=Paper=
{{Paper
|id=Vol-2523/paper10
|storemode=property
|title=
Stages of Design of Digital Twin for Local Enterprise
|pdfUrl=https://ceur-ws.org/Vol-2523/paper10.pdf
|volume=Vol-2523
|authors=Regina Starodubtseva,Vlada Kugurakova,Darina Vorobyeva
|dblpUrl=https://dblp.org/rec/conf/rcdl/StarodubtsevaKV19
}}
==
Stages of Design of Digital Twin for Local Enterprise
==
https://ceur-ws.org/Vol-2523/paper10.pdf
Stages of Design of Digital Twin for Local Enterprise
Regina Starodubtseva[0000-0003-4414-368X], Vlada Kugurakova[0000-0002-1552-4910],
and Darina Vorobyeva[0000-0001-9997-2907]
Kazan (Volga region) Federal University, Kremlevskaya 18, Kazan, Russia
vlada.kugurakova@gmail.com
Abstract. The work is devoted to the consideration of the stages of development of a
digital twin for the local enterprise. With the aid of the digital twin it is possible to
manage the enterprise, control problems in it, ensure the normal functioning of the
enterprise objects due to the monitoring of a status of facilities and equipment based
on big data obtained from Industrial IoT-sensors, as well as the monitoring of perfor-
mance and current status of employees on the basis of the interpretation of bio-signals
obtained in real time. The plant for preparation of super-viscous oil with disposal
facilities was chosen as a local enterprise. This data is collected for the unit of predic-
tive analysis and prediction in real time of pre-emergency situations for the rapid
taking action to prevent these threats. In order to develop a plausible prototype of the
digital twin of the enterprise, Unity cross-platform development environment for
computer games was used.
Keywords: Digital Twin, Virtual Reality, Augmented Reality, Unity Real-Time Development
Platform
1 Introduction
Digital systems for personal use have long moved from dry data to the more natural
form of the information delivery to a user. Production systems tend to the same and, it
seems, due to technological progress in the field of automation finally step over the
abyss, after which the control systems become inherently more complex, but at the
same time easier for direct human use. The improvement of such systems will require
joint efforts of specialists of mathematical, computer, and, of course, psychological
sciences, but will provide a huge increase in the efficiency of enterprises, as well as
will contribute to the mental comfort of people working directly with the system.
Digital economy and the related terms such as digital twin, digital copy, digital
double, digital transformation, digitalization are already firmly established in the ter-
minology of both IT-specialists and economists, but there are different interpretations
of them. “A Digital Twin is an integrated multiphysics, multiscale, probabilistic simu-
lation of an as-built vehicle or system that uses the best available physical models,
sensor updates, fleet history, etc., to mirror the life of its corresponding flying twin”
[1]. According to this definition, in our understanding digital twin is an interactive 3D
representation using user-friendly visual interfaces (clear and needless for additional
Copyright © 2019 for this paper by its authors. Use permitted under Creative
Commons License Attribution 4.0 International (CC BY 4.0).
85
interpretation) of the mathematical model of the technological object, updated in real
time on the basis of information from the array of sensors. According to the vice pres-
ident of GE Software Research Colin J. Parris, “Digital twin is a living model that
drives a business outcome” [2].
Besides the remote comfortable work of top management with the current state of
the enterprise, for the work of middle managers and their subordinates, workforce in
the production itself, it is necessary to develop applications that allow to get the same
interactive data in augmented reality format. Thus, the aim of the research is the de-
velopment of the most effective and realistic simulation of the work of the local en-
terprise in real time, simulation of the sensors that monitor the condition of facilities
and equipment, as well as recording of the status of employees; the creation of func-
tionality of simple and intuitive user interface (UI/UX), data generation on the basis
of which visualization of simulation of work of the enterprise will be implemented.
The remainder of this paper is organized as follows. In Section 2, there is a sum-
mary of the literature on digital twin research. Section 3 introduces the digital twin
framework for the concrete enterprise, including the architecture and functional de-
velopment. In Section 4, a case study shows the emergence of new problems while
developing the digital twin framework. Conclusions are discussed in Section 5.
2 Related Works
The term “Twin” appeared as part of NASA's Apollo program for which two space
crafts were built, one of which remained on earth to display the condition of the first
one and was called “the twin” [3]. Usage of a digital twin creates new opportunities
for improving business processes of companies due to the accurate and timely data.
Nowadays the concept of a digital twin is considered actively as it is a strategic path-
way. According to [4] a digital twin is a powerful tool which has the following ad-
vantages: Visibility, Predictive, What if Analysis Understand and explain behaviours,
Connect disparate systems such as backend business applications. Described ad-
vantages provide great prospects for applying of a digital twin. However, the litera-
ture review reveals, that the development of the digital twin is still at its infancy as
literature mainly include concept papers, but some concrete applied case-studies al-
ready exist.
The authors of [5] consider the concept of a digital twin in a whole, its application
in product design, production scheduling, production itself and in prediction of nor-
mal and critical events. The article [6] specifically describes the application of the
digital twin for modeling of the construction work of the railway station buildings for
King’s Cross station. The results of this study can provide construction participants
with reasonable guidance on the use of the digital twin in railway station projects for
the planning, design and operation of an economical, efficient and environmentally
friendly construction project. The work [7] describes some possible cases of usage of
a digital twin for the wiring harness for Mercedes-Benz Cars. Digital twin allows to
find vehicle faults caused by damage of the wiring harness, that consequently makes
it possible to solve the problem quickly and cost-effectively. As for the petrochemical
86
industry, the research [8] represents approaches to time series data processing as well
as frequency unification, time lag issues, and the demand for immediacy for model-
ling of a digital twin.
The paper [9] concerns the use of sensors for Industrial Internet of Things and rep-
resents a SmartJacket. This solution is a jacket that carries data collecting sensors and
safety elements such as RGB LED sleeve strips. Upon the occurrence of a specific
event RGB LED sleeve stripe of the corresponding color illuminates, alerting the
employee about the emergency situation.
An essential part of a digital twin is an ability to forecast some emergency situa-
tions in an enterprise. According to [10], a digital twin propose a solution of diagnos-
tics (static and dynamic), optimization and prediction tasks due to the following com-
ponents: the initial digital model, describing the processes and relations between some
of control parameters and being optimized parameters; the knowledge base, filled
with initial information about desired KPI; the database, able to store current data
from control object; the ability to connect to DCS / PSS / ICS to send them a control
signals; the execution environment, which can run digital models; the chosen algo-
rithms of system identification and optimization. The paper [11] provides the model
of emergency situations and accidents prediction for reduce the frequency of their
occurrence at building sites. The authors analyzed data obtained from environmental
sensors and generated association rules that represent the relationship between the
accident types and causes. For this reason, preprocessing, association rule generation,
and visualization are executed step by step. An experimental implementation using
open-source R is conducted for demonstration of the accident prediction model. The
investigation [12] provides an architecture framework to implement the cyber-
physical production system, that monitors quality in metal-casting processes and pre-
dicts the occurrence of factory situations in real time based on the technologies such
as the IoT, big data, and simulations. Moreover, such system acts as a coordinator to
form optimal decisions related to the re-creation of production schedules.
In the current article we represent our own vision of a digital twin on the example
of the development of the plant for preparation of super-viscous oil with disposal
facilities.
3 Our Approaches
One of the small local enterprises in the Republic of Tatarstan was chosen as a pre-
liminary site for the implementation of digital transformation of the enterprise produc-
tion process management. The plant for preparation of super-viscous oil with disposal
facilities “Kamenka” [13] was chosen as it provides not only the relative simplicity of
a technical device, but also the ability to simulate all technological processes accord-
ing to data obtained from IoT-sensors. A comfortable realistic three-dimensional in-
terface for monitoring the situation at the site was developed for this enterprise. The
architecture of the developed system is described below.
87
3.1 Architecture of the System
Database. For the specific solution database will have four entities (tables) such as
User, Requirement, Notification, Task. More detailed database architecture is repre-
sented in the Figure (see Fig. 1).
Fig. 1. Database architecture
The table User contains information about each member of staff, personal infor-
mation, including username and password for authentication in the system. It is as-
sumed that each employee will have access to the system through the application on a
mobile device, which the employee will carry with him throughout the time spent in
the enterprise, in other words, the entire working day. He will receive notifications
(Notification) via this application. Notifications are simple messages from the admin-
istrator or other employees, or tasks (Task) that need to be executed (prevent or fix
machinery breakdown, or take the place of another co-worker). The table User also
includes such information as surname, name, date of birth, sex (male/female), job
role, work status (busy/free), email and password.
The table Requirement contains all information about equipment of the enterprise
including tank vessels, tubes, collectors, etc. It is assumed that each equipment intro-
duced into the enterprise must be registered in the system, which will store such in-
88
formation as the equipment identification plate (category), serial number, run life,
installation date and status (working properly/malfunction/exited run life). If the sta-
tus of the equipment is “malfunctioning” or “exited run life”, an administrator re-
ceives an alert and creates a task (Task) to repair or remove the equipment from pro-
duction. The task must be executed by someone of the employees.
The table Notification contains information about equipment or employee status
notifications, and it is associated with the User (User_ID) and Requirement (Re-
quirement_ID) tables. Besides ID and Foreign Key the entity of a notification have
the following attributes: priority (deviation from the norm is critical/not critical), title
(by which the deviation is caused), notification text, the date of the deviation appear-
ance and activity (the problem is fixed/not fixed).
Finally, the table Task contains information about tasks that are sent to employees
for execution, and it is associated with the User (User_ID) and Requirement (Re-
quirement_ID) tables. When an administrator receives automatic notifications about
the status of enterprise’s facilities, he creates a task object. It is assumed that employ-
ees will choose a task from the list of tasks: “accept”, or skip the incoming task, i. e.
“reject”. Thus, the entire workflow will consist of the execution of tasks that income
on the mobile devices. It will allow to optimize production activities and save time.
The Task entity will consist of attributes such as a task title, description, creation date,
and status (completed/not completed).
The database architecture should contain status of all features such as Prediction
entity including workload, integrator, cost part of the object for predicting the possi-
bility of equipment replacement. The database architecture should be also extended
with UserState entity, which has information about health and environment around
the employee. This information includes data obtained from sensors.
Database query. All database queries are implemented in the DbMySQLUtils class.
They act as methods, so inheriting from this class it is possible to access the database.
Since some queries take a good deal of time and slow down the main stream, requests
are sent to the database in separately created streams. For this reason, Sys-
tem.Threading utility is used.
All data from the database is read after a specified period of time to remove exces-
sive load from the database. This algorithm is described in the ReadAllInfoThread()
method (see Fig. 2). When starting the application, the system will immediately read
the data, then update the data again after a specified pause, and so on. If other queries
to the database (INSERT, UPDATE, DELETE) are executed, the stream, that reads
all data, will stop, then the stream with a certain query will be started and then the
ReadAllInfoThread() method with the newly created stream will be started again.
89
Fig. 2. Method that reads all data from the database
Movement of employees. The movement of employees will be carried out accord-
ing to their tasks: the employee, having received an alert (a call from a head, automat-
ic notification of the problem, scheduled tasks, etc.), must go to the location of the
specified division.
For debugging, the procedure of the current location of the employee was simulat-
ed, which was implemented through the serialization of data into a JSON-file of sev-
eral checkpoints. In the future, the movement will be implemented not via coordinates
from the JSON-file, but relative to deviations from the norm of equipment and staff
conditions, emergency situations, and other factors.
Indicators of equipment sensors. Serialization and deserialization of indicators of
sensors of the equipment were implemented by the same principle. For tubes, tanks,
collectors two indicators were selected – pressure and temperature. The condition of
the equipment will be determined on the basis of these indicators. Each indicator has
three zones: green, yellow and red. Each zone is an interval in which one or another
indicator can be located. The green zone means that the indicator is normal, the yel-
90
low zone – a slightly deviation from the norm, the red one – a critical deviation from
the norm.
Indicators of employee sensors. Similarly to serialization and deserialization of
equipment sensors data of employees are recorded and read. First of all, it is pressure,
pulse and temperature. Employee pressure is measured in two values: upper indicator
(arterial tension) and lower indicator (venous blood pressure). Both indicators are
measured in millimeter of mercury (mm Hg). The temperature is measured in stand-
ard degrees Celsius (°C). Pulse is measured in beats per minute.
However, for comprehensive workplace safety in the oil industry with the aid of
sensors also it is possible to measure the electrocardiogram, record the respiratory rate
and heart rate, body position and activity of the staff, as well as to supplement the
system with sensors of environmental quality, for example, sensors that monitor the
presence of poisonous gas in the working environment, that provides monitoring of
the health of the employee. Moreover, with the aid of the alarm installed in the wear-
able device, the employee can be timely warned about dangerous situations or be able
to warn about them by pressing the button.
3.2 Functional Development
Zoom and view from different angles. An important part of the implementation of
digital twins are natural interfaces that give mental comfort to the user working direct-
ly with the system [14].
The zoom implementation for the desktop version was programmed as standard via
a computer mouse Scroll Wheel. Viewing from different angles is implemented via
User Interface (UI) buttons. When being clicked by a computer mouse wheel on a
particular employee or equipment the point of click becomes closer.
View employee data and equipment information. It was necessary to add the
ability to view employee data and equipment information with the aid of using entity-
relationship model from the database (Requirement and User). The fields in which
this information will be displayed are panels Requirement InfoPanel and Employee
InfoPanel (see Fig. 3), located on the right side of the screen. These panels will ap-
pear as often as a user clicks the left mouse button on the corresponding digital object.
Fig. 3. Current employee data
91
Notifications about problems. When a problem occurs in the system (sensor indi-
cators are in the red or yellow zone), a notification object (Notification) appears. De-
pending on the degree of criticality of each indicator, a certain algorithm determines
in which zone the overall condition of the equipment or employee is located. Moreo-
ver, depending on the indicators of previous notifications, the decision is made
whether to deactivate (disable) the latest notifications. Deactivated notifications are
notifications included in the archive. They can be viewed in the employee or equip-
ment card by clicking on the button with the warning icon (triangle with an exclama-
tion mark). Notifications related to an equipment or employee can be sorted by the
following categories: All messages, Accepted, Rejected, and Recent. When is being
hovered by the mouse over the warning button on the main scene, a drop-down list
with the same categories appears (see Fig. 4).
Label is an instant visualization of the status of an employee or equipment. In
order to instantly determine the status of an employee or equipment, the label tool is
developed. The label appears on the top of the object, and can be yellow if the devia-
tion is not critical, red – in case of critical deviation (see Fig. 5).
Fig. 4. Drop-down list of notifications
Fig. 5. Employee with a warning label
92
4 Emergence of New Problems
The problem of overload of connection of the application with the database occurs
when there are too many requests sent to the server. In order to avoid this problem it
was necessary to develop an algorithm by which all queries will be distributed in
accordance with the priority given to them. Thus, a queue of queries is created, where
the higher priority ones are executed at the beginning. If the amount of data in the
database is too large, it is necessary to distribute it to several different servers. The
next step in the implementation of the pilot application will be solving this load
problem to determine the load limits for the physical architecture organization.
5 Conclusion
To sum up, digital twin is in the process of development and requires the introduction
of many functions for the correct operation of the system. At this stage, first of all, it
is necessary to solve the problem of overload of the communication channel with the
database. In addition, it is necessary to add a display function and a corresponding
sort of a notification on the stage-screen with notifications and on Requirement
InfoPanel and Employee InfoPanel panels.
However, besides the current problems, many important tasks were solved:
designing the application architecture, functionality for editing and displaying
information about equipment and employees, creation of notifications and functional
content of the scene such as scaling, camera movement, displaying UI/UX elements
when pressing or hovering over certain buttons, as well as automatic display of
warning labels in places of problems.
Acknowledgements
This work was funded by the subsidy of the Russian Federation Government to
support the Program of competitive growth of Kazan Federal University among
world-class academic centers and universities.
93
References
1. Glaessgen, Edward, and David Stargel: The digital twin paradigm for future NASA and
US Air Force vehicles, 53rd AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dy-
namics and Materials Conference 20th AIAA/ASME/AHS Adaptive Structures Confer-
ence 14th AIAA (2012).
2. Parris, C.J., Laflen, J.B., Grabb, M. L., and Kalitan, D.M.: The future for industrial ser-
vices: the digital twin. Infosys Insights, 42–49 (2016).
3. Rosen, R., von Wichert G., Lo, G., and Bettenhausen, K.D.: About the importance of au-
tonomy and digital twins for the future of manufacturing, IFAC-Papers OnLine 48 (3),
567–572 (2015).
4. Kuehn, W.: Digital twins for decision making in complex production and logistic enter-
prises. Int. J. of Design & Nature and Ecodynamics 13 (3), 260–271 (2018).
5. Qi, Q. and Tao, F.: Digital twin and Big Data towards smart manufacturing and industry
4.0: 360 degree comparison, in IEEE Access 6, 3585–3593 (2018).
6. Kaewunruen, S. and Xu, N.: Digital twin for sustainability evaluation of railway station
buildings, Frontiers in Built Environment 4, art. no. 77 (2018).
7. Tharma, R., Winter, R., and Eigner, M.: An approach for the implementation of the digital
twin in the automotive wiring harness field, Proceedings of International Design Confer-
ence, DESIGN 6, 3023–3032 (2018).
8. Min, Q., Lu, Y., Liu, Z., Su, C., and Wang, Bo: machine learning based digital twin
framework for production optimization in petrochemical industry, International Journal of
Information Management (2019).
9. Marcon, P., Arm, J., Benesl, T., Zezulka, F., Diedrich, C., Schröder, T., Belyaev, A.,
Dohnal, P., Kriz, T., and Bradac, Z.: New approaches to implementing the SmartJacket in-
to industry 4.0, Sensors (Switzerland) 19 (7), No. 1592 (2019).
10. Kostenko, D., Kudryashov, N., Maystrishin, M., Onufriev, V., Potekhin, V., and Vasi-
liev, A.: Digital twin applications: Diagnostics, optimisation and prediction, Annals of
DAAAM and Proceedings of the International DAAAM Symposium 29 (1), 0574–0581
(2018).
11. Kwon, J.-H. and Kim, E.-J.: Accident prediction model using environmental sensors for
industrial internet of things, Sens. Mater. 31 (2), 579–586 (2019).
12. Lee, J.H., Do Noh, S., Kim, H.-J., and Kang, Y.-S.: Implementation of cyber-physical pro-
duction systems for quality prediction and operation control in metal casting, Sensors
(Switzerland) 18 (5), No 1428 (2018).
13. Khayrullin, L.O., Kugurakova, V.V., and Starodubtseva, R.A.: Digital real time twin of en-
terprise entity. CEUR Workshop, in print (2019).
14. Kugurakova, V.V., Elizarov, A.M., Khafizov, M.R., Lushnikov, A.Yu., and
Nizamutdinov, A.R.: Towards the immersive VR: measuring and assessing realism of user
experience. In 23nd International Conference on Artificial ALife and Robotics, 146–152
(2018).
94