collection of different digital artifacts but form a system in which these
artifacts are connected in a structured way and consequently contain meta-information (Rosen et al. 2015).
Several components need to be considered: a physical component, a virtual component, a service system,
the data, and a connection between all of them (Tao et al. 2019). The use cases for DTs are diverse, spanning
Closing the Implementation Gap of Digital Twins
Twenty-eighth Americas Conference on Information Systems, Minneapolis, 2022
3
several dimensions from product design and production to optimization and maintenance (Rosen et al.
2015; Tao et al. 2019), but are commonly focused on control, simulation, and monitoring, with data
becoming a central driver. Tao et al. (2019) discuss the use of DTs in value chains in which the DT provides
services to the customers. Therefore, the DT can be considered a platform to enable human engagement in
I4.0 and to support the value creation of multiple actors.
Abstract Design Knowledge and Reference Models
DPs represent prescriptive knowledge intended to aid practitioners in translating abstract concepts into a
more tangible form for practical application. However, this is often not sufficient because the development
of IT artifacts is a continuous, complex, and context-specific process (Chandra Kruse et al. 2016; Sein et al.
2011). Chandra Kruse et al. (2016) and Amabile et al. (1996) observe that the successful application of DPs
depends on the designers' knowledge base, resulting in something that could be described as an
implementation gap. This gap describes a distance between the proposed solution and available resources,
such as design knowledge, ranging from tacit knowledge among developers to design principles and
technical models. In a sense, the issue also relates to a problem of bridging between two experts, one
explicating his knowledge and the other applying it. To narrow this implementation gap, this work attempts
to propose a reference model for DT platforms, as designers commonly use such models as a basis for their
work during the design phase of IS. Reference models constitute a simplified or optimized representation
of a system, which enables the derivation of design proposals from an ideal concept (Rosemann and van
der Aalst 2007). Furthermore, our reference model follows the understanding of Becker et al. (2007) and
Rosemann and van der Aalst (2007) and is intended to apply to a class of abstract application areas and
thus exhibit both generality and adaptability. Such a generally applicable reference model has to be
transformed into a specific model for practical application in most cases (Fettke and Loos 2003). Therefore,
we develop specific mechanisms that allow the reference model to be customized.
Research Approach
We aim to develop a reference model for manufacturing DTs by synthesizing the design outcomes of four
case companies. We aim to support companies in their I4.0 initiatives by providing best practices for DT
design in an easily applicable way. For this purpose, the design science research framework of Kuechler and
Vaishnavi (2008) is combined with reference modeling approaches (Rosemann and van der Aalst 2007).
Awareness of the problem
In 2018, we launched a multi-year design study on DT development in the manufacturing context. Four
mechanical engineering companies participated, one developing a DT in sales, two a DT in product lifecycle
management, and one a DT in service and maintenance. The data collection on which the DPs and the
continuing reference model are based was multifaceted: qualitative questionnaires, eight hours of semi-
structured interviews (Fontana and Frey 1994), six hours of focus group discussions (Morgan 1997), 20
hours of workshops, and more than 30 hours of direct and indirect observations (Mayring 2004). The data
were both open and deductively coded. The design knowledge was translated into DPs following Gregor,
Chandra Kruse, & Seidel (2020). The DT developers were provided with these DPs but struggled to
operationalize them. Overall, we concluded that the problem was an implementation gap that made it
difficult for practitioners to incorporate the complex design knowledge of DTs into the development
process.
Suggestion
Our suggestion is to develop an adaptive reference model on a conceptual level, which means it should
abstract the used technology. This way, it is general enough to represent the design knowledge in a case-
independent way but at the same time gives enough support to develop complex systems like the DT. This
conceptual level allows heterogenic groups of DT developers, target users, and researchers to discuss and
plan DT implementations tailored to the companies' value creation. The envisioned reference model closes
the implementation gap and thereby addresses the scientific side of the problem of DT development. While
there are already some reference models for DT, they neither address codifying design knowledge nor target
software developers (Bevilacqua et al. 2020; Zheng and Sivabalan 2020). We refer to an established
Closing the Implementation Gap of Digital Twins
Twenty-eighth Americas Conference on Information Systems, Minneapolis, 2022
4
approach to derive an adaptive reference model in a design study based on literature and empirical input
(Hönigsberg et al. 2019). Our goal is to reduce the development effort for DTs and imprint the gained design
knowledge (DPs) into a reference model, which can be best realized by combining a configurative and a
generic adaptive reference model. Hence, we combine two types of adaptive reference models. Thus, with
the suggested solution, an instance-specific model can be configured and then manually adapted for an
optimal fit.
Development
The basic knowledge for developing DTs stems from our multi-year study on DT design. In the development
process, we captured theory ingrained design knowledge in DPs to guide the DT design in our four
manufacturing companies. These DPs were refined in the empirical cases. When the project team
encountered the "implementation gap" problem, the abstract DPs were reflected and detailed, drawing on
DT literature (Wache and Dinter 2021). In the literature review, the Work System Method (Alter 2013) was
used as a framework for structuring the scientific discourse of DTs, to understand DTs as a socio-technical
system and thereby addressing the practitioner side of the problem of DT development, the over-technical
DT understanding. A multi-view reference model was derived from the literature review results, explicating
the prescriptive knowledge of the DP from the research case. The intermediate results were validated during
the development in a focus group and workshops with 6-15 participants from our case companies. First, we
validated the literature-based assumptions on which DT aspects are relevant for specific implementation
cases in focus group interviews (cf. choice board and configuration rules). Second, we validated the
literature-derived DT functions and corresponding architectural assumptions in workshops (cf. reference
model views). The functions with corresponding architectural components were refined and logically
arranged to derive use case and architectural diagrams.
Evaluation
The reference model was applied to the four scenarios of the case companies to evaluate the applicability
and adequacy of the results. This scenario-based evaluation was discussed in expert interviews with four
developers of DT platforms involved in the DT development projects, and two IS scholars who specialized
in design research and DP development. The interviews lasted approximately 30 minutes each and included
a presentation and demonstration of the reference model followed by a semi-structured interview using an
interview guide. The results of the interviews were analyzed to identify potential improvements and gather
additional input. The revised reference model was presented to the interviewees to confirm that the
improvements were satisfactory. After one iteration, the results were deemed adequate.
The Digital Twin Project - Construction of the Reference Model
Awareness of the Implementation Gap and a Suggestion to Close it
During the design study's intensive requirements analysis and problem formulation from 2018 to the
beginning of 2020, interim results were repeatedly discussed and evaluated with the participating
companies. However, it became increasingly apparent that both the software development companies and
the mechanical engineering companies found the rather theoretical results of the design study difficult to
access. On the one hand, there was an impression of "cryptic research babble" and an attitude of "that's nice
for research, but I don't need all that in practice." On the other hand, the more detailed DT design
knowledge available was overly technical and inadequate to discuss the DT design as an integral part of the
case companies' value creation. The companies described an implementation gap in which the distance
from the abstracted knowledge to their system development reality is too big, and the overly technical DT
descriptions are too far from the companies' scenarios.
To close this implementation gap, the design project results were transferred into a semi-abstract design
artifact. Halfway between the discussed instance design and the abstract DPs, conceptual models for design
description needed to be developed. In our cases, a dilemma arose as, on one hand, the users of the design
knowledge preferred a representation as close as possible to the implementation, so little transfer efforts
and few adaptations to the own implementation are necessary. On the other hand, this representation made
the results less generalizable, whereby the design knowledge exhibited a small projectability to the other
Closing the Implementation Gap of Digital Twins
Twenty-eighth Americas Conference on Information Systems, Minneapolis, 2022
5
case companies (Vom Brocke et al. 2020). This dilemma is also known in other design knowledge capture
and transfer approaches, namely in reference modeling (Becker et al. 2007). Therefore, our proposal to
solve the implementation gap problem described above is to apply the solution mechanism of reference
modeling to find a middle ground between abstract and detailed guidance. More specifically, adaptive
reference modeling was applied to address our described dilemma. Here, the design knowledge from the
research cases is transferred into a model that gives generally valid implementation suggestions that are
adaptable to the instance case. The specific design of actors and services can be configured in an
implementation-specific way via an adaptation mechanism in the reference model (Rosemann and van der
Aalst 2007).
Development of the Reference Model
The DT literature was analyzed using the Work System Method as a lens, and the resulting aspects of DTs
could be classified into three main topics (Wache and Dinter 2020). First, with Strategy & Environment,
Processes & Activity, Actor and Product Level, the type of DT and its deployment is described. The type and
the use determine the DT's functionality and structure. Second, Information & Analysis and Service &
Function describe what functions it should provide during use. Third, Technological Link and
Infrastructure describe the underlying technological basis of a DT Platform. The literature review results
lead to the following three components of the reference model: the choice board, the modeling views
(functional and architectural view), and the configuration rules.
Choice Board: The following table represents the choice board of the reference model. On the left side of
Table 1, a question is stated, and on the right side, there is a choice to answer this question. The questions
can be answered from top to bottom, and at the end of this configuration, the reference model suggests a
functional and architectural scope for the desired DT. This tool can be used to capture the vision for the DT
to be developed at an early planning stage. This type of morphological choice board is well suited to
represent the variable solution space for reference models. The grey highlighting exemplifies one of our
case scenarios as a configuration in the reference model (grey = selected, white = not selected). The
configuration result is highlighted in the functional and architectural view as well.
Question
Choice
Which corporate objective
will the DT support?
Product
lifecycle
management
Product-
service system
Sustainability
management
Sales
management
In which life cycle phase will
the DT be used?
Design/plan
Production/build
Service/run
Who will use the DT?
Customer
Producer /internal
Supplier spanning
To which product level will
the DT refer?
Unit
level/component
System
level/machine/plant
System of systems
/production line/factory
Table 1. Choice Board
Functional View and Architectural View: The views have been modeled using the Unified Modeling
Language (UML) due to its widespread use. The choice of a widely used standard for system modeling
supports our intention to close the implementation gap of DTs by choosing an easily accessible and
understandable medium for developers, target users of the DT, and researchers alike. In addition to input
from the literature, three DPs from the multi-year study were used to inform the reference model's
construction. The DPs in the short form are (1) Cyber-physical (re-) configurability, (2) Smartness of the
product, and (3) IT platform with a microservice architecture as a boundary object (Wache and Dinter
2021). The functional view (Figure 1) was modeled as a use case diagram. Thus, not only the different
functions and activities but also the associated actors could be represented. Based on the literature on DTs,
the functions were grouped into three phases: Configuration (Plan), Order & Production (Build), and
Operations (Run). This grouping reflects that DT can be used in the planning phase, in the production
process, or in the operation of the finished plant. In addition, some administration functions have been
identified that need to be supported regardless of the focused phase. Corresponding to this functional view,
an architectural view was derived from the literature and our cases. For clarity, a three-layer architecture
Closing the Implementation Gap of Digital Twins
Twenty-eighth Americas Conference on Information Systems, Minneapolis, 2022
6
was chosen (Figure 2). Front ends were modeled for the various actors in the graphical user interface (GUI)
layer; in the application view, components behind a central access component represent subsystems that
are logically linked to certain functional scopes of the functional view. The different components, in turn,
access services, which provide partial functionalities. For example, the configuration component accesses
the physical and virtual configuration service to configure the machine as well as the associated sensors and
data flows. The data models to be considered are represented in the persistence layer and possible interface
systems in the external system view.
The first DP Cyber-physical (re-) configurability is exemplified by the presence of a configuration function
in the use case diagram as well as the presence of the configuration component with both physical and
virtual configuration services. The second DP Smartness of the product is embodied by the fact that in the
use case diagram, analysis functions such as simulation and optimization are made available to the user in
all lifecycle phases of the DT. This DP is reflected by the analysis service's central position, which is
interconnected with almost all components in the architecture view. The third DP IT platform with a
microservice architecture as a boundary object leads to the system being designed as a common platform
from the customer to the supplier. This DP is reflected both in the use case diagram by the actors and in the
architecture diagram by the various front ends and the possibility of integrating external systems.
Figure 1 Functional View
Adaptation Mechanisms and Rules: The type and use of DTs determine the system's functionality and
architecture. Thus, there is a logical relationship between the choices in the choice board and the individual
elements in the functional and architectural view (cf. highlighting in Table 1, Figures 1 and 2). The DT can
be configured on a conceptual level using the choice board. Several rules have been defined that select
functions and the corresponding architecture components when a choice is made in the choice board. The
configured model is created in the first step, which specifies the recommended scope for the two diagrams
User
Configuration
Place
order
View partner
catalog
Manage
user
Use
communication
Operations
Simulate, optimize
production process
Use plant
services
Control
production
process
Orchestrate
services
Manage data
protection and security
Show life cycle
and history
Manage
tickets
Diagnose
plant
Visualize
plant
Use 3D
visualization
Use AR
visualization
Prepare
configuration for
production
Monitor plant
condition
Upload, open
CAD files
Maintain data models
and rules
Perform
(predictive)
maintenance
Configure
plant
Receive
recommendations
Compare, save,
load, validate
configuration
View
component
information
Search, select,
position components
Adjust parameter setting
(color, material,
measurements)
Internal
Supplier
Manage plant
knowledge
Simulate,
optimize plant
Order & Production
Administration
Administrator
Restricted
visibility/
accessibility
Customer
Simulate,
optimize
design
Use production
services
Monitor
production
process
Manage production
process knowledge
Control
plant
Closing the Implementation Gap of Digital Twins
Twenty-eighth Americas Conference on Information Systems, Minneapolis, 2022
7
for a specific DT to be developed. In the second step, the model can be adapted (manually) with a generic
adaptation to generate a better fit for the specific application scenario. An example of a combination of
several such rules affecting both views: IF Design/Plan AS life cycle phase THEN INCLUDE Configure plant
AS Function AND Configuration component AS Architectural Component USING Physical configuration
service AND Virtual configuration Service. This chaining of rules does not correspond to all rules triggered
when Design/Plan is selected in the Choice Board but corresponds to a continuous example from the Choice
Board to the architecture view. For example, a rule for generic adaptation is: IF REMOVE Place order AS
Function THEN REMOVE Order Service AS Architectural component AND Order Model AS Persistence
Model. This type of rule does not configure the model in the classical sense but ensures consistency between
views when manipulated manually.
Figure 2 Architectural View
Evaluation of the Reference Model
Several interviews were conducted for the evaluation, involving four DT developers and two researchers,
one focusing on DPs, and another with a background in reference models. The DPs and the reference model
were shown and explained to the interviewees before they answered some questions of the interview guide.
A summary of the collected feedback is provided hereafter. Applying the reference model to the case
companies' scenarios was deemed suitable support by the developers, above all for an early development
phase. All interviewees were positive about whether the reference model is helpful for developers of DT
platforms. It was confirmed by the four developers that the reference model makes the rather abstract DPs
more accessible for the target group of practitioners and that they would prefer the reference model for
their development projects. However, the extent of support provided by the reference model would depend
on the application context and the individuals. Thus, the interviews with developers revealed a different use
of DPs and reference model depending, e.g., on the assumed software development methodology (waterfall
model, agile software development). There was unanimity among all interviewees that the DPs play a
strategic role in system development, whereas the reference model is used during operational software
development. Concerning DPs, there were differing views among developers and researchers as to whether
DPs should describe a system in full (practitioner viewpoint) or rather have a focus on a thematic area
(academic viewpoint). With regard to the reference model, both interviewed groups emphasized that its
extensibility is of great importance so that further DPs could be reflected and the reference model applies
GUI
Application
Customer
Frontend
Internal
Frontend
Supplier
Frontend
Configuration
component Security
service
Administrator
Frontend
Visualization
service
User management
service
Central access
component
Administration
component
Physical
configuration
service
Production
component
Orchestration
service
Knowledge
management
component
Simulation
service
Order service Communication
service
Virtual configuration
service
Data management
service
Maintenance
component
Recommendation &
optimization service
Real time data
management service Ticket service Diagnostic
service
Persistence / models
Analysis
model
Configuration
model
Ordering
model
Knowledge
base
Visualization
models
Production
model
Maintenance
models
Orchestration
model
Security
model
Master data
Analysis
service
Twin control
component
External Systems
E-Commerce
Plattform Community CRM DWH ERP SCM PDM CAP/PPS CAD
Cloud
Edge
Fog
On-Premise
Closing the Implementation Gap of Digital Twins
Twenty-eighth Americas Conference on Information Systems, Minneapolis, 2022
8
to different contexts, thus facilitating the instantiation of the design knowledge. With respect to
extensibility, both groups repeatedly expressed the desire for the addition of a data view that maps the DT
data schemas. In addition, the interviewees in both groups were missing an overarching process perspective
or meta-process that outlines how the target group of developers should use the reference model to
incorporate and implement the resulting suggestions into the development process.
After the interviews, the revision of the reference model focused on extending the views to homogenize their
granularity. Thus, the production part of the use case and the architecture diagram were extended by
elements.
Discussion
Our study developed a reference model to address the implementation gap by providing
designers/practitioners with a configurative tool that integrates the abstract DPs into more practical, use-
case-based approaches for DT design. We successfully answered the research question on how a DT
reference model should be constituted to make design knowledge available to IS scholars and practitioners
to close the implementation gap. The addition of our reference model converts the required knowledge base
for designers by integrating DPs into a guiding architecture while still allowing practitioners the necessary
freedom to develop a DT based on their needs. This principle is generally useful when technically
independent conceptual models are used in development projects to synchronize the world view of the
various project stakeholders. We assume that our model can also be used for companies in the same
industry. Using a socio-technical lens to extract the characteristics of DTs from the literature and transfer
the findings into a reference model leveraging DPs from a DT design study, we were able to address both
sides of the problem designers are facing while developing DTs. Our study addresses a gap in the research
on the design of I4.0 technologies such as DTs. In addition to accumulating knowledge about systems like
DTs (Baiyere et al. 2020), we also focus on the mechanism of making this knowledge accessible to
practitioners. This is an important undertaking because companies need the necessary technical expertise
to successfully implement digitalization initiatives in their companies (Legner et al. 2017; Vial 2019).
Of course, not all DPs are 'cryptic research babble'; there is a continuum between abstraction and specificity
in which DPs exist (Wache et al. 2022). Nonetheless, it can be observed that nuances of the encoded design
knowledge in DPs are lost in use. The transfer of DPs into a form closer to and more understandable for the
target group of developers was feasible. This transfer was tested in an evaluation, and it was shown that the
resulting format is more accessible to practitioners. Thus, it could be demonstrated that the implementation
gap could be successfully addressed. With this transfer of prescriptive knowledge to the target group of
practitioners, our study addresses one of the goals of design-oriented IS research (Chandra Kruse et al.
2016; Gregor et al. 2020). Design-oriented IS research seeks to accumulate instance knowledge and make
it accessible to other contexts and scenarios. In this context, design knowledge can be characterized as
descriptive and prescriptive. Other recent work is concerned with transferring descriptive knowledge into
prescriptive knowledge (Möller et al. 2021), which benefits the accumulation of prescriptive knowledge.
Our work can also be located in the process of knowledge generation and transfer but starts later, as we de-
abstract prescriptive design knowledge, i.e., express it less abstractly and more contextually to make it more
accessible to practitioners. Our work ties in with the findings of Möller et al. (2021) by representing different
aspects/characteristics/degrees of detail of the design knowledge spectrum through our proposed reference
model. Furthermore, the developed reference models represent a continuation of DTs' conceptualization
(van der Valk et al. 2021), where the development of DPs and reference models for DTs are necessary next
research steps. Van der Valk et al. (2021) argue that practitioners can improve their understanding of DTs
by critically examining their proposed archetypes. Our work contributes another building block to this more
accessible knowledge base on DTs for practitioners, introducing a DP-based reference model. It enables
practitioners to develop a company-specific DT, which acts as a guideline for developing DTs.
Nevertheless, our work has some limitations. Our research examined DTs through a socio-technical lens,
considering DTs as a platform that enables human engagement in an I4.0 context, thereby limiting the
applicability of our reference model to the domain of manufacturing. Our reference model is not readily
applicable to other DT domains such as buildings or healthcare. The practitioners evaluated the content of
our proposed reference model based on their experience of developing a DT platform during our multi-year
study on DTs. The two researchers, one from the field of DPs and one from the field of reference models,
had no connection to our research case and thus provided an external perspective onto the reference model.
Closing the Implementation Gap of Digital Twins
Twenty-eighth Americas Conference on Information Systems, Minneapolis, 2022
9
We acknowledge that further evaluation is required to validate our findings beyond the scope of our
research case by transferring the reference model onto other cases. During our evaluation of the reference
model, it became apparent that the reference model could be extended by several additional and zoomed-
in detail views to increase its overall usefulness. Thus, the current scope of the reference model can be seen
as a limitation.
Conclusion
Our study aims to create a reference model for DTs to support developers during the development process.
We address an implementation gap perceived by practitioners when they are confronted with abstract DPs
and convert them into implementation. Based on the characteristics of DTs from the literature and our case
companies, a multi-view adaptive reference model has been developed. As shown in the evaluation, the
reference model represents an abstract knowledge vehicle that can support the DT development already in
the early phases of development projects. In summary, our research contributes to the knowledge base by
integrating reference modeling with the more abstract DPs to create a semi-abstract knowledge artifact.
Future research should extend the evaluation to more cases, and further views and view refinements for the
DT reference model might be developed. A particular focus should be on further closing the implementation
gap by expanding the reference model from the abstract level to a level that is no longer independent of
technology and contains specific technical implementation approaches.
Acknowledgements
The research in this paper was supported by a grant from the German Ministry for Research and Education
(BMBF), project name: Co-TWIN, no: 02P17D146.
REFERENCES
Alter, S. 2013. “Work System Theory: Overview of Core Concept, Extensions, and Challenges for the
Future,” Journal of the Association for Information Systems (14:2), pp. 72–121.
Amabile, T. M., Collins, M. A., Conti, R., Phillips, E., Picariello, M., Ruscio, J., and Whitney, D. 1996.
Creativity in Context: Update to the Social Psychology of Creativity, New York: Routledge.
Baiyere, A., Topi, H., Venkatesh, V., Wyatt, J., and Donnellan, B. 2020. “Internet of Things (IoT) - A
Research Agenda for Information Systems,” Communications of the Association for Information
Systems (47:1), pp. 564589.
Becker, J., Delfmann, P., and Knackstedt, R. 2007. “Adaptive Reference Modeling: Integrating
Configurative and Generic Adaptation Techniques for Information Models,” in Reference Modeling
Efficient Information Systems Design Through Reuse of Information Models, J. Becker and P.
Delfmann (eds.), Heidelberg: Physica, pp. 2758.
Beverungen, D., Müller, O., Matzner, M., Mendling, J., and vom Brocke, J. 2019. “Conceptualizing Smart
Service Systems,” Electronic Markets (29:1), pp. 7–18.
Bevilacqua, M., Bottani, E., Ciarapica, F. E., Costantino, F., Donato, L. Di, Ferraro, A., Mazzuto, G.,
Monteriù, A., Nardini, G., Ortenzi, M., Paroncini, M., Pirozzi, M., Prist, M., Quatrini, E., Tronci, M., and
Vignali, G. 2020. “Digital Twin Reference Model Development to Prevent Operators’ Risk in Process
Plants,” Sustainability (12:3), pp. 1–17.
Vom Brocke, J., Winter, R., Hevner, A., and Maedche, A. 2020. “Special Issue Editorial – Accumulation
and Evolution of Design Knowledge in Design Science Research: A Journey through Time and Space,”
Journal of the Association for Information Systems (21:3), pp. 520544.
Chandra Kruse, L., Seidel, S., and Purao, S. 2016. “Making Use of Design Principles,” in DESRIST 2016:
Tackling Society’s Grand Challenges with Design Science, J. Parsons, T. Tuunanen, J. Venable, B.
Donnellan, M. Helfert, and J. Kenneally (eds.), St. John’s: Springer, pp. 3751.
Dietz, M., and Pernul, G. 2020. “Digital Twin: Empowering Enterprises Towards a System-of-Systems
Approach,” Business and Information Systems Engineering (62:2), pp. 179–184.
Fettke, P., and Loos, P. 2003. “Multiperspective Evaluation of Reference Models Towards a Framework,”
in Conceptual Modeling for Novel Application Domains: ER 2003, Lecture Notes in Computer Science
(LNCS) (Vol. 2814), M. A. Jeusfeld and Ó. Pastor (eds.), Berlin: Springer, pp. 8091.
Closing the Implementation Gap of Digital Twins
Twenty-eighth Americas Conference on Information Systems, Minneapolis, 2022
10
Fontana, A., and Frey, J. H. 1994. “Interviewing: The Arts of Science,” in Handbook of Qualitative Research,
N. K. Denzin and Y. S. Lincoln (eds.), Thousand Oaks: SAGE Publications, Inc., pp. 361376.
Glaessgen, E., and Stargel, D. 2012. “The Digital Twin Paradigm for Future NASA and U.S. Air Force
Vehicles,” in 53rd Structures, Structural Dynamics and Materials Conference, Honululu, pp. 1–14.
Gregor, S., Chandra Kruse, L., and Seidel, S. 2020. “The Anatomy of a Design Principle,” Journal of the
Association for Information Systems (21:6), pp. 16221652.
Grieves, M. 2014. “Digital Twin: Manufacturing Excellence through Virtual Factory Replication,” Digital
Twin White Paper.
Grieves, M., and Vickers, J. 2017. “Digital Twin: Mitigating Unpredictable, Undesirable Emergent Behavior
in Complex Systems,” in Transdisciplinary Perspectives on Complex Systems: New Findings and
Approaches, F.-J. Kahlen, S. Flumerfelt, and A. Alves (eds.), Cham: Springer, pp. 85113.
Hönigsberg, S., Kollwitz, C., and Dinter, B. 2019. “Designing a Reference Model for Digital Product
Configurators,” in 14th International Conference on Wirtschaftsinformatik, Siegen, pp. 214–228.
Kagermann, H., Wahlster, W., and Helbig, J. 2013. “Recommendations for Implementing the Strategic
Initiative INDUSTRIE 4.0: Final Report of the Industrie 4.0 Working Group,” Frankfurt/Main.
Kuechler, B., and Vaishnavi, V. 2008. “On Theory Development in Design Science Research: Anatomy of a
Research Project,” European Journal of Information Systems (17:5), pp. 489–504.
Legner, C., Eymann, T., Hess, T., Matt, C., Böhmann, T., Drews, P., Mädche, A., Urbach, N., and Ahlemann,
F. 2017. “Digitalization: Opportunity and Challenge for the Business and Information Systems
Engineering Community,” Business and Information Systems Engineering (59:4), pp. 301308.
Mayring, P. 2004. “Qualitative Content Analysis,” in A Companion to Qualitative Research, U. Flick, E. von
Kardoff, and I. Steinke (eds.), London: SAGE Publications, Inc., pp. 266270.
Möller, F., Haße, H., Azkan, C., van der Valk, H., and Otto, B. 2021. “Design of Goal-Oriented Artifacts from
Morphological Taxonomies: Progression from Descriptive to Prescriptive Design Knowledge,” in 16th
International Conference on Wirtschaftsinformatik, Duisburg-Essen, pp. 117.
Morgan, D. L. 1997. The Focus Group Guidebook, Thousand Oaks: SAGE Publications, Inc.
Rosemann, M., and van der Aalst, W. M. P. 2007. “A Configurable Reference Modelling Language,”
Information Systems (32:1), pp. 123.
Rosen, R., Von Wichert, G., Lo, G., and Bettenhausen, K. D. 2015. “About the Importance of Autonomy and
Digital Twins for the Future of Manufacturing,” IFAC-PapersOnLine (28:3), pp. 567572.
Schoormann, T., Möller, F., and Hansen, M. R. P. 2021. “How Do Researchers (Re-)Use Design Principles:
An Inductive Analysis of Cumulative Research,” in DESRIST 2021: The Next Wave of Sociotechnical
Design, L. Chandra Kruse, S. Seidel, and G. I. Hausvik (eds.), Kristiansand: Springer, pp. 188194.
Sein, M. K., Henfridsson, O., Purao, S., Rossi, M., and Lindgren, R. 2011. “Action Design Research,” MIS
Quarterly (35:1), pp. 3756.
Shafto, M., Conroy, M., Doyle, R., Glaessgen, E., Kemp, C., LeMoigne, J., and Wang, L. 2012. “Modeling,
Simulation, Information Technology & Processing Roadmap,” Washington.
Tao, F., Zhang, H., Liu, A., and Nee, A. Y. C. 2019. “Digital Twin in Industry: State-of-the-Art,” IEEE
Transactiions on Industrial Informatics (15:4), pp. 24052415.
van der Valk, H., Haße, H., Möller, F., and Otto, B. 2021. “Archetypes of Digital Twins,” Business and
Information Systems Engineering. (https://doi.org/10.1007/s12599-021-00727-7).
Vial, G. 2019. “Understanding Digital Transformation: A Review and a Research Agenda,” Journal of
Strategic Information Systems (28:2), pp. 118144.
Wache, H., and Dinter, B. 2020. “The Digital Twin – Birth of an Integrated System in the Digital Age,” in
53rd Hawaii International Conference on System Sciences, Maui, pp. 54525461.
Wache, H., and Dinter, B. 2021. “Digital Twins at the Heart of Smart Service Systems - An Action Design
Research Study,” in 29th European Conference on Information Systems, Marrakech, pp. 117.
Wache, H., Möller, F., Schoormann, T., Strobel, G., and Petrik, D. 2022. “Exploring the Abstraction Levels
of Design Principles: The Case of Chatbots,” in 17th International Conference on Wirtschaftsinformatik,
Nürnberg, pp. 115.
Zheng, P., and Sivabalan, A. S. 2020. “A Generic Tri-Model-Based Approach for Product-Level Digital Twin
Development in a Smart Manufacturing Environment,” Robotics and Computer-Integrated
Manufacturing (64:101958), pp. 118.
... Als Lösungsmechanismus (LM3) zu den verschiedenen spezialisierten Lösungen (SF3) wurden die Best Practices des Entwicklungsteams durch die Integration der Lösungsansätze zu einem Referenzmodell zusammengeführt. Eine Anwendungsfallsicht im Modell erfasst die konkreten Szenarien der Unternehmen und eine Architektur-Sicht bildet die konkreten technischen Lösungen der Entwickler ab (Wache et al. 2022). Durch die Integration der verschiedenen Instanzbestandteile im Referenzmodell und die integrierte Weiterentwicklung der Einzellösungen wurde auch das angestrebte übergreifende Konzept der wissenschaftlichen Partner adressiert. ...
Article
Full-text available
Zusammenfassung Für die Digitalisierung der Fertigungsindustrie ist es notwendig, dass Unternehmen in neuen Partnerkonstellationen kollaborieren, um neuartige digitale Services anzubieten. Die Zusammenarbeit bei der Entwicklung von den entstehenden Service Ökosystemen hat einen großen Einfluss auf die später möglichen Geschäftsmodelle. Im vorliegenden Beitrag werden anhand eines Beispielprojektes für die Entwicklung einer Digital Twin Plattform Spannungsfelder und Lösungsmechanismen vorgestellt, die beim Aufbau von digitalen Softwareplattformen und Service Ökosystemen auftreten können. Die Spannungsfelder erstrecken sich über alle Projektphasen und resultieren aus drei Faktoren: einer fehlenden gemeinsamen Weltsicht, dem Bedürfnis nach lokaler Optimierung sowie einem Fokus auf Effizienzsteigerungen statt Innovationen. Abschließend werden Leitfragen formuliert, mit denen Spannungsfelder identifiziert werden können, um Lösungsmechanismen zu entwickeln und somit die erfolgreiche Gestaltung eines Service Ökosystems sicherzustellen.
Conference Paper
Full-text available
Formulating design principles is the primary mechanism to codify design knowledge which elevates its meaning to a general level and applicability. Although we can observe a great variety of abstraction levels in available design principles, spanning from more situated to more generic levels, there is only limited knowledge about the corresponding (dis-)advantages of using a certain level of abstraction. That is problematic because it hinders researchers in making informed decisions regarding the (intended) level of abstraction and practitioners in being oriented whether the principles are already contextualized or still require effort to apply them within their situation. Against this backdrop, this paper (1) explores different abstraction levels based on a sample of 69 design principles from the chatbot domain as well as (2) provides a preliminary positioning framework and lessons learned. We aim to complement methodological guidance and strengthen the principles' applicability, leading to knowledge reuse.
Article
Full-text available
Currently, Digital Twins receive considerable attention from practitioners and in research. A Digital Twin describes a concept that connects physical and virtual objects through a data linkage. However, Digital Twins are highly dependent on their individual use case, which leads to a plethora of Digital Twin configurations. Based on a thorough literature analysis and two interview series with experts from various electrical and mechanical engineering companies, this paper proposes a set of archetypes of Digital Twins for individual use cases. It delimits the Digital Twins from related concepts, e.g., Digital Threads. The paper delivers profound insights into the domain of Digital Twins and, thus, helps the reader to identify the different archetypical patterns. The Version of Record is available online at: http://dx.doi.org/10.1007/s12599-021-00727-7 under the CC BY Licence
Chapter
Full-text available
Accumulating prescriptive design knowledge, such as design principles (DP), is one of the fundamental goals in design science research projects. As previous studies have examined the use of DPs in practice to advance the development and communication of such principles, we argue that this attention also needs to be paid to how and for what researchers (re-)use DPs. Hence, this paper explores DP usage in cumulative (information systems) research based on the analysis and coding of a sample of 114 articles with 226 in-text citations. In doing this, we aim at contributing to the valuable discourse on DP reuse and accumulation by focusing on usage in research, present preliminary types of DP usage extracted from cumulative literature, as well as raise the awareness for guiding user and designer in how to (re-)use and how to allow for reuse of DPs.
Conference Paper
Full-text available
Today’s rapidly advancing technology is shaping many aspects of our personal and working lives. In the context of the continuing digitalization of the manufacturing industry Digital Twins (DT) have emerged as a new kind of networked information system, which converges real-world assets with virtual counterparts. Smart service systems (SSS) provide a new perspective on these information systems enabling a connection between the technical concepts of industry 4.0 with service science. We conduct an action design research study in the machine manufacturing industry and use the SSS as a guiding perspective for our DT design. This paper contributes three design principles for DTs in manufacturing: (1) cyber-physical (re-)configurability, (2) smartness of the product, and (3) IT platform as a boundary object. The design was evaluated with several manufacturing companies based on a user interface prototype.
Article
Full-text available
The Internet of Things (IoT) is emerging as an integrated set of digital innovations with the potential to unleash unprecedented opportunities as well as to create significant challenges from both technological and societal perspectives. The emergence of IoT heralds a new dimension of a digital era with impact and influence that are not yet fully clear. This signals the opening of valuable opportunities for scholarly inquiries, particularly for information systems (IS) scholars. We posit that, as the IS discipline sits at the intersection of technical, business, and social applications of IT, which are also the essential dimensions of the impact of IoT, IS scholars are well positioned to understand and contribute to advancing research on this new topic and associated phenomena. This paper outlines the distinctive attributes of IoT and their implications for existing traditions of IS research. It further highlights some illustrative research perspectives from which IoT can be studied by IS scholars. We highlight a research agenda for IS in two different ways: first, by suggesting four categories of implications on IS research: (1) introduction of the physio-digital continuum; (2) multi-level exploration of IS; (3) composite affordances; and (4) heterogeneity; and second, by introducing four thematic impact domains: (1) impact on organizations; (2) impact on technology; (3) impact on individuals; and (4) impact on society.
Article
Full-text available
This essay derives a schema for specifying design principles for information technology-based artifacts in sociotechnical systems. Design principles are used to specify design knowledge in an accessible form, but there is wide variation and lack of precision across views on their formulation. This variation is a sign of important issues that should be addressed, including a lack of attention to human actors and levels of complexity as well as differing views on causality, on the nature of the mechanisms used to achieve goals, and on the need for justificatory knowledge. The new schema includes the well-recognized elements of design principles, including goals in a specific context and the mechanisms to achieve the goal. In addition, the schema allows: (i) consideration of the varying roles of the human actors involved and the utility of design principles; (ii) attending to the complexity of IT-based artifacts through decomposition;(iii) distinction of the types of causation (i.e., deterministic versus probabilistic); (iv) a variety of mechanisms in achieving aims; and (v) the optional definition of justificatory knowledge underlying the design principles. We illustrate the utility of the proposed schema by applying it to examples of published research. Available at: https://aisel.aisnet.org/jais/vol21/iss6/2
Article
Full-text available
In the literature, many applications of Digital Twin methodologies in the manufacturing, construction and oil and gas sectors have been proposed, but there is still no reference model specifically developed for risk control and prevention. In this context, this work develops a Digital Twin reference model in order to define conceptual guidelines to support the implementation of Digital Twin for risk prediction and prevention. The reference model proposed in this paper is made up of four main layers (Process industry physical space, Communication system, Digital Twin and User space), while the implementation steps of the reference model have been divided into five phases (Development of the risk assessment plan, Development of the communication and control system, Development of Digital Twin tools, Tools integration in a Digital Twin perspective and models and Platform validation). During the design and implementation phases of a Digital Twin, different criticalities must be taken into consideration concerning the need for deterministic transactions, a large number of pervasive devices, and standardization issues. Practical implications of the proposed reference model regard the possibility to detect, identify and develop corrective actions that can affect the safety of operators, the reduction of maintenance and operating costs, and more general improvements of the company business by intervening both in strictly technological and organizational terms.
Article
Digital twin (DT) is one of the most promising enabling technologies for realizing smart manufacturing and Industry 4.0. DTs are characterized by the seamless integration between the cyber and physical spaces. The importance of DTs is increasingly recognized by both academia and industry. It has been almost 15 years since the concept of the DT was initially proposed. To date, many DT applications have been successfully implemented in different industries, including product design, production, prognostics and health management, and some other fields. However, at present, no paper has focused on the review of DT applications in industry. In an effort to understand the development and application of DTs in industry, this paper thoroughly reviews the state-of-the-art of the DT research concerning the key components of DTs, the current development of DTs, and the major DT applications in industry. This paper also outlines the current challenges and some possible directions for future work.