Conceptualization of Digital Twins in an Education

Services Environment: A Straw Man Proposal

Allan Sylvester, Rebecca Mines, Rodreck David, Jennifer Campbell-Meier

To cite this version:

Allan Sylvester, Rebecca Mines, Rodreck David, Jennifer Campbell-Meier. Conceptualization of Dig-
ital Twins in an Education Services Environment: A Straw Man Proposal. Proceedings of the 56th
Hawaii International Conference on System Sciences, HICSS Conference Office Department of IT
Management, Shidler College of Business, University of Hawaii at Manoa, Jan 2023, Lahaina, Maui,
Hawaii, United States. �hal-04666810�

HAL Id: hal-04666810

[Link]
Submitted on 2 Aug 2024

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.

Distributed under a Creative Commons Attribution - NonCommercial - NoDerivatives 4.0

International License
 Proceedings of the 56th Hawaii International Conference on System Sciences | 2023

Conceptualization of Digital Twins in an Education Services Environment:

A Straw Man Proposal
Allan Sylvester Rebecca Mines Rodreck David Jennifer Campbell-Meier
Victoria University of Victoria University of Em-Lyon Business Victoria University of
Wellington Wellington School Wellington
New Zealand New Zealand France New Zealand
[Link]@[Link] [Link]@[Link] rdavid@[Link] [Link]-
meier@[Link]

Abstract energy, and infrastructure. Previous research shows that

Digital twins have been used in manufacturing to many organizations, particularly in the manufacturing
describe, predict, and prescribe responses to complex sector, see digital twin technology as means to realize
problems. The digital twin is a constellation of significant commercial benefits by enhancing the
technologies that mirror physical objects in the virtual experiment space that informs improvements in such
world, including what has happened, is happening and areas as operations, product, and service innovation.
could or should happen in the future for the mirrored Digital twinning seeks to bring together the best
object. What is common in previous conceptualizations features of planning, simulation, forecasting, and big
of digital twins is that there is a physical boundary to data into a cohesive digital artefact that can be used to
the extent that digital twins can mirror real objects provide a digital representation of reality.
(sometimes including the objects’ environments). We The term “digital twin” has long been used to
propose a blended approach, using McKinsey’s straw describe a virtual, digital equivalent of a physical
man and Parmar et al.’s. (2020) framework, to offer a product (Zhang, Ma, Sun, Lin, & Thürer, 2019) and a
more rigorously structured process for arriving at a connection between a virtual object and its physical
refined conceptualization of digital twins in the counterpart (Van der Valk et al., 2020) such as between
educational service environment. digital cars, aircraft, buildings, or even cities, and their
physical equivalents. Such a connection requires that
Keywords: digital twins, service environments, various data be collected in real-time, from the physical
education object to its digital mirror or twin (Glaessgen and
Stargel 2012). This includes data such as: sensory data,
granular “atomic” level data, geometric data, current,
1. Introduction
past, and predicted data (Grieves and Vickers 2017). A
high level of synchronicity is also important as it enables
It has become normal for organizations to seek
the digital twin to maintain an “ultra -realistic” version
strategic advantages from leveraging digital
of its physical counterpart (Glaessgen and Stargel
transformation. This requires organizations to have both
2012). This, in-turn, means that a digital twin requires
the capacity and capabilities to self -evolve and
highly durable systems support for handling high
anticipate current and future changes rather than being
frequency and high velocity data (Rathore et al., 2021).
reactive (Berisha -Gawlowski, Caruso & Harteis, 2021).
The digital twin is a constellation of technologies
Digital systems and the data they generate can provide
that mirrors physical objects virtually, capturing data
opportunities for organizations to understand how they
about what has happened, what is happening, and/or
deliver their services, and to predict and prescribe how
what could or should happen in the future for the
they could respond to complex, yet necessary changes
‘twinned’ object. For example, every Telsa car, or
that could provide strategic advantages. Often, this
aircraft developed by GE has a digital twin (El Saddik,
requires tools for testing ideas and assumptions before
2018; Saracco, 2018). These products report back to
full deployment of changes in processes, practices,
their parent companies daily, about their experiences
technologies, and other facilities that organizations use
and such data are used by simulation programs on the
to deliver services. The ability to test hypotheses in
digital twin to discover anomalies and provide
near-real time without disrupting the real-world value
corrective actions. The multitude of digital twins, for
creation environment makes digital twin technology a
Tesla as for GE, makes possible to learn from multiple
very attractive proposition for organizations in various
experiences. For example, Tesla , Inc. indicates that it
sectors, including manufacturing, healthcare, transport,
gets the equivalent of 1.6 million miles “a day” of

URI: [Link]
978-0-9981331-6-4 Page 5643
(CC BY-NC-ND 4.0)
driving experience and this is fed back to each car in a
continuous learning process (Saracco, 2018).
While simulation certainly is a central aspect of
how digital twins work, particularly though predictive
capabilities (Van der Valk, 2020), digital twins are more
than simulations in that they do not necessarily carry
intentionality, but rather, they indiscriminately capture Figure 1. Cyber-physical progression of
real-time data about the functioning of an object, such a product cycle using a digital twin
that even unintended events, anomalies, and
discrepancies between intended design use and actual
use can be captured. This tells us how something that 2.2 Digital Twins in the Services Context
was designed is being used, which goes beyond
forecasts based on use, or predicted use. What is common in previous research on the
Manufacturing was quick to leverage the benefits application of digital twinning in manufacturing is that
there is a physical boundary to the extent that digital
of statistical process control in the first half of the 20th
century (Shewhart & Deming, 1986) and services twins can mirror real objects (sometimes including the
caught up in the later part of the century. For example, objects’ environments). There is, however, a noticeable
trend in the application of the term in service
applying lean six sigma to services is now a relatively
normal practice. A similar situation is unfolding as environments such as health care services, education,
manufacturing has been seeking and optimizing their and financial services (cf. Elayan et al., 2021; Liljaniemi
& Paavilainen, 2020; Sepasgozar, 2020), where such
value creation processes with digitalization and digital
twins representing the current state of the art. It seems services do not have a clear-cut physical boundary. The
only logical that the service sector will seek to learn and services are often delivered through multiple digital
systems and services. For example, imagine trying to
capitalize on that knowledge base as it has done before.
mirror the information services offered to a patient in
primary care: Would the patient record be the main
artefact to be “twinned”? Would this also include the
2. Background to Digital Twins clinical services information involved? What of the drug
administration from pharmacies, and the associated
In the following discussion, we describe the records? What about insurance services? What model of
context of digital twins in manufacturing and explain digital twinning can be applied in such a scenario? The
why this becomes challenging when extended to a same can also be said for education services, social
services setting. We focus on education services as a use services, and so forth.
case and identify the research gap explored in this paper. As such, digital twins in services environment
must encapsulate several fragments of a “defined
2.1 Digital Twins in Manufacturing Industries service” or “service agent” or “client”. This includes
parts of systems that collect data about such a service,
In the field of manufacturing, digital twins are at agent, or client. In a way, each actor, client, or agent in
the heart of creating cyber-physical systems (CPS), the service environment can have several fragments of
which integrate physical and digital components (Tao et their digital twin generated through data collection by
al., 2019). In such cases, a digital twin serves as a virtual various information systems. For example, social media
mirror of the physical system. For example, such platforms such as Twitter, Facebook, and Instagram, all
systems have recently been implemented to capture and collect parts of an individual’s “self” attributes (e.g.,
mirror developments in cities (Ketzler et al., 2020), preferences, choices, and interests). Governments and
automotive industries (Biesinger & Weyrich, 2019), and local authorities also collect other parts, and so do
production lines for other physical products. The digital systems such as education, healthcare, and department
version is often used to track how well the physical stores (club card systems, transaction histories).
product is running, collect data about experiences in its Similarly, companies where individuals work collect
use, and possible inefficiencies, without necessarily personal information that can easily profile an
affecting the use of the physical/CPS version (Figure 1). individual’s unique characteristics.
Digital twinning is also used to carry out Moreso, unlike machine agents such cars,
simulations to create representations of the physical airplanes, robots, and other inanimate gadgets often
system that integrate these two parts, physical and twinned in the manufacturing industry, human agents
cyber, to a full closed loop control system in which the are the key actors twinned in the service context.
physical system is controlled by the virtual one. Twinning human agents introduces subjective elements

Page 5644
of human behavior that are quite complex. This includes There is limited research that has explored the
unpredictability, changing human interests, choices, and use of digital twins to augment the delivery of education
decisions; all of which makes digital twinning in the services. Previous research related to education tends to
services context a complex problem. focus on digital twins of engineering and architectural
Until recently (and in some ways this remains concepts used for learning (cf. Liljaniemi &
so), all these fragments are dispersed and there exist Paavilainen, 2020; Dai & Brell-Çokcan 2022), but this
laws and regulations on how such information can be does not directly address broader service delivery itself.
shared across multiple agencies. Services and their We draw conjectures from manufacturing and
supporting IT systems are often dispersed and based on organizational research (e.g., Biesinger & Weyrich,
actor-to-actor interactions that can change depending on 2019; Berisha -Gawlowski et al., 2021) that there are
the situation. This makes it complex to use digital twins benefits of applying digital twins to educational
as they are traditionally applied in manufacturing. The services. On the organizational side, there is an
complexity of embedding a digital twin is therefore opportunity to firstly review systems architecture or
dependent on the nature of the service targeted. Th e structure to identify how the systems are structured to
original conceptualization of digital twins may not both maintain a record of the student and deliver
necessarily fit into such service environments where the learning. Secondly, to identify the hierarchy of systems
original “mirrored” entity is not physically bound and following systems or data life cycle of the student
exists as a conglomeration of multiple data and journey and of university processes. Lastly, there is also
information systems (see Table 1). Thus, a unified an opportunity to enhance data collection and analytics
definition and (re)conceptualization of digital twinning performed by the education services providers.
that considers the services environment is required. Creating a digital replica of, (a) the student’s
digital footprints as they traverse an educational system,
Table 1. Comparisons between digital twinning (b) the record of a student’s interactions (touchpoints)
in manufacturing and services contexts with an educational system is unique because it captures
Manufacturing Services Context nuances from distinct systems that are often
Industry disintegrated. A student may leave digital footprints
Tightly coupled Multiple independent even though they have not specifically interacted with a
system, for example, logging onto Wi-Fi and entering a
systems often systems from which the
building are not typical touchpoints in the way that an
integrated as part of a digital twin can be
product created information request or enrollment in a course would
normally be tracked.
“Object of interest” may
Object of interest is span multiple
physically bound organizations with 2.4 Research Gap
different access rules
Machine to machine Human to machine The existing literature is sparse when it comes to
interaction is more interaction is more information systems use for service delivery in higher
prevalent but human to prevalent but machine education. Digital twins can inform how certain
machine interaction is to machine interaction is information systems will perform under various
also present also present economic, environmental, and social conditions, and
identify the drivers of possible disruptions.
Boundary of the real
The nuances of human experience and culture are
Boundary of the real “object” is difficult to
beyond the scope or unable to be interpreted by the
object is clear cut (easy define – often
typical instruments for data collection (such as sensors
to identify) intertwined with other
and post-hoc data extraction from settings such as Wi-
“objects”
Fi access or building security systems) in digital twin
scenarios. Perhaps in the future AI based video agents
2.3 Digital Twins in Educational Services will be able to read visual cues while people are
interacting with avatars, but for now, sensors and data
Education is a service sector that must operate at collection approaches are limited to relatively crude
considerable scale, generates large volumes of measures. The point is that understanding these human
transactional data, and requires many touchpoints for complexities and their dependencies on underlying data
funders, providers, consumers (a.k.a., students), and use policies ca n provide detailed insights for managing
their support agents. As such, education is to us a educational services. Therefore, to unpack this
brilliant natural laboratory to explore the use and issues complexity in the educational services setting, we firstly
with digital twins in a service context. specified the heuristics for developing a digital twin in

Page 5645
the educational services environment. Secondly, we 2. Discuss The team critiques the proposal,
made use of an exploratory vignette to explore the detailing arguments for and/or against
conceptualization of digital twinning in this context. the concept. This may lead to
This combined approach used traditional consulting deconstruction or reconstruction of
techniques scaffolded with theoretical foundations. This parts of the concept at this stage.
is fully discussed in the following section on methods. 3. Clarify The process continues in the ‘clarify’
Our exploration was guided by the followin g stage, where feedback and criticism
research question: How can the concept of digital begin to shape the next iteration of
twinning be developed in an educational services the concept. The previously identified
context? arguments are tested against this new
conceptualization to confirm its
3. Methods suitability for the challenge at hand.
4. Decision The proposal is then finalized. The
To explore how digital twinning can be goal is to evolve a draft concept into
developed in an educational services context, we a version which carries the full
utilized the straw man proposal process as an expository support of all consulting members.
tool. We adapted McKinsey’s Straw Man Proposal
Process (Think Insights, 2022) by blending it with the 3.2 Principles & Evolutionary Process Model
principles and evolutionary process model for building for Building a Digital Twin
a digital twin put forward by Parmar et al. (2020 p.732).
As we needed structure to our proposal process in a way
3.1 McKinsey’s Straw Man Proposal Process that considers digital twins as the object of interest, we
blended the straw man proposal with Parmar et al.’s
The straw man proposal process (see Table 2) principles and evolutionary process model for building
has previously been used by researchers as a a digital twin (see Parmar et al., 2020 p.732; Figure 2).
methodological tool that provides a framework for Although Parmar et al.’s (2020) focus was on the
design modelling and theorization (Cantalupo et al., process for developing digital twins in the
2018). For example, researchers have used the straw manufacturing context, the logical organization of these
man proposal when they have conjectures about how a principles was applicable and helpful in structuring our
new system, process, or solution might work, and wish conceptualization process. Below, we outline the
to give themselves and others a model to challenge and principles and how we applied them in this study.
evolve (cf. Veeraraghavan & Reisslein, 2011).
Applied in previous research the straw man
approach provided an opportunity for researchers to
conduct multiple iterations, tear apart initial ideas, and
rebuild on their initial theorization. The straw man
proposal has also been used to formulate and evolve
hypotheses with limited data before full-scale
quantitative tests can be performed (Hattis et al., 2002).
As a methodological process, the straw man approach is
also used in applied industrial research and consulting
(e.g., by companies such as McKinsey and Intel, cf. Figure 2. Parmar et al.’s principles and evolutionary
Cantalupo et al., 2018; Think Insights, 2022). Used in process model for building a digital twin
this paper, the straw man approach offers a useful
starting point for conceptualizing digital twinning in the 1. Start with what you have
educational services environment. This principle would normally require undertaking a
comprehensive audit of information systems, data
Table 2. McKinsey’s Straw Man Proposal Process collection points, and the student experience, to create
an initial model of the digital twin. For this exercise, this
Step Description
step is carried out in sketch form only but should still
1. Concept In the first stage of the straw man capture the essential touch points of the interaction.
proposal process, an idea to resolve a
particular challenge is raised with a 2. Set the data free
consulting team. Much of the data collected though information systems
in the University’s enrollment processes is captured in

Page 5646
“silos”. To successfully create a digital twin, data needs
to be “set free” from their respective silos. This can also
become one of the value-generating functions of a
digital twin; so, the more integrated the twin is, the more
useful it can be. Therefore, improving the accessibility
and opportunities to analyze data in timely manner.

3. Move the digitalization frontier

There are some stages in most enrollment process which Figure 3. The adapted and merged model
are only partially digitalized. To realize the full potential
of a digital twin, we must actively seek out those stages In the ‘concept’ stage of the straw man proposal,
and digitalize them. Data generated in these newly we ‘start with what we have.’ Here, we derived our
digitalized stages should be integrated into the initial straw man concept from a roundtable
organizational model, so that the twin can more brainstorming activity involving four participants: (1) a
accurately represent its “real-time” twin. Service senior tutor working with student learning systems (2)
designers have an important role to play at this stage. an associate dean - students, (3), a program director and
(4), an administrator. We selected these participants for
4. Seek new digital opportunities the expert panel because of their knowledge of student
As the twin evolves and continues to be improved upon, journeys and of the various systems and datapoints
new opportunities emerge for how the twin can be used created in these journeys. This ‘conceptualizing’
for organizational improvement. In sum, the insights process was underpinned by our assumption that a
generated from the twin can “identify and execute new digital twin can be successfully created that mirrors
digital opportunities to further extend the reach and educational services delivery.
coverage of the digital twin” (Parmar et al., 2020 p.16). In the second stage, ‘discuss’, we ‘set the data
free’ through a use case. The ’use case’ we selected
5. Increment the twin focuses on the journey of a post-graduate student
This principle provides an opportunity to reflect on enrolling at a university. This use case was selected as
learnings, incorporating them in iterative feedback, an example because it is one of the more straightforward
before moving the digitalization frontier and seeking enrollment processes. Any university enrollment
new digital opportunities. The expectation is that a process is represented by a complex set of use cases. For
digital twin will create a virtuous loop where new the purposes of this exercise, we deliberately choose one
products and services can emerge because learners, of the more tightly bound cases to help manage the
educators and institutional leaders can gain deeper scope and create clarity on the concept of DTs in the
understanding of what, how and why some educational educational service environment. The use case
services are successful whilst others are not. highlights the key information system touch points (i.e.,
where the prospective student generates data at each
3.3 The Adapted Methodological Approach step of the enrollment journey). Discovering these touch
points requires document and process analysis. We
Our blended approach provided us with a paired our use of a case with document analysis of an
framework to structure to our proposal process, framing interactive Student Orientation Module developed to
digital twins as the object of interest (see Figure 3). We give a procedure advice to students enrolling in a Master
were able to offer a more rigorously structured process of Information Studies (MIS) program. As we navigated
for arriving at a refined conceptual model for digital this enrollment process, we noted the information a
twins in the educational services environment. In doing prospective student was required to supply, and the
so, we took a concept in need of exploration and various information systems a student would interact
development, then simultaneously explored the with on the journey.
feasibility and scope of that concept in a bounded In the next stage, we use the insights gleaned
context and outlined how the concept could be from our discussion to (re)conceptualize our initial
actualized. Our use of this blended approach in the digital twin concept. In clarifying the concept at this
educational services environment is a novel approach stage, we also delivered building blocks for nascent
that extends previous applications of the straw man theory in the digital twinning conversation, in the form
proposal process and Parmar’s principles and of an incremented digital twin concept.
evolutionary process model (see Figure 3).

Page 5647
 In the ‘decision’ stage, we merged principles 3 loop starting with initial agency, experience and
and 4 of Parmar et al.’s (2020) framework; ‘move the touchpoints with learning platforms, data capture, and
digitalization frontier’ and ‘seek new digital feedback
opportunities.’ We structured our ‘decision’ as a
discussion of our findings (Section 5) to reflect on the 4.2 Discussion
learnings from our straw man process and finalize the
proposal. In this stage, we also discuss the benefits, The initial straw man model we developed (Figure
applications, challenges, and limitations of our digital 4) identified weaknesses and areas of improvement to
twin concept. present a better conceptualization. To do this, we made
scenario analysis using use cases to examine the agency
4. Findings of the two main actors in our model (1) the student, and
(2) the education services provider(s).
The expository process we followed enabled us We constructed a use case of a student enrolling in
to identify key elements for developing a digital twin in the MIS program. We present two perspectives of this
the education services context. We also made progress case from our analyses. [Appendix A: Use case
on illustrating how these elements could be organized diagrams. Figure 5 (Use Case 1) models the student and
(see Figure 7) to capture data during interactions with their touchpoints with systems in the enrollment process
key agents and how this can create a feedback loop that (including data generated at each of these touchpoints).
continuously improves services whilst informing the Figure 6 (Use Case 2) models the student’s interactions
key actors (students and services providers). has with people in the enrollment process]. These two
revealed that digital twins in a service science context perspectives capture the enrollment experience in the
general and education specifically is a domain MIS program and reveal new insights to further the
warranting further exploration. development of our straw man. Using use cases in this
way revealed the student’s experiences and agency
4.1 Conceptualization (decision making, mistakes, etc.) which changes the
shape of the digital twin as it grows.
Our initial conjectures were that a digital twin in We recognize our assumption that data collected
the higher education context we examined would throughout the enrollment process will be accessible to
heavily rely on the interaction of two agents (pursing the twin in real time, as necessary to inform the enrollee
self-agencies): (i) that of the student and (ii) that of the experience and decision-making capability. We assume
institution providing the services through various these data may be held in a data lake. However, with the
personnel. To interact, these two human-centric agents turn to cloud-based systems, data collected through the
would use learning platforms and rely on the capture and touchpoints with the University website, enrollment
(re)use of data to inform each other on how they were system, finance system, student records and so forth,
progressing in the learning journey (see Figure 4). may be held separately by service vendors, or may be
subject to confidentiality or privacy concerns which
might otherwise impede the twins’ access.
Secondly, our current twin conceptualization
assumes the student’s pathway through the enrollment
process follows the most logical and rational order, with
a determinate number of decisions available at each
stage. The student may make choices along the journey,
such as not checking their email or returning their offer
of study, which might prompt background processes to
Figure 4. Initial “straw man” concept generate further touchpoints as university staff (or
systems) intervene. These potential background
We present our straw man in (Figure 4) as an processes are invisible to us as we explore our use case
initial digital twin concept that links students, their from a student perspective. A key insight here is that the
experiences, and touchpoints with learning platforms, structure of the twin (including the sources of data
capture of experiential data, and feedback for decision informing the twin) are shaped by the choices of the
making to both students and education services student, as it is not pre-determinable what the enrollee
providers. Thus, there are opportunities for the agents in will do or when they will do it (despite that we know
our model to have constant feedback based in their what they should do a nd when they should do it). A
experiences and make better decisions that shapes their further insight is that the twin in this context will always
agencies. This creates a reinforcing cyclical feedback reflect the student perspective, and as such there will be

Page 5648
certain ‘black box’ systems which might shape the 5. Discussion of Findings
student journey but may not otherwise be “modellable”
in an educational services twin. A clarified twin concept In this section we discuss insights delivered from
should recognize the complexity of apparently simple our straw man process and arrive at an answer to our
interactions, allowing for new data sources as the twin guiding research question: How can the concept of
mirrors the student and increments itself. digital twinning be developed in an educational services
Our use case analysis also highlights a key strength context? We also propose avenues for developing
in our concept, which is that the role of student agency further research in the form of moving the digitalization
and experience is promoted as a main actor in their frontier and seeking new digital opportunities.
education journey. An incremented digital twin in this Incrementing our twin in the ‘Clarify’ stage of
case should elaborate on student agency which may our adapted methodological approach delivered
influence future iterations. important insights about the general applicability of
digital twins in the educational services environment.
4.3 Clarification Firstly, we extend on findings from the literature and
discover that a further important difference between
Based on our critique of our initial straw man twins in service contexts and twins in manufacturing is–
concept, we clarified how the digital twin can be applied in the case of the latter, that the object of the twin does
to the education services in the case that we explored not have agency which fundamentally shapes the twin.
(see Figure 7). We observed that, unlike the application This finding represents a valuable insight: this exercise
of digital twins in manufacturing (Figure 1), we will not demonstrates that further research exploring this
get the same results from the same processes and concept, while challenging, represents an important
procedures because of the effect of student agency and opportunity for service improvement.
human complexity at the various touchpoints. We can Our incremented digital twin conceptualization
improve the “experiences” for both the institution and also revealed limitations of twinning in the educational
the individual student by exploring how individual services environment more generally. We learn it is not
preference influences the digital twin. We present our possible to have a perfect digital twin as, unlike cars and
clarified digital twin to offer deeper insight into the snack foods, we are not able to control all the variables
elements of an educational services digital twin in this that influence student behaviors through systems. There
context. This incremented twin captures the reality of a is no easy answer to explain why students choose to
twin existing in a digital space, capturing data from behave in the way they do when they interact with
systems determined by necessity (for example, university systems. Often the factors and reasons are as
accessing the learning platform during a course) and complex as the people themselves.
because of student agency (such as the additional Having a digital twin of an educational journey
touchpoints generated by an individual student’s can provide further insights and access to data at
interests or decisions). The digital twin also improves individual, group, and institution levels of analysis. At
management and operational activities on the the individual level (the student) a digital representation
organizational side, as it delivers descriptive, predictive, of their education context can be used simplify complex
and prescriptive analytics. encounters, provide near real time data to enhance
This clarified digital twin better captures how student agency and autonomy. At the group academic
an implemented twin will increment itself to continually educator level digital twins can be used to model
deliver insights about its proxy. cohorts, observe behavioral trends of a group, and
provide rapid insight for agile and effective systems. At
the institutional level, digital twins can give rapid and
accurate insights into asset utilization and management,
student success and completion rates and provide
visibility into aspects of educational processes that
would have otherwise been lost in the noise of a
complex service environment. What we can do is
explore the processes the students engage with. Instead
of a true digital twin the process becomes a proxy that
may indicate where touch points are necessary.
Rather than a strict conceptualization of being a
Figure 7. Revised straw man, i.e., second digital identical twin the metaphor may well be that of a
iteration concept digital fraternal twin – made of the same stuff but not
quite identical enough to be an exact analogue – they

Page 5649
still have something to tell us about the journey, but so The use of digital twins in services is still new.
many things we can’t control for. The focus has mostly been on its application in
manufacturing. Thus, in services industry such as
6. Limitations and opportunities education, research literature is understandably sparse.
The intent of this paper is to provide a stimulus for
The “straw man concept” used in this paper (Figure conversation about digital twins in services generally
7) is exploratory in nature. The straw man approach is and in education services. To give ourselves and others
applied to deliver an interesting and logically sound something to react to we used the straw man approach
concept as its end goal. Therefore, this examination of to provide a foundational illustration of what is possible
how the digital twin concept typical in manufacturing using one modest use case from the large corpus of uses
and physical goods can be applied to services generally cases typically to be found in any university.
and education. We do not claim this to be a By engaging ourselves with the digital twin
generalizable design template. While the outcome of our concept we have modelled user interaction and feedback
conceptualization exercise is necessarily tentative, and that can be generated and used for decisions. The digital
therefore limited, we also believe that it provides a twin opens-up to the possibility of integrating real-time
useful starting point that will provoke and inform future student experiences in computational decision-making
rigorous research in this area. processes and thereby get immediate localized
Currently, the Parmar et al.’s (2020) guidin g information to and about specific individuals, cohorts,
framework is not a series of testable constructs. A and teachers which considers their current university
Critical Success Factor (CSF) analysis could be used to pathways, schedules, preferences, and experiences.
move the framework from principles to constructs that On the organizational side, there is an
could be used to develop an empirical model. opportunity to review systems architecture, processes,
We expect to move forward with iterations that and organizational structure to identify how the systems
involve extensive data collection and experimentation to create, retrieve, update, and dispose of data about and
further develop our concept into testable and for students.
scientifically applicable models, ontologies, and design
science proposals. Future work will also require a more
References
grounded capture of rich data from student experiences
as well as explication of criteria and specifications for Berisha-Gawlowski, A., Caruso, C., & Harteis, C. (2021). The
constructable IS artefacts. This will be paired with concept of a digital twin and its potential for learning
experiments that confirm our assumptions and organizations. In: Digital Transformation of Learning
speculation about the value that can be derived by Organizations (pp. 95-114). Springer, Cham.
creating both students’ agency and improved decision Biesinger, F., & Weyrich, M. (2019). The facets of digital
making for education service providers. twins in production and the automotive industry. In: 23rd
Apart from our own research program on applying International Conference on Mechatronics Technology
digital twins to education services, we hope that this (ICMT) (pp. 1-6). IEEE.
paper establishes a trajectory for future research by Cantalupo, C., Eastep, J., Jana, S., Kondo, M., Maiterth, M.,
Marathe, A., ... & Trinitis, C. (2018). A strawman for an
others. hpc powerstack. Technical Report No. LLNL-TR-
756268. Intel Corporation and Lawrence Livermore
7. Conclusions National Lab. (LLNL), United States. Retrieved from:
[Link]
This paper has shown that initiating a digital twin Dai, R., & Brell-Çokcan, S. (2022). Digital twins as education
support in construction: a first development framework
in educational services can help with identifying
based on the Reference Construction Site Aachen West.
problems and gaps in systems that students interact with
Construction Robotics, 6(1), 75-83.
on campus. Digital twins in education may become a El Saddik, A. (2018). Digital twins: The convergence of
useful tool at the individual level for enhancing student multimedia technologies. IEEE Multimedia, 25(2), 87-9.
agency and to help to build student autonomy through Elayan, H., Aloqaily, M., & Guizani, M. (2021). Digital twin
university processes. For administrators, identifying for intelligent context aware IOT healthcare systems.
areas where students need advice and assistance can IEEE Internet of Things Journal, 8(23), 16749-16757.
lead to better services and staffing. In addition, digital Glaessgen, E., & Stargel, D. (2012). The digital twin paradigm
twins can be used to collect data for critically analyzing for future NASA and US Air Force vehicles. In: 53rd
Structures, Structural Dynamics, and Materials Conf.,
the systems and processes of the wider education
AIAA (p. 1818).
delivery system.

Page 5650
Grieves, M., & Vickers, J. (2017). Digital twin: Mitigating Sepasgozar, S. M. (2020). Digital twin and web-based virtual
unpredictable, undesirable emergent behavior in complex gaming technologies for online education: A case of
systems. In Transdisciplinary Perspectives on Complex construction management and engineering. Applied
systems (pp. 85-113). Springer, Cham. Sciences, 10(13), 4678.
Hattis, D., Baird, S., & Goble, R. (2002). A straw man Shewhart, W. A., & Deming, W. E. (1986). Statistical method
proposal for a quantitative definition of the RfD. Drug from the viewpoint of quality control. Courier Corp.
and Chemical Toxicology, 25(4), 403-436. Tao, F., Qi, Q., Wang, L., & Nee, A. Y. C. (2019). Digital
Ketzler, B., Naserentin, V., Latino, F., Zangelidis, C., twins and cyber–physical systems toward smart
Thuvander, L., & Logg, A. (2020). Digital twins for manufacturing and industry 4.0: Correlation and
cities: A state of the art review. Built Environment, 46(4), comparison. Engineering, 5(4), 653-661.
547-573. Think Insights. (June 7, 2022). Straw Man Proposal:
Liljaniemi, A., & Paavilainen , H. (2020). Using digital twin Brainstorming, the McKinsey Way. Retrieved from:
technology in engineering education–course concept to [Link]
explore benefits and barriers. Open Engineering, 10(1), brainstorming-the-mckinsey-way/.
377-385. Van der Valk, H., Hunker, J., Rabe, M., & Otto, B. (2020).
Parmar, R., Leiponen, A., & Thomas, L. D. (2020). Building Digital twins in simulative applications: a taxonomy. In:
an organizational digital twin. Business Horizons, 63(6), 2020 Winter Simulation Conference (WSC) (pp. 2695-
725-736. 2706). IEEE.
Rathore, M. M., Shah, S. A., Shukla, D., Bentafat, E., & Veeraraghavan, M., Li, J., & Reisslein, M. (2011). A
Bakiras, S. (2021). The role of AI, machine learning, and strawman proposal for future diverse internets. In 2011
big data in digital twinning: A systematic literature IEEE Symposium on Computers and Communications
review, challenges, and opportunities. IEEE Access, 9, (ISCC) (pp. 792-795). IEEE.
32030-32052. Zhang, H., Ma, L., Sun, J., Lin, H., & Thürer, M. (2019).
Saracco, R. (2018). The rise of digital twins. IEEE Future Digital twin in services and industrial product service
Directions. 2018.01.16. systems: review and analysis. Procedia CIRP, 83, 57-60.

Page 5651
9. Appendix A

Figure 5. Use Case 1

Figure 6. Use Case 2

Page 5652