Human-Intelligent Systems Integration (2022) 4:23–33

https://doi.org/10.1007/s42454-022-00041-x

REVIEW ARTICLE

A unified view of a human digital twin

Michael E. Miller1 · Emily Spatz2

Received: 16 October 2021 / Accepted: 30 January 2022 / Published online: 12 March 2022
This is a U.S. government work and not under copyright protection in the U.S.; foreign copyright protection may apply 2022

Abstract
The term human digital twin has recently been applied in many domains, including medical and manufacturing. This term
extends the digital twin concept, which has been illustrated to provide enhanced system performance as it combines system
models and analyses with real-time measurements for an individual system to improve system maintenance. Human digital
twins have the potential to change the practice of human system integration as these systems employ real-time sensing and
feedback to tightly couple measurements of human performance, behavior, and environmental influences throughout a prod-
uct’s life cycle to human models to improve system design and performance. However, as this concept is relatively new, the
literature lacks inclusive and precise definitions of this concept. The current research reviews the literature on human digital
twins to provide a generalized structure of these systems, provide definitions of a human digital twin and human digital
twin system, and review the potential applications of these systems within product design, development, and sustainment.
This review of the existing literature suggests that components of human models sufficient to provide robust human digital
twins are likely to be derived across multiple fields of study. Thus, development of these systems would benefit an open
multi-disciplinary research effort.

Keywords Human digital twin · Model-based systems engineering · Human performance

1 Introduction be different, placing further stress on logistics for sustain-

ment of these products. Digital twin technology has been
An outcome of the industrial revolution and mass production developed in recent years to address the problems which
is the standardization of products. This process has many arise from product and environmental differences (Tuegel
advantages, including the ability to create many product et al. 2011). The digital twin concept includes constructing
copies from a single design and the ability to leverage scale a digital representation or model of an individual product
to reduce overall product cost. Another advantage is the to improve the accuracy of maintenance and performance
ability to use standardized parts during system operation predictions for individual products (Kobryn 2020). Thus, a
and sustainment to reduce the cost of maintenance. While digital twin has been described as the model of a component,
the variability between products is controlled during manu- product, or system developed by a collection of engineering,
facturing, their use often varies, increasing the differences operational, and behavioral data which support executable
between products. If these differences are not accounted models, where the models evolve over the lifecycle of the
for, the variability decreases the ability to properly esti- system and support the derivation of solutions which assist
mate product performance and plan maintenance cycles. the real time optimization of the system or service (Boschert
Furthermore, specific products within a product family can and Rosen 2016).
Recently, this term has been extended to humans using
the term “human digital twin.” This term has been applied
* Michael E. Miller in diverse fields, including medicine (Chakshu et al. 2019;
[email protected]; [email protected] Corral-Acero et al. 2020; Hirschvogel et al. 2019; Liu et al.
1
Department of Systems Engineering and Management, Air
2019; Lutze 2020), sports performance (Barricelli et al.
Force Institute of Technology, 2950 Hobson Way (AFIT/ 2020), manufacturing ergonomics (Caputo et al. 2019;
ENV), Wright Patterson AFB, OH 45433, USA Greco et al. 2020; Sharotry et al. 2020), and product design
2
Defense and Intel, KBR Government Solutions – U.S., 3725 (Demirel et al. 2021)(Constantinescu et al. 2019; Demirel
Pentagon Blvd Suite #210, Beavercreek, OH 45431, USA

13
Vol.:(0123456789)
24 Human-Intelligent Systems Integration (2022) 4:23–33

et al. 2021). Although the human digital twin concept may be system lifecycle. The overall goal of this paper is to sup-
analogous to digital twins of products, there are distinct dif- port the development of this technology by proposing com-
ferences, including increased underlying variability between mon terminology and a vision for applying this technology
humans and the fact that humans often employ products to to support product development and deployment, with an
achieve their goals. Commonalities can be observed in the emphasis on product development and deployment within
use of the digital twin concept across the fields where it the United States Department of Defense acquisition system.
has been applied. For example, each of these fields dis- It is hoped that socialization of a common definition and
cusses constructing and applying a model of humans which potential applications of this technology within the human
is informed by data collected from sensors which provide factors, ergonomics, and human systems integration com-
insight to human behavior and performance in real world munities will lead to synergistic efforts which improve the
settings. However, there are also differences. For example, development of this technology.
in medicine or sports performance, the models often focus
on improving the human or human performance while in
manufacturing and product design, the model’s focus is on 2 Method
improving either the process or the artifacts with which the
human interacts. As a result, the components which are mod- The current research consisted of (1) conducting a review of
eled within each of these applications differ. literature pertaining to human digital twins, (2) developing
A clear example of a human digital twin from the lit- a structural model of a human digital twin based upon the
erature pertains to manual material handling within manu- components of human digital twins from the literature, (3)
facturing or warehousing applications (Greco et al. 2020; applying this literature to derive an inclusive definition, and
Sharotry et al. 2020). In this example, sensors are deployed (4) synthesizing and developing use case models based upon
with the human in their environment which monitor the potential applications inferred by the literature.
human’s kinematic motion as they perform work within the
environment. Other information, such as objects being lifted, 2.1 Literature review
or forces required to activate items within the environment
are also collected. This data populates simulation models The literature search included a Google Scholar search on
which estimate the fatigue in various muscle groups of the the term “human digital twin” which was complemented by
operator. These fatigue estimates support assessments and a search on “digital twin” and either “human,” “man,” or
potential changes in work or rest schedules, handling pro- “woman.” Each search was limited to dates between 2011
cesses, or material handling tools to improve overall worker and July 2021. The search on human digital twin returned
health, safety, and productivity. Thus, a model of a human 99 articles while the complementary search returned 15,000
attribute is applied within a closed-loop system which ful- results. Each of the human digital twin and the abstracts of
fills the criteria for a digital twin (Boschert and Rosen 2016). the first 1000 articles on digital twin was then reviewed. As
While models of human behavior, performance, and state the purpose of this paper was predominantly to understand
have been developed for decades, the key concept of human the structure and application of human digital twins to sys-
digital twins involves including these models within a tightly tem design, development, and operation, the abstract review
coupled, closed-loop system, where this system permits the determined if the articles discussed the creation or applica-
model to evolve with data from a specific real-world human tion of digital twins which included a model of a human as
as they act within an environment and to provide feedback part of the digital twin. A total of 54 unique papers which
with the goal of enhancing the performance and well-being included a model of a human either as the digital twin or as
of the human to which it is coupled. a component of a digital twin were identified. For example,
While the applications for human digital twin technolo- a model of a human may represent an athlete where perfor-
gies differ, advances in one field may inform application mance is affected by nutrition, rest, and prescribed physi-
of this technology in other fields. In fact, it is common for cal activity while a model with a human as a component
models which were developed to understand human perfor- may include a workstation in a factory where performance
mance for medical or sports applications to inform other is influenced by factors such as nutrition, rest, and activity
applications, such as product design. Therefore, the current of the human but also by attributes of the system in which
research seeks to leverage a review of existing literature they are performing, such as material handling equipment
regarding digital twins, and more specifically human digital and procedures.
twins, to derive a common definition of the human digital Tables 1 and 2 show the application space and the pub-
twin, a taxonomy of the components useful for construction lication year for the resulting papers. As shown, most of
of the human digital twin, and use cases of this technol- the papers either pertained to the manufacturing or medical
ogy within product development and sustainment across the domains. Four additional papers were focused specifically on

13
Human-Intelligent Systems Integration (2022) 4:23–33 25

Table 1  Application areas for Friedenthal et al. 2014). Diagrams are also provided for
human digital twins based on Manufacturing 26
structuring information (i.e., package diagrams) and sup-
literature review Medical 14
porting mathematical modeling (i.e., parametric diagram).
Biomechanics applications 4
SysML was chosen to represent aspects of human dig-
Sports and fitness 4
ital twins as the digital twin concept has been described
General product design 2
as a method to extend Model-Based Systems Engineering
Military 1
(MBSE) from the design to the operation and sustainment
Disaster management 1
phase of a product (Boschert and Rosen 2016) and SysML
Organization design 1
has been developed to support MBSE. Furthermore, this
Identity management 1
language is an object-based language which permits the
description of a generic system, sometimes through a refer-
ence architecture, and the extension of this description to
Table 2  Publication year for specific instances or instantiations (King et al. 2020) and
papers in literature review 2016 1
has been used to model the human’s contribution to sys-
2018 4
tem performance (M. Watson et al. 2017a, b; M. E. Watson
2019 16
et al. 2017a, b). Thus, the generic system description pro-
2020 23
vided through the diagrams in this paper can be extended to
2021 10
illustrate the application of human digital twins in specific
applications or instances.
Each of the articles pertaining to human digital twins
understanding and modeling human biomechanics, predomi- were analyzed to determine the structural components and
nantly with a focus on reducing musculoskeletal injuries in use cases they discussed. These elements were then used
the workplace. Four additional papers focused on health to compose a SysML block definition diagram to illustrate
and fitness applications, primarily for sports training. The the infrastructure for a human digital twin. Furthermore,
remaining papers appeared to uniquely address five addi- additional components were added based upon additional
tional application areas, including general product design functions which were implied by the discussion of human
(Demirel et al. 2021), military application (Li et al. 2020), digital twins. Similarly, use cases were extracted from each
disaster management (Fan et al. 2021), organization design of the articles from the literature review. These use cases
(Parmar et al. 2020), and identity management (Zibuschka were interpreted with respect to product design and applica-
et al. 2020). tion within the United States Department of Defense (DoD).
As shown in Table 2, practically all the publications Separate use case diagrams were then developed for the
appeared within the past 4 years. The earliest paper does application of knowledge gained from human digital twins
not mention the words human digital twin, but discusses to product design and operation. While the use cases are
the construction of a digital athlete, which is updated based specific to the DoD, which acquires products for its own
upon sensed information from the application environment use, many of these use cases can be generalized to other
(Alderson and Johnson 2016). The earliest papers using organizations where the organization responsible for product
the term “Digital Twin” while referring to human models acquisition and the organization employing the product may
were published in 2018 with one of these papers appearing differ. Overall, these diagrams define the general compo-
in the medical literature (Bruynseels et al. 2018) and three nents and product development-oriented use cases of human
appearing within the manufacturing literature (Graessler and digital twins.
Poehler 2018a, 2018b; Kemény et al. 2018).

2.2 Modeling structure and use cases

3 Structural definition of a human digital
The structure and use case models of human digital twins twin
were created using the Systems Modeling Language
(SysML). This language provides a standardized language As shown in the block definition diagram of Fig. 1, digital
and family of diagrams for representing systems during twin systems typically include a real-world entity, i.e., the
design and development. The SysML diagrams include dia- real-world twin, and a digital twin where the digital twin is
grams for representing system requirements (i.e., require- a representation of one or more attributes of the real-world
ments diagram), structure (i.e., block definition and inter- twin. The real-world and digital twins are linked through an
nal block diagrams), and behavior (i.e., use case, activity, interchange component such that changes in one twin can
state machine, and sequence diagrams) (Delligatti 2013; produce changes in the other (Barricelli et al. 2020; Grieves

13
26 Human-Intelligent Systems Integration (2022) 4:23–33

Fig. 1  Block definition diagram

showing primary components of
a digital twin system

2015; Latif and Shao 2020; Mendi et al. 2021; Sharotry et al. a process analogous to the perceptual loop (Neisser 2014).
2020; van der Valk et al. 2020). However, in a digital twin system, this sequence includes a
This interchange component permits the digital twin to be repetitive sequence with the following stages:
updated as the real-world entity acts within its environment.
The digital twin then executes embedded models which are 1. Sensors on the real-world twin sense the actions and
informed by the data it receives from the real-world twin to performance of the real-world twin within its real-world
simulate one or more aspects (e.g., structure or behavior) of environment, as well as relevant state information about
the real-world twin with the goal of determining changes in the real world.
the real-world twin which might produce an improvement 2. The interchange component conveys the sensed data to
in the real-world twin. the digital twin.
sSharotry and colleagues discuss a digital twin system 3. An analysis is performed to determine whether the data
in a manufacturing environment which consists of a human is consistent with the digital twin’s current model pro-
conducting material handling tasks (Sharotry et al. 2020). jections and adjusts or builds a model to explain any
The system includes a data collection module which includes differences.
motion capture, biometric suites among other sensors, a data 4. The model is applied to create projections of future
analysis and forecasting model, and a database which con- behavior within a virtual environment.
tains the data which is collected for analysis. In this system, 5. The projection of future behavior is compared to a goal
performance and operator fatigue metrics are provided to the state.
individuals in the environment to support improved manu- 6. Based on this analysis, the system determines if a mod-
facturing performance and to help reduce human injury. ification to the structure or behavior of the system is
Furthermore, this model can be used to identify material likely to move the system towards a goal state, and if
handling steps which induce substantial fatigue, permit- so, selects a modification to the physical element or its
ting these steps to be evaluated and redesigned. Within this behavior which moves the real-world twin towards the
example, the real-world twin includes the manufacturing goal state.
environment, the human, and the data collection subsystems. 7. The modification is conveyed to the real-world twin and
The database provides the interchange component, and the applied to the real-world twin.
data analysis and forecasting model provides the digital rep-
resentation or the digital twin. This series of steps is repeated, adjusting the digital mod-
As illustrated by these examples, human analysts apply els in the digital twin and the real-world twin to move the
the digital twin to conduct analysis, develop courses of system towards its goal state. Ideally, the coupling of the
action, and then apply the courses of action in the real world. digital and real-world twins by the interchange component
The courses of action identify potentially useful changes in provides the ability to rapidly create robust models of real-
the structure or behavior of the real-world twin. In these sys- world performance, permit these models to be applied to
tems, the analyst or the interchange component then causes project future real-world performance, and to improve the
changes in the real-world twin, which ideally improves performance of the real-world system.
the performance of this entity. Importantly, the two-way To create accurate predictions, the models represent not
exchange between the digital and real-world twins provides only the real-world entity but the environment in which the
the digital twin the ability to sense the real world, creates real-world twin is acting. This fact is not evident in much
an understanding of the world, and acts upon it, completing of the literature as the digital twin concept is applied within

13
Human-Intelligent Systems Integration (2022) 4:23–33 27

controlled environments. Thus, these systems were assumed Within the literature, the real-world twin includes a
to be closed; that is, they do not interact significantly with human within the environment. Sensors are used to pro-
their environment. However, systems are often open, vide real-time or near real-time information about the
readily exchanging information, energy, and matter with human and the environment (Corral-Acero et al. 2020;
their environment (Blanchard and Fabrycky 2006). Thus, it Sharotry et al. 2020) and this real-time feedback may be
becomes important to model the environment to understand augmented with non-real time feedback, such as medical
the entity’s interface with the environment. Specifically, the diagnoses, professional electro-cardiograms, or informa-
interface between these components require understanding tion from imaging devices (Corral-Acero et al. 2020).
of the spatial configuration, energy, information, and Additionally, the human can be queried to input subjective
material transfer between the environment and the system information (e.g., mood) or information which is difficult
(Jain et al. 2010). However, from an architectural standpoint, to track (e.g., nutrition information) into tracking applica-
the environmental model does not necessarily need to tions (Barricelli et al. 2020). In manufacturing and product
be part of the digital twin. Instead, the environmental design scenarios, the human also interacts with physical
model may be separated from the digital twin system and devices (i.e., system), such as manufacturing equipment,
simply be associated with this system. This is a significant material to be handled, or products which can be instru-
architectural decision as the human digital twin system will mented (Barricelli et al. 2020). Additionally, the real-world
not necessarily require a broad model of the environment but twin typically includes a process or procedure (Latif and
only requires models which reflect the information, energy, Shao 2020). Although the process is not a physical item, it
and matter that the system exchanges with the environment. is important in understanding the behavior of the human.
Therefore, the complexity of the environmental model may Finally, while not discussed explicitly in the literature, it
be constrained if it is constructed to capture the necessary may also be important for some mechanism to be present
elements as opposed to the creation of a more generic and to modify the physical environment. The processes may be
encompassing model of the environment. The real-world changed but to change the physical environment, it is nec-
twin must also determine the state of the environment and essary to include physical mechanisms, shown as actuators
communicate this state to the digital twin. in Fig. 2. As shown, near the head of the arrows associ-
Reviewing the literature, different components of the dig- ated with these entities, the human digital twin systems,
ital twin system are illustrated by different authors. There- by definition, include a human, one or more sensors, and
fore, it is possible to integrate the concepts across these at least 1 process. The machine and actuators are optional
papers to provide an architectural view of a generic digital components as they do not exist in all systems, including
twin. Figure 2 shows a more detailed block definition dia- the sports systems.
gram based upon an analysis of the digital twin systems in Of the three components, the interchange component
the literature. As shown earlier, this system is comprised of receives the least attention within most of the published
a real-world twin, a digital twin, and an interchange compo- literature. Although the literature is clear that this inter-
nent. However, alluded to before, the real world and a repre- change component permits two-way communication
sentation referred to as the virtual world are each associated between the real-world entity and the digital entity, its
with this system, where these two entities are linked by their components are not clearly discussed. However, beyond
state information. communications, this component can be envisioned as the

Fig. 2  Block definition diagram showing an exploded view of a generic human digital twin system

13
28 Human-Intelligent Systems Integration (2022) 4:23–33

warehouse for data or the database, although the database digital twin. It is clear based upon the literature that a
is sometimes viewed as part of the digital twin (Latif and human digital twin system (HDTS) should include a digi-
Shao 2020). Furthermore, this component can be responsi- tal representation of a real-world human. This real-world
ble for analysis to clean the data, fusing information from human may be an individual, whom we wish to character-
the various sensors associated with the real-world twin and ize and model or a human class, where this human class
integration of new data with the existing data (Tao et al. represents a group of humans with common traits, charac-
2019a). teristics, behavior, etc. The human digital twin (HDT) is
Finally, we can review the components of the digital then the digital representation of the real-world human or
twin within the digital twin system. The digital twin cer- real-world twin. The HDT may exist purely as a mathemat-
tainly contains a model of the human. This model can ical representation of the individual or class of individuals.
include mechanistic models, which model well under- Alternately, the HDT may exist as a virtual entity which
stood physical, chemical, biological, and physiological can be rendered within a virtual or real-world system.
processes, as well as, statistical models which rely upon The digital representation can include both first princi-
the data collected from the real-world twin (Corral-Acero ples models which are based on fundamental understand-
et al. 2020). It can optionally contain additional models ing and statistical models. Various attributes of one or
of the machine or the process which is to be performed more humans can be modeled (Alexander et al. 2020).
(Latif and Shao 2020; Tao et al. 2019a). Also necessary is These include attributes in at least one of the following
a prediction engine which generates perturbations to the categories:
virtual world, human, machine, or process to understand
how changes in these elements affect the performance a Physical: including anthropometric attributes, biome-
of the human or the system (Tao et al. 2019a). To assess chanics attributes, times required for task completion,
the outcomes, a performance goal is also required such eye movements, and injuries
that the results produced by the prediction engine can be b Physiological: measures such as heart rate, heart rate
assessed. A mission description is also likely necessary variability, galvanic skin response, muscle tension,
which informs the process model of potential process steps blood oxygen level, brain electrophysiologic signals,
which are considered viable. pupillometry, blink rates and timing, peripheral blood
Although not shown in Fig. 2, it is also common for a pro- flow, and gastronomical activity. However, these are
cess manager to interact with the digital twin to determine if intended to correlate to higher level physiological meas-
any of the modifications evaluated by the prediction engine ures such as fatigue, circadian cycle, alertness level, and
should be applied to the real-world twin. While these modifi- activation or engagement level.
cations may be communicated through the interchange compo- c Perceptual performance: auditory sensitivity, ability to
nent to the real-world twin, physical changes to the real-world decipher speech in different languages, visual sensitivity,
twin may require human intervention. color sensitivity, contrast sensitivity, pressure sensitiv-
It is worth noting that during design, the real-world twin ity, pain thresholds, temperature sensitivity, etc.
may be under development and therefore the real-world d Cognitive performance: knowledge, skills, abilities or
twin or even a prototype of the real-world twin may not aptitudes, workload level, situation awareness, decision-
exist to participate in the digital twin system. Under these making abilities, intuitive/analytic bias, etc.
circumstances, and perhaps others as well, it can become e Personality characteristics: attributes such as personality
useful to include the visualization engine, as shown in Fig. 2 type, propensity to trust, and propensity towards suspi-
(Havard et al. 2019; Ma et al. 2019; Wang et al. 2021). This cion.
visualization engine can provide a virtual or augmented f Emotional state: feelings experienced by the individual
reality rendition of the digital twin in the virtual or real within their current circumstance, including levels of
worlds which a designer or potential user can interact with depression, and anxiety.
to improve insights and understanding until the real-world g Ethical stance: representation of the individual’s belief
twin is available. system, including values, beliefs, and mores.
h Behavior: actions taken by the individual to interact with
the system. These actions may be influenced by any of
4 Defining human digital twin and human the earlier attributes.
digital twin systems
As such, we can define a human digital twin as an inte-
Based upon the prior discussion, we can attempt to define grated model which facilitates the description, prediction,
a human digital twin and a human digital twin system. or visualization of one or more characteristics of a human
We need to consider several aspects to define the human or class of humans as they perform within a real-world

13
Human-Intelligent Systems Integration (2022) 4:23–33 29

environment. A human digital twin system is a pairing of and validation, manufacturing, and use. Although HDTS
a real-world human twin and a human digital twin which might be useful during any phase of the lifecycle, Fig. 3
includes a model of physical, physiological, personality, and Fig. 4 depict the use cases which apply during product
perception, cognitive performance, emotion, or ethics of a development or acquisition, and operations and sustainment,
human; where the real-world human and human digital twin respectively.
are integrated such that a change in the real-world human As Fig. 3 shows, during the product development and
or its digital representation produces change(s) in the other. acquisition life cycle phase, HDTS can be used to aid the
This human digital representation likely includes both design of various portions of the system to be acquired.
mechanistic and statistical models of one or more attributes This might start with simulation of the human operating
related to one or more of the human defined properties. the product within the system of systems (SoS) to under-
stand operational needs and the benefits of enhancements.
As these needs are defined, modifications or new prod-
5 Human digital twin systems use cases uct concepts may be developed and evaluated (Laamarti
2019). In this use case, models which exist within the
Besides defining HDTS and investigating their structure, it is human digital twin can be applied early in system devel-
also useful to understand how they might be applied. Within opment to evaluate the utility of various designs and to
the literature, use cases are discussed within the health analyze the impact of these differences through trade space
industry where HDTS support well-being, health care, and analysis (Tao et al. 2019b). This might include evalua-
development of medical devices (Laamarti et al. 2020). tion and understanding of various manpower, personnel,
Additionally, use cases are discussed for the design and man- and training options, as well as options regarding the per-
ufacturing of products, including design of human–machine formance or other attributes of the system which impact
interaction, as well as the development of product service important emergent behaviors, i.e., “ilities” of the system.
and use (Ma et al. 2019). Furthermore, these models might consider modifications
For product development and application, use cases may to the user interface, items which enhance human physical
be considered for different product lifecycle phases; that or cognitive performance, and personal protective equip-
is, they may be useful during product design, verification ment (PPE) to prevent performance degradation. Perhaps,

Fig. 3  Example product devel-

opment and acquisition-based
use cases

13
30 Human-Intelligent Systems Integration (2022) 4:23–33

Fig. 4  Operations and sustain-

ment use cases imagined based
upon literature

some models might be applied to understand the trade- material components of the system to focusing on changes
offs in operator or maintenance procedures to aid mis- in the tactics, techniques, and procedures (TTPs) which are
sion design or to optimize the performance of operators employed when applying the system. Similarly, these models
or maintainers within the system. As further indicated in may be used to evaluate and improve training or personnel
this diagram, these models and HDTS will likely require selection methods (Ma et al. 2019). Additionally, the human
sharing and collaboration among different communities digital twin may be used to understand readiness of person-
which interact during product design and acquisition. nel who interact with the systems or understand the errors or
It should be noted that early in a development cycle of circumstances in which errors are made (Tao et al. 2019b).
a novel product, the real-world portion of the HDTS will Finally, the HDTS may be applied in real time to determine
undergo design, and therefore, the real-world system or and allocate taskwork among team members (Ma et al.
representative, trained operators may not exist. Thus, a 2019). The corresponding use cases are shown in Fig. 4.
representative real-world system will not be available. Dur- As shown in Fig. 4, as the HDTS is deployed with the
ing these portions of the project, the digital portion of the system to the field, we expect that the operators will develop
HDTS will be required to operate in an open loop fashion. and employ new or revised TTPs which increase the util-
However, as prototypes of portions of the system are devel- ity and performance of the product (Cox and Szajnfarber
oped and early proxies for system operators and maintainers 2017). The resulting innovations help to redefine taskwork
are used to participate in demonstrations, simulations, or and improve training. Furthermore, the models themselves
evaluations, data from these events can be used to begin may be used to explore changes to taskwork beyond those
to close the loop from the real-world to the digital models captured from observing user behavior, as observed in the
for the system and the humans. As this data is included, the manufacturing-oriented literature. Additionally, the HDTSs
level of model fidelity and user confidence in these models may be employed to analyze and understand human errors,
should improve. Besides changes in model fidelity, the level which result in either near misses or mishaps to suggest
of detail represented by the models is likely to change as one changes in TTPs, training, or taskwork to improve safety.
progresses through the system development and acquisition Furthermore, utilization of personal protective equipment,
process from early needs development through acquisition. as well as environmental exposure to information, energy, or
Similar models can also be applied during development of matter (e.g., radiation or known carcinogens), may be moni-
the manufacturing and logistics processes to affect human tored to support occupational health interventions. Further-
interaction as the real-world twin is developed. more, the HDT may be capable of understanding attributes
As the system enters operations and sustainment, the of the humans which enhance performance, aiding refine-
use cases depicted in Fig. 4 become relevant. In this phase, ment of personnel selection.
the focus of HDTS shift from focusing on changes to the

13
Human-Intelligent Systems Integration (2022) 4:23–33 31

Although TTPs, training, and personnel selection may all As shown, the HDTS include a real-world twin, a digital
improve system performance, perhaps the greatest opportu- twin, and an interchange component which facilitates real-
nity is understanding human readiness. However, this use time monitoring of the real-world human as they perform
case likely diverges significantly from the human digital in an environment using the system or product to provide
twins that would be useful during system development. In robust models and predictions which facilitate system
fact, this use case likely relies upon significant interaction improvements. HDTS rely upon recent advances in low-
with the management and medical personnel who are not cost sensors, data storage, and data mining technologies,
as likely to interact with the other use cases illustrated in together with the ongoing conversion to digital engineering
Fig. 4. Although not apparent from our earlier discussion, as components. The integration of these components facili-
the HDT applied in medical and sports performance focus on tates feedback loops which support improvements both in
understanding performance of systems which are internal to the design of products or systems for improved performance
the human while HDT applied in product design or manufac- in real-world settings as well as improvements in TTPs and
turing tend to focus on human behavior and the interaction training to enhance user performance during product or sys-
of the human with systems and the environment. As such, tem operations and sustainment.
there is a basis for arguing that the HDT used for under- Successful implementation of HDTS will require the
standing readiness may be unique from the HDT used for development of robust models of various aspects of the
practically all other use cases shown in Fig. 3 and Fig. 4. human at different scales for different applications. These
Nevertheless, each of these HDTs relies on input from the models will need to address aspects of the human to include
same real-world humans and the HDT models within these physical, physiological, perceptual, cognitive, behavioral,
two areas certainly interact. Therefore, a HDT developed to emotional, personality, and ethical structure and perfor-
support readiness and HDT developed for other use cases mance. These models will need to describe and account
will likely require some level of integration. Thus, there for the variability among and uniqueness of individuals to
will be a need to integrate models of the human which are be suitable for application within HDTS. The challenge,
intended to understand health or health related human per- however, is the fact that component models which address,
formance with models of the human which are intended to typically narrowly defined, aspects of the human are being
understand their performance when interacting with systems developed within several disciplines for disparate applica-
within an environment. tions as illustrated by the literature review. To accelerate
the deployment of this technology, a concerted, multi-dis-
ciplinary, effort will be required to understand, define, and
6 Summary and conclusion synthesize these models into HDTS applications.

This paper sought to define a common structure, definition,

and set of use cases for HDTS as they pertain to systems, Author contribution Conceptualization: MEM; methodology: MEM;
formal analysis and investigation: MEM, ET; writing—original draft
generally, and to product development, specifically. As dis- preparation: MEM; writing—review and editing: ET; funding acquisi-
cussed, the HDTS terminology has evolved in the past few tion: MEM; resources: MEM; supervision: MEM.
years and is discussed predominantly in research papers
related to medical and manufacturing systems, although Funding Partial financial support was received from the Air Force
other applications including fitness, product design, and Research Laboratories.
military, among others, are present within this literature. Declarations
This research provides a definition, structure, and use
The views expressed in this article are those of the authors and do not
cases for these systems, as defined within the reviewed lit- necessarily reflect the official policy or position of the Department of
erature. However, the development of HDTS as defined in the US Air Force, US Department of Defense, nor the US Government.
this paper are relatively new, and therefore, the definitions,
structure, and use cases are likely to evolve as these systems Ethics approval Not applicable.
are employed and used. While an effort was made to capture Consent to participate Not applicable.
the structure and behavior of these systems within SysML
models and then to integrate these models, these activities Consent for publication This work was prepared by employees or con-
required interpretation by the authors. Therefore, these mod- tractors of the United States Government as part of their official duty,
and thus, the work is not subject to copyright within the USA. The
els may not always represent the intent of the authors of the authors consent to publish this work.
original articles. For each of these reasons, it is hoped that
the definition, structure, and use cases of HDTS provided in Conflict of interest The authors declare no competing interests.
this paper will provide an integrated view from which these
definitions can evolve.

13
32 Human-Intelligent Systems Integration (2022) 4:23–33

Open Access This article is licensed under a Creative Commons Attri- Twin” to enable the vision of precision cardiology. European
bution 4.0 International License, which permits use, sharing, adapta- Heart Journal, 41(48), 4556-4564B. https://d​ oi.o​ rg/1​ 0.1​ 093/e​ urhe​
tion, distribution and reproduction in any medium or format, as long artj/​ehaa1​59
as you give appropriate credit to the original author(s) and the source, Cox A, Szajnfarber Z (2017) System user pathways to change. In Madni
provide a link to the Creative Commons licence, and indicate if changes AM, Boehm B, Wrwin DA, Ghanem R, Wheaton MJ (eds) 15th
were made. The images or other third party material in this article are Annual conference on systems engineering research disciplinary
included in the article's Creative Commons licence, unless indicated convergence: implications for systems engineering research, pp
otherwise in a credit line to the material. If material is not included in 1–9
the article's Creative Commons licence and your intended use is not Delligatti L (2013) SysML distilled: a brief guide to the systems mod-
permitted by statutory regulation or exceeds the permitted use, you will eling language. Addison-Wesley
need to obtain permission directly from the copyright holder. To view a Demirel HO, Irshad L, Ahmed S, Tumer IY (2021) Digital twin-driven
copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. human-centered design frameworks for meeting sustainability
objectives. J Comput Inf Sci Eng 21(3). https://d​ oi.o​ rg/1​ 0.1​ 115/1.​
40506​84
Fan C, Zhang C, Yahja A, Mostafavi A (2021) Disaster city digital
References twin: a vision for integrating artificial and human intelligence for
disaster management. Int J Inf Manag 56(2019):102049. https://​
Alderson J, Johnson W (2016) THE PERSONALISED “DIGITAL doi.​org/​10.​1016/j.​ijinf​omgt.​2019.​102049
ATHLETE”: an evolving vision for the capture, modelling and Friedenthal S, Moore A, Steiner R (2014) A practical guide to SysML.
simulation, of on-field athletic performance. 34th International Elseivier Science and Technology
Conference on Biomechanics in Sports, 23–26 Graessler I, Poehler A (2018a) Integration of a digital twin as human
Alexander T, Huiskamp W, Fromm L, Lewis MD, DiMatteo R, Gun- representation in a scheduling procedure of a cyber-physical
zelmann G, MGroarty C, Gallant S, Blais D, Genc I (2020) Ref- production system. IEEE International Conference on Industrial
erence architecture for human behaviour modelling (STO-TR- Engineering and Engineering Management, 2017-Decem, 289–
MSG-127). https://​www.​sto.​nato.​int/​publi​catio​ns/​STOTe​chnic​ 293. https://​doi.​org/​10.​1109/​IEEM.​2017.​82898​98
al Reports/Forms/Technical Report Document Set/docsethome- Graessler I, Poehler A (2018b) Intelligent control of an assembly sta-
page.aspx?ID=4581&FolderCTID=0x0120D5200078F9E8704 tion by integration of a digital twin for employees into the decen-
3356C409A0D30823AFA16F6010066D541ED10A62C40B2AB tralized control system. Procedia Manuf 24:185–189. https://​doi.​
0FEBE9841A61&List=92d5819c-e6ec-4241-aa4e-57bf918681b- org/​10.​1016/j.​promfg.​2018.​06.​041
1&RootFolder=https://​www.​sto.​nato.​int/​publi​catio​ns/​STOTe​ Greco A, Caterino M, Fera M, Gerbino S (2020) Digital twin for mon-
chnic​al Reports/STO-TR-MSG-127 itoring ergonomics during manufacturing production. Appl Sci
Barricelli BR, Casiraghi E, Gliozzo J, Petrini A, Valtolina S (2020) (Switzerland) 10(21):1–20. https://​doi.​org/​10.​3390/​app10​217758
Human digital twin for fitness management. IEEE Access Grieves M (2015) Digital twin: manufacturing excellence through vir-
8:26637–26664. https://​doi.​org/​10.​1109/​ACCESS.​2020.​29715​76 tual factory replication. In White paper. https://w ​ ww.r​ esear​ chgat​ e.​
Blanchard BS, Fabrycky W (2006) Systems engineering and analysis net/​publi​cation/​27521​1047_​Digit​al_​Twin_​Manuf​actur​ing_​Excel​
(4th edn). Pearson Education, Inc. lence_t​ hroug​ h_V​ irtua​ l_F
​ actor​ y_R​ eplic​ ation. Accessed Sept 2021.
Boschert S, Rosen R (2016) Digital twin - the simulation aspect. Havard V, Jeanne B, Lacomblez M, Baudry D (2019) Digital twin and
In P. Hehenberger & D. Bradley (Eds.), Mechatronic Futures: virtual reality: a co-simulation environment for design and assess-
Challenges and Solutions for Mechatronic Systems and Their ment of industrial workstations. Prod Manuf Res 7(1):472–489.
Designers (pp. 59–74). Springer Nature. https://​doi.​org/​10.​1007/​ https://​doi.​org/​10.​1080/​21693​277.​2019.​16602​83
978-3-​319-​32156-1_1 Hirschvogel M, Jagschies L, Maier A, Wildhirt SM, Gee MW (2019)
Bruynseels K, de Sio FS, van den Hoven J (2018) Digital twins in An in silico twin for epicardial augmentation of the failing heart.
health care: ethical implications of an emerging engineering para- Int J Numer Methods Biomed Eng 35(10). https://d​ oi.o​ rg/1​ 0.1​ 002/​
digm. Front Genet 9(FEB):1–11. https://​doi.​org/​10.​3389/​fgene.​ cnm.​3233
2018.​00031 Jain R, Chandrasekaran A, Erol O (2010) A Framework for end-to-end
Caputo F, Greco A, Fera M, Macchiaroli R (2019) Digital twins to approach to Systems Integration. Int J Ind Syst Eng 5(1):79–109.
enhance the integration of ergonomics in the workplace design. Int https://​doi.​org/​10.​1504/​IJISE.​2010.​029763
J Ind Ergon 71(January):20–31. https://​doi.​org/​10.​1016/j.​ergon.​ Kemény Z, Beregi R, Nacsa J, Glawar R, Sihn W (2018) Expanding
2019.​02.​001 production perspectives by collaborating learning factories - per-
Chakshu NK, Carson J, Sazonov I, Nithiarasu P (2019) A semi-active ceived needs and possibilities. Procedia Manuf 23(2017):111–
human digital twin model for detecting severity of carotid sten- 116. https://​doi.​org/​10.​1016/j.​promfg.​2018.​04.​002
oses from head vibration—a coupled computational mechanics King D, Jacques D, Gray J, Cheney K (2020) Design and simulation of
and computer vision method. Int J Numer Methods Biomed Eng a wide area search mission: an implementation of an autonomous
35(5):1–17. https://​doi.​org/​10.​1002/​cnm.​3180 systems reference architecture. In K. H. Bae, F. S. Kim, L.-M. S.,
Chakshu NK, Sazonov I, Nithiarasu P (2021) Towards enabling a Z. Zheng, T. Roeder, & R. Thiesing (Eds.), Proceedings of the
cardiovascular digital twin for human systemic circulation using 2020 Winter Simulation Conference (pp. 540–551). IEEE. https://​
inverse analysis. Biomech Model Mechanobiol 20(2):449–465. ieeex​plore.​ieee.​org/​stamp/​stamp.​jsp?​arnum​ber=​93841​12&​casa_​
https://​doi.​org/​10.​1007/​s10237-​020-​01393-6 token=g_​CYe3w​x3K8A​AAAA:​9HIoC​IeYeo​mnP1H​ptMU3​kD-​
Constantinescu C, Rus R, Rusu CA, Popescu D (2019) Digital twins oI18O2​ W8jQP ​ GmWXO ​ ICMTC ​ MhhAv​ jAeLj​ YNoX0​ nrlTO ​ zCPV​
of exoskeleton-centered workplaces: challenges and develop- 1J-_​qug&​tag=1
ment methodology. Procedia Manuf 39:58–65. https://​doi.​org/​ Kobryn PA (2020) The digital twin concept. In Frontiers of Engineer-
10.​1016/j.​promfg.​2020.​01.​228 ing: Reports on Leading-Edge Engineering from the 2019 Sym-
Corral-Acero J, Margara F, Marciniak M, Rodero C, Loncaric F, Feng posium (pp. 17–24). National Academy of Engineering. https://​
Y, Gilbert A, Fernandes JF, Bukhari HA, Wajdan A, Martinez doi.​org/​10.​17226/​25620
MV, Santos MS, Shamohammdi M, Luo H, Westphal P, Leeson P,
DiAchille P, Gurev V, Mayr M, … Lamata P (2020) The “Digital

13
Human-Intelligent Systems Integration (2022) 4:23–33 33

Laamarti F (2019) Towards standardized digital twins for health, sport, analysis and feedback of repetitive work in the manual material
and well-being. University of Ottawa. https://​doi.​org/​10.​20381/​ handling industry. In R. Bae, K.H., Feng, B., Kim, S., Lazarova-
ruor-​23746 Molnar, S., Zheng, Z., Roeder, T. and Thiesing (Ed.), Proc. of the
Laamarti F, Badawi HF, Ding Y, Arafsha F, Hafidh B, Saddik AEl 2020 Winter Simulation Conference (pp. 2637–2648)
(2020) An ISO/IEEE 11073 standardized digital twin framework Tao F, Sui F, Liu A, Qi Q, Zhang M, Song B, Guo Z, Lu SCY, Nee
for health and well-being in smart cities. IEEE Access 8:105950– AYC (2019a) Digital twin-driven product design framework. Int
105961. https://​doi.​org/​10.​1109/​ACCESS.​2020.​29998​71 J Prod Res 57(12):3935–3953. https://​doi.o​ rg/1​ 0.1​ 080/0​ 0207​543.​
Latif H, Shao G (2020) A case study of digital twins for a manufactur- 2018.​14432​29
ing process involving human interactions. In R. Bae, K.H., Feng, Tao F, Zhang H, Liu A, Nee AYC (2019b) Digital twin in industry:
B., Kim, S., Lazarova-Molnar, S., Zheng, Z., Roeder, T. and state-of-the-art. IEEE Trans Industr Inf 15(4):2405–2415. https://​
Thiesing (Ed.), Proceedings of the 2020 Winter Simulation Con- doi.​org/​10.​1109/​TII.​2018.​28731​86
ference (pp. 2659–2670). http://​repos​itorio.​unan.​edu.​ni/​2986/1/​ Tuegel EJ, Ingraffea AR, Eason TG, Spottswood SM (2011) Reen-
5624.​pdf gineering aircraft structural life prediction using a digital twin.
Li S, Yang Q, Xing J, Yuan S (2020) Preliminary study on the appli- Int J Aerospace Eng 2011. https://​doi.​org/​10.​1155/​2011/​154798
cation of digital twin in military engineering and equipment. van der Valk H, Hunker J, Rabe M (2020) Digital twins in simulative
Proceedings - 2020 Chinese Automation Congress, CAC 2020, applications: a taxonomy. In Bae R, Feng KH, Kim B, Lazarova-
7249–7255. https://​doi.​org/​10.​1109/​CAC51​589.​2020.​93269​11 Molnar S, Zheng S, Roeder Z, Thiesing T (eds) Proc. of the 2020
Liu Y, Zhang L, Yang Y, Zhou L, Ren L, Wang F, Liu R, Pang Z, Deen Winter Simulation Conference, pp 2695–2706
MJ (2019) A novel cloud-based framework for the elderly health- Wang Q, Jiao W, Wang P, Zhang Y (2021) Digital twin for human-
care services using digital twin. IEEE Access 7:49088–49101. robot interactive welding and welder behavior analysis. IEEE/
https://​doi.​org/​10.​1109/​ACCESS.​2019.​29098​28 CAA J Automatica Sinica 8(2):334–343. https://​doi.​org/​10.​1109/​
Lutze R (2020) Digital twin based software design in ehealth - a new JAS.​2020.​10035​18
development approach for health/ medical software products. Pro- Watson ME, Rusnock CF, Miller ME, Colombi JM (2017a) Inform-
ceedings - 2020 IEEE International Conference on Engineering, ing system design using human performance modeling. Syst Eng
Technology and Innovation, ICE/ITMC 2020. https://​doi.​org/​10.​ 20(2):173–187. https://​doi.​org/​10.​1002/​sys
1109/​ICE/​ITMC4​9519.​2020.​91985​46 Watson M, Rusnock CF, Colombi JM, Miller ME (2017b) Human-
Ma X, Tao F, Zhang M, Wang T, Zuo Y (2019) Digital twin enhanced centered design using system modeling language. J Cogn Eng
human-machine interaction in product lifecycle. Procedia CIRP Decis Making 11(3):252–269. https://​doi.​org/​10.​1177/​15553​
83:789–793. https://​doi.​org/​10.​1016/j.​procir.​2019.​04.​330 43417​705255
Mendi AF, Erol T, Dogan D (2021) Digital twin in the military field. Zibuschka J, Ruff C, Horch A, Roßnagel H (2020) A human digital
IEEE Internet Computing 7801. https://d​ oi.o​ rg/1​ 0.1​ 109/m
​ ic.2​ 021.​ twin as building block of open identity management for the inter-
30551​53 net of things. Lecture Notes in Informatics (LNI), Proceedings
Neisser U (2014) Cognitive psychology: classic edition. In Cognitive - Series of the Gesellschaft Fur Informatik (GI), P-305, 133–142
Psychology (Classic). Psychology Press
Parmar R, Leiponen A, Thomas LDW (2020) Building an organiza- Publisher's note Springer Nature remains neutral with regard to
tional digital twin. Bus Horiz 63(6):725–736. https://​doi.​org/​10.​ jurisdictional claims in published maps and institutional affiliations.
1016/j.​bushor.​2020.​08.​001
Sharotry A, Jimenez JA, Wierschem D, Koldenhoven RM, Valles D,
Koutitas G, Aslan S (2020) A digital twin framework for real-time

13