Professional Documents
Culture Documents
Download full ebook of Advances In Design Simulation And Manufacturing Vi Proceedings Of The 6Th International Conference On Design Simulation Manufacturing The Lecture Notes In Mechanical Engineering 1St Edition Vitalii Iv online pdf all chapter docx
Download full ebook of Advances In Design Simulation And Manufacturing Vi Proceedings Of The 6Th International Conference On Design Simulation Manufacturing The Lecture Notes In Mechanical Engineering 1St Edition Vitalii Iv online pdf all chapter docx
Download full ebook of Advances In Design Simulation And Manufacturing Vi Proceedings Of The 6Th International Conference On Design Simulation Manufacturing The Lecture Notes In Mechanical Engineering 1St Edition Vitalii Iv online pdf all chapter docx
https://ebookmeta.com/product/recent-advances-in-intelligent-
manufacturing-select-proceedings-of-icame-2022-lecture-notes-in-
mechanical-engineering-harish-kumar-editor/
https://ebookmeta.com/product/advances-in-systems-engineering-
select-proceedings-of-nsc-2019-lecture-notes-in-mechanical-
engineering-v-h-saran-editor/
https://ebookmeta.com/product/sustainable-manufacturing-and-
design-woodhead-publishing-reviews-mechanical-engineering-
series-1st-edition-kaushik-kumar-editor/
https://ebookmeta.com/product/13th-chaotic-modeling-and-
simulation-international-conference-springer-proceedings-in-
complexity-christos-h-skiadas-editor/
https://ebookmeta.com/product/14th-chaotic-modeling-and-
simulation-international-conference-springer-proceedings-in-
complexity-christos-h-skiadas-editor/
https://ebookmeta.com/product/proceedings-of-the-27th-
international-conference-on-systems-engineering-
icseng-2020-182-lecture-notes-in-networks-and-systems-182-henry-
Lecture Notes in Mechanical Engineering
Advances in Design,
Simulation and
Manufacturing VI
Proceedings of the 6th International
Conference on Design, Simulation,
Manufacturing: The Innovation
Exchange, DSMIE-2023,
June 6–9, 2023, High Tatras, Slovak
Republic - Volume 1: Manufacturing
Engineering
Lecture Notes in Mechanical Engineering
Series Editors
Fakher Chaari, National School of Engineers, University of Sfax, Sfax, Tunisia
Francesco Gherardini , Dipartimento di Ingegneria “Enzo Ferrari”, Università di
Modena e Reggio Emilia, Modena, Italy
Vitalii Ivanov, Department of Manufacturing Engineering, Machines and Tools,
Sumy State University, Sumy, Ukraine
Mohamed Haddar, National School of Engineers of Sfax (ENIS), Sfax, Tunisia
• Engineering Design
• Machinery and Machine Elements
• Mechanical Structures and Stress Analysis
• Automotive Engineering
• Engine Technology
• Aerospace Technology and Astronautics
• Nanotechnology and Microengineering
• Control, Robotics, Mechatronics
• MEMS
• Theoretical and Applied Mechanics
• Dynamical Systems, Control
• Fluid Mechanics
• Engineering Thermodynamics, Heat and Mass Transfer
• Manufacturing
• Precision Engineering, Instrumentation, Measurement
• Materials Engineering
• Tribology and Surface Technology
Advances in Design,
Simulation and
Manufacturing VI
Proceedings of the 6th International Conference
on Design, Simulation, Manufacturing:
The Innovation Exchange, DSMIE-2023,
June 6–9, 2023, High Tatras, Slovak Republic -
Volume 1: Manufacturing Engineering
Editors
Vitalii Ivanov Justyna Trojanowska
Sumy State University Poznań University of Technology
Sumy, Ukraine Poznań, Poland
Ján Pitel’
Technical University of Košice
Prešov, Slovakia
© The Editor(s) (if applicable) and The Author(s), under exclusive license
to Springer Nature Switzerland AG 2023
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors
or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
The book “Advances in Design, Simulation and Manufacturing VI” was organized
into two volumes according to the main conference topics: Volume 1—Manufacturing
Engineering and Volume 2—Mechanical and Materials Engineering. Each volume is
devoted to research in design, simulation, and manufacturing in the main conference
areas.
The first volume consists of four parts. The first part includes recent developments
in smart manufacturing. It presents intelligent solutions for sensor selection in plant
retrofitting. This part also includes improving machine vision systems for collabora-
tive assembly applications. Recent developments in digital twins for industrial robotics,
robotic arms in a virtual environment, and designing a workplace in virtual and mixed
realities are presented in this part. The evolution of competence management in man-
ufacturing industries and digital business models is also discussed. Finally, the first
part includes studies on recent developments in digital and information technologies
according to the Industry 4.0 strategy.
The second part includes studies in information management systems, particularly
ensuring quality management at manufacturing enterprises, improving information man-
agement systems, and perspectives of lean management. Studies in increasing the pro-
duction process efficiency in new production facilities, implementing intelligent trans-
port systems, and improving logistics analysis are additionally presented. Moreover, this
part also analyzes the impact of standardized reusable packaging on supply chain design
and environmental efficiency.
The third part of manufacturing technology is devoted to modeling and surface modi-
fication of steel by laser radiation, fabrication and post-processing of metal products, and
verifying additively manufactured and bulk metals. It presents studies in implementing
computer-aided joint design systems, applying information technologies to ensure qual-
ity characteristics of the working surfaces during mechanical processing, and improving
the quality of cutting tools based on convolutional models. This part also includes recent
developments in contact processes during drilling, microcutting of titanium alloys and
steel during grinding, and substantiation of chip removal models during milling. More-
over, issues regarding the stress-deformed state of the cylinder liner’s working surface,
turning threads for drill pipes, surface relief formation in peripheral end milling, and
just-in-time gear grinding wheel dressing are presented. Studies in material removal
for aluminum alloys and surface roughness assessment during the milling of pure and
carbon black-reinforced polypropylene materials are discussed in this part. Finally, the
third part proposes ways to improve drives for a multi-operational milling machine,
automatic workpiece clamping mechanisms of lathes to expand technological capabil-
ities, and an impact of solid lubrication on the diamond grinding characteristics for
difficult-to-machine materials.
The fourth part regarding engineering education is based on studies in a dual edu-
cational system for the professional training of workers and the entrepreneurial mindset
development of engineering students. A digital twin for remote learning is also pro-
posed in this part. Finally, this part proposes a professional and creative competencies
formation model for engineers based on information and digital technologies.
We appreciate the partnership with Springer Nature, iThenticate, EasyChair, and our
sponsors for their essential support during the preparation of DSMIE-2023.
Preface vii
Thank you very much to DSMIE Team. Their involvement and hard work were
crucial to the success of the conference.
DSMIE’s motto is “Together we can do more for science, technology, engineering,
and education”.
Steering Committee
General Chair
Co-chair
Program Committee
DSMIE Team
Smart Manufacturing
Tracking of Trucks Using the GPS System for the Purpose of Logistics
Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Peter Trebuňa, Marek Mizerák, Miriam Pekarčíková, Marek Kliment,
and Matúš Matiscsák
Manufacturing Technology
Engineering Education
Dmytro Adamenko(B)
Abstract. The chemical, process, and energy generation industries are essential
production industries worldwide. However, these traditionally powerful industrial
sectors have been under ever-increasing cost pressure in recent years, so the com-
panies are continuously trying to improve wage- and energy-intensive products
as well as services through the use of digitalization of production processes, e.g.,
to use operational data to optimize processes by tailoring them to customer needs
and, at the same time, cutting operation and production costs. The machines and
equipment in these industries are typically operated for several decades and can-
not be directly integrated into the newly developed digitized processes. One of the
possibilities to perform a so-called smartification of the machines and equipment
is to retrofit them with sensors and data processing hardware. The present paper
presents a methodology for the systematic selection of sensors for retrofitting
existing and operating plants. For this purpose, the requirements for the selec-
tion process have been established. The advantages and disadvantages of existing
methods for sensor selection are presented and analyzed. Subsequently, a new
methodology for sensor selection is presented and successfully validated in the
retrofit process of an existing and still operating for 20 years 900 kW wind turbine.
1 Introduction
Asset-intensive industries such as chemicals, oil and gas, mining, metals, pulp and paper,
and power production have been turning to new technologies to increase the reliability
and availability of their equipment while keeping maintenance costs under control [1].
Using digital tools and advanced analytics capabilities alongside traditional lean tech-
niques, they aim to predict and prevent equipment failures, increase labor productivity,
and streamline the management of external contractors [2].
Recent studies show that maintenance expenses account for a large portion of the
vulnerable operating costs of manufacturing companies and can be as high as 25% of
operating costs [3]. So, 82% of the companies have faced unplanned machine break-
downs in recent years. For 40% of plant operators, one hour of downtime cost more than
e1 million [4]. Thus, maintenance is a large part of the cost-intensive activities that can
be optimized, and maintenance costs are the only type of cost that can be influenced
after a plant has been put into operation [5].
Maintenance is often planned at specific intervals based on manufacturer recom-
mendation or operator experience. Industry 4.0 approaches and any process monitoring
generally require a meaningful process variable that allows conclusions to be drawn
about the current status of the production process or a machine [6]. Condition-based
maintenance ensures full utilization of the service lifecycle of systems and components
and saves additional maintenance work due to more frequent service downtimes. The
prerequisite for the realization of the condition-based maintenance approach is the avail-
ability of actual process data, which must be measured and transferred for evaluation
with the help of sensors and supporting hardware to the virtual models [7] that can
describe the behavior and simulate the remaining useful life of the system so that predic-
tive maintenance can be realized [8]. The transition from preventive to condition-based
maintenance strategies becomes necessary to meet the described requirements [9].
In factories, assets are kept in operation as long as they continue producing high-
quality products [10]. It is not unusual to see machines working for several decades still
in use on the shop floor [11]. Older machinery is usually in good condition but do not
have the necessary infrastructure to collect the measurement data and derive from it the
condition of the components. Investments in older equipment allow users to maximize
functionality with new sensors and software features. Due to the extremely high cost
of new machines, retrofit is being considered. Investments in modernizing equipment
result [12] in a higher utilization [13] of existing assets [14] have also been considered.
Sensor selection is often based on machine operator experience and sensor manufac-
turer expertise. However, it is not always clear which measurements must be measured
and how they must be performed. Since machines and equipment with a variety of com-
ponents can be exposed to different mechanical, electrical and other loads, selecting
the correct sensor becomes a great challenge, as the resistance and compatibility of the
measurement principle of sensors with the environmental conditions must be ensured.
Another challenge [15] is to find a method suitable for multiple machines [16] within a
single shopfloor [17].
2 Literature Review
Scientific works and industrial practice have established that the digitization of industry
and the transformation of production processes are necessary to secure the competitive-
ness of companies [18]. The long service lives and significant investments in machinery
and equipment have led to the formation of a market for their upgrade through additional
or replacement components. This market is constantly growing, and its turnover in 2021
is estimated at 45 billion dollars [19].
One of the main challenges of the retrofit process is the strong heterogeneity of
machines within a single shopfloor, so each machine requires its retrofit methodology.
Some of the global industrial companies, such as ABB [20], Festo [21], or TROX [22],
offer specialized retrofit kits that adapt the older systems of these manufacturers and
provide the corresponding functional extensions.
Still, no holistic method for retrofitting existing machines and systems can be trans-
ferred to several machines presented. From the practical examples in the literature and
Sensor Selection for Smart Retrofitting of Existing Plants 5
presented retrofit concepts [13, 23], a generally applicable scheme for the process of a
retrofit was extracted. The general retrofit process roughly consists of three steps [24].
First, the decision to implement the retrofit is to be made based on an analysis of oper-
ating, investment, and operating costs with predictive maintenance [25]. Subsequently,
the selection of the components that are relevant for securing the plant operation and
characteristic measured variables that can indicate a possible failure of defined compo-
nents takes place [26]. At the end of the process, the actual state of the plant or machine
is acquired, and sensors, hardware, and suitable mounts for attaching selected additional
components are selected [27].
Selection of sensors for the producing machines and plants is challenging because
many factors and requirements must be considered. A series of conflicting objectives
arise when instrumenting a plant with a wide variety of transducers. Sensor systems
with high sensitivity are desired so that a change in a physical process variable results in
the most significant change in the sensor’s output variable. Often, this requires a certain
proximity to the active point. The closer measurements are made to active components,
the less favorable the environmental conditions for sensors often become. Depending
on the distance from the active point, sensors are stressed by high temperatures, vibra-
tions, or the effects of process fluids [6]. The selection is complicated by a multitude of
sensors on the market with different operating principles, measurement characteristics,
and specific necessary environmental conditions to ensure the reliable measurement and
transmission of the signal.
For the realization of an optimal sensor selection, the requirements for the selection
process were established. These were derived from the available sensor selection methods
of Löpelt [28], VDMA [29], Hesse [30], Meß [31], and Classe [32], but also from the
context of the entire retrofit process. The most important of these has been summarized
in Table 1. Subsequently, the requirements were mapped to the considered methods.
The sensor selection approaches presented in the literature differ to some extent
in the intended use for which they were designed and cover a wide range of sensor
selection tasks. So, the approaches of Löpelt and VDMA focus on the retrofit processes.
The methods of Hesse, Meß, and Classe focus on sensor selection for new designs.
The known approaches require very application-dependent, specific information that
is dependent on the product requirements and, apart from abstract recommendations for
activities such as those offered in the well-known approaches to retrofitting and sensor
selection, can hardly be presented in a generally valid way in their diversity. The situation
is similar to reformulating the sensor task into a measurand to be recorded, for which a
high level of metrological experience may be necessary.
Table 1. (continued)
3 Research Methodology
The present paper aims to derive the methodology for sensor selection while considering
the conditions of the existing machinery and production processes. To achieve this, the
requirements for the sensor selection process are first established, the known methods
[28–32] for sensor selection are analyzed concerning the requirements, and an optimized
procedure for sensor selection is derived. The better suitable process is to be judged based
on more requirements being fulfilled compared to the known methods. Variables to be
Sensor Selection for Smart Retrofitting of Existing Plants 7
measured and the operating and plant conditions are assumed to be known from previous
steps of the retrofit process.
For the realization of the sensor selection, a 5-step methodology was proposed. The
central aspect of the developed methodology should be the interaction between sensor
selection and coordination with the functional structure or the requirements and, thus,
the integration of loops. Therefore, following the argumentation in the previous chapter,
the arrangement of the steps is chosen as far as possible. So, the list of requirements for
the sensor can be completed meaningfully. Using the integration of a permanent result
display, and the thus guaranteed traceability of the exclusion steps in the configurator
operation, this can be facilitated by the computer-aided procedure. An adjustment should
be made if no result remains according to the display before the requirements list is
processed. To facilitate operation, user access to selection steps that have already been
run through should it be possible to make specific corrections, after which the results are
updated without restarting the configurator. The selection process is illustrated in Fig. 1.
At the beginning of the process, the value to be measured, the installation space, and
the ambient condition should be known. The sequence starts with the selection of the
measurand. For this purpose, the user is presented with a list of common, industry-typical
physical values in which the one required for the measurement task can be searched.
This is important because, after each step, an exclusion of sensors takes place, in this
case those that are not suitable for measuring the measurand. If required, the database
can be extended to include other measured variables and other characteristics. In the
same step, the user enters the required measuring range by selecting a minimum and
maximum value in the appropriate unit. The configurator compares this input with the
entered measuring ranges of the sensor database to reject sensors that do not fit. Thus, all
sensors with minimum values that are above the entered minimum value and maximum
values that are below the entered maximum value will not be considered further.
In the requirements for the measurement process, the signal properties required to
fulfill the measurement task are queried on a measurement-quantity-specific basis. Most
cases should include linearity, resolution, accuracy, or hysteresis. However, it is not
advisable to generalize this selection for all measurement quantities since the technical
8 D. Adamenko
requirement criteria also change due to differences in the measurement strategy. The
sensor’s bandwidth determines whether the measurement task can be fulfilled concern-
ing technical requirements. Instead of the bandwidth, however, response times are an
essential feature for temperature sensors. In this step, the input is made via the required
value in the respective unit. In the same step, the conditions at the measurement location
are described. The user is offered to describe the possible influences on the measurement
location. In order to select a sensor according to the available space and the electrical con-
nection requirements, mounting conditions are also requested in this step. This includes
the size of the sensor, the type of connection, the type of mounting, and the supply
requirement. This information can usually be obtained from the data sheets provided
by the manufacturers. At this step of the developed methodology, the material proper-
ties (magnetic, conductive, and insulating) and environmental conditions (temperature,
pressure, visibility and light conditions) are also recorded and queried.
In the next step, the results are filtered for environmental disturbance variables.
According to VDMA, relevant environmental variables that can influence the quality
of the measurement are temperature, vibration and shock loads, ambient pressure, light
radiation, contact with media, and electromagnetic fields. Classe, Hesse, and Hering
supplement this list with dust and dirt particles, explosive environments, and radioac-
tive radiation. Furthermore, the behavior with respect to electromagnetic interference
and the effect of external interference fields on the radiation of a sensor are classified.
The information on vibration and shock resistance refers to the ability of the sensor to
withstand these loads mechanically without suffering damage.
Within the framework of the developed methodology, all disturbance variables that usu-
ally lead to impaired measurement results on the sensor product ready for use have been
recorded. This means that disturbances inherent in the active principle but compensated
for standards in the sensor design are not included in the list. These are less interesting
for sensor selection than for sensor manufacturing. Some of the listed items can also
be assigned to the measurand-dependent characteristic conditions, such as the division
of the ambient temperature into the corresponding ranges of low (lower than −30 °C),
high (above 50 °C), and normal temperatures. The same categorization has also been
performed for ambient pressure.
After filtering the sensors resistant to the disturbance variables, the information in
the last step concerns the preference of the output signal or desired communication
interfaces. As a result of the safety analysis of the system, ideally performed by the user,
information about the effect severity of a sensor system failure is available. In order to
support the redundant arrangement of the sensor system that may be deemed necessary
by this analysis result, the configurator solution can offer the advantage of suggesting
several sensors, even of different operating principles, after the selection process has
been run through. This can be the case if several sensors are available after using the
configurator.
Suppose no sensor is available as a result of the selection according to the strict
exclusion principle. In that case, it is possible to carry out a utility value and weak point
analysis of the sensors that are not one hundred percent suitable using user access to
all steps in retrospect to weigh up the extent to which non-optimal circumstances can
be accepted. Naturally, knock-out criteria should be defined for this purpose, which
immediately excludes sensors in case of non-compliance.
In order to design how cataloging could look like considering the mentioned distur-
bance variables, advantages, and influencing factors of manufacturer sensors, a scheme
according to Roth [33] is used in the following. Accordingly, a division of the information
into main, outline, and access parts takes place to ensure clarity and quick information
access. A difference to the definition of a construction catalog after Roth is that the
approximate completeness is fulfilled only conditionally with the cataloging of man-
ufacturer sensors. Although the compilation aims to be prepared for many different
applications, complete coverage of all products on the market is not intended in this
framework. The one-dimensional structure is intended to serve as a uniform structure.
First and foremost, the outline section contains the measured variable to which the
following contents are related. As central, immediately to be considered further outline
criteria apply against the background of the installation of the sensor into an existing
product, the present or constructively still to be realized kind of the event change. The
latter is assigned to the material of the measuring object and the permissible switching
10 D. Adamenko
distance. The permissible materials are partially one-sided and not mutually exclusive.
Like the sensing distances, whose structured arrangement would be arbitrary due to the
different specifications of either a fixed value or a range, they do not assume a clear
structuring function. Nevertheless, they occur in the outline part because of the close
connection to the event change and their characteristic as a measurand-specific condition.
Thus, a different outline part would be required for each measurand.
The main part consists of the sensor model and the associated operating principle.
The access part contains the other properties queried in the configurator mode, via the
selection of which the most favorable solution for prevailing environmental conditions
and interfaces can be found. The access part is subdivided into requirements for the
equipment to fulfill the measuring task. In this case, only the frequency range, disturbance
variables, environmental influences, installation conditions, which include the electrical
connection and the supply voltage, and the type of output signal.
Validation of the developed methodology has been realized with the retrofit of
an operating wind turbine. The variables to be measured were defined as rotational
speed, wind speed, wind acceleration, and elongation of the tower. The environmental
conditions were recorded and evaluated with a questionnaire.
The validation showed that suitable sensors could be found with the developed
methodology because corresponding models are available in the library. In the use cases
for the four defined measurands, only sensors were identified that could meet the techni-
cal requirements for measurement or were similar or even identical to the design of the
sensors already in use. The sensors issued in these four cases and not considered suit-
able for the task were too oversized regarding functional properties and, therefore, not
worth the potential cost and installation effort to the system operator. In the future, this
problem could be counteracted by an additional display of the cost boundary conditions
during the sensor output by the configurator for a convenient overview or the query of
the intended service life.
The selection of the search criteria thus seems to provide a reliable function in
most of the present examples, apart from the few points mentioned. It was left to the
user to collect the necessary data to work through the individual steps to find suitable
manufacturer products without additional search effort and information requirements.
In some cases, the decision already made by the responsible team of the company was
confirmed, or, especially in the case of acceleration and wind speed measurement, the
field of vision was extended to equally possible alternatives. No reason was given as to
why the sensor types that were not used had not been applied so far. The compilation of
the questions needed for the configurator arose from the orientation on the steps of the
developed method, which suggests that even without using a configurator, the procedure
can lead to the successful finding of a suitable sensor.
5 Conclusions
The great importance of sensors emerges from the strategies discussed for the approach
to retrofit measures and the requirements for systems to be retrofitted. Examining the
strategies showed that integrating a detailed methodology for sensor selection in this
context is still outstanding. Such a methodology, intended to support the planner of a
Sensor Selection for Smart Retrofitting of Existing Plants 11
retrofit, was developed and presented. It is based on the conditions and structures of a
plant and, with the help of task-related requirement queries, limits the possible sensor
selection according to the exclusion method.
In particular, the intensive treatment of disturbance variables dependent on the active
principle represents an essential supplement to existing approaches in the developed
method. The detailed treatment of measurand-specific trade-offs in sensor selection,
which is not addressed in every approach or is touched upon superficially, was a crucial
filtering process during the validation.
The validation results demonstrate that the developed method can be suitable for
gathering information about the plant structure and requirements needed for a sensor
selection for industrial applications, given a previously defined metric. By integrating
sensors into existing plants, it is possible to gather essential knowledge about the process
and advance the digitization of operations. On the way to a configurator that can cover
a wide range of possible use cases, collecting comprehensive information about various
measurement situations beyond the examples covered in the validation is still pending.
Acknowledgment. This work was supported by the German Federal Ministry for Economic
Affairs and Energy within the Central Innovation Program for SMEs.
References
1. Maciel, R., Safarik, P.: Power Plant 4.0: Embracing Next-Generation Technologies for Power
Plant Digitization. McKinsey & Company (2020)
2. VDMA – Fit for Service (2017). https://www.vdma.org/viewer/-/v2article/render/1117285.
last accessed 2021/09/21
3. ConMoto: Value-oriented maintenance: Strategic dimension of the wrench. ConMoto
Consulting Group (2011)
4. Vanson Bourne: After the Fall: The Costs, Causes & Consequences of Unplanned Downtime.
GE Digital (2017)
5. Leidinger, B.: Value-Oriented Maintenance: Reduce Costs, Maintain Availability, 2nd edn.
Gabler Verlag, Wiesbaden (2017)
6. Bonitz, J., Hamm, L., Müller, P., Hermann, S.: Improved process supervision by highly
sensitive strength receivers CNT. ZWF 115(7–8), 488–491 (2020). https://doi.org/10.3139/
104.112373. (in German)
7. Blum, A., al Diban, R., Rudolph, M., Weickhardt, C.: Optical multi-sensor network for main-
tenance: implementation in a machine tool application. ZWF 115(5), 299–302 (2020). https://
doi.org/10.3139/104.112283. (in German)
8. Zhai, S., Reinhart, G.: Predictive maintenance as a pioneer for maintenance-oriented
production control. ZWF 113(5), 298–301 (2018)
9. Paessler. IT Explained: Retrofitting. https://www.paessler.com/it-explained/retrofitting. last
accessed 2021/09/21
10. BearingPoint: Predictive Maintenance Study 2021. BearingPoint, Frankfurt/Main (2021)
11. Schuh, G., Anderl, R., Gausemeier, J., ten Hompel, M., Wahlster, W.: Industrie 4.0 Maturity
Index. Managing the Digital Transformation of Companies (Acatech STUDY). Herbert Utz
Verlag, Munich (2017)
12. Zwick/Roell: Service Reliability and Upgradeability for Older Testing Machines.
Zwick/Roell, Ulm (2016)
12 D. Adamenko
1 Introduction
Collaborative robots and their implementation in production are an essential part of the
Industry 4.0 concept [1]. Collaborative robots offer increased flexibility and mobility
than classical industrial robots [2] and can also simplify or increase the productivity
of applications where classic industrial robots were previously used. According to IFR
statistics [3], as of 2021, 39,000 collaborative robots were installed. Compared to 2020,
sales increased one and a half times; compared to 2017, more than three and a half times.
At the same time, the percentage of collaborative robots compared to the total number
of industrial robots sold in 2017 is 2.75%, and this ratio for 2021 is 7.54%.
Despite the increased popularity of collaborative robots in production, the relative
share of installations of collaborative robots is currently low. The authors noted that one
of the reasons for the slow introduction of collaborative robots is the need to ensure a high
level of worker safety [4]. In addition to physical safety, the mental safety of the worker
must also be ensured to reduce worker anxiety and stress during the human-machine
interaction by making the interaction more accessible and intuitive [5]. Moreover, it is
necessary to simplify the work with modern technologies due to the lack of qualified
workers. According to the survey, both enterprises that have implemented the Industry
4.0 technologies and enterprises that are just planning to modernize production feel a
lack of qualified workers [6], negatively affecting the speed of spreading the Industry 4.0
concept. Due to the presence of open questions in collaborative robotics, this area is still
open for new research, methodology development, and definition of basic requirements.
The authors chose machine vision systems as a field of study. Machine vision sys-
tems have become one of the fundamental robotics technologies at the adaptation stage
of information collection technologies (Robotics 2.0) and environment perception tech-
nologies (Robotics 3.0). The use of sensors (including cameras) and the correct analysis
of their readings increase the adaptability and autonomy of industrial robots. Moreover,
without information from sensors, it is impossible to realize the concept of Industry 4.0
[7] fully.
This work is devoted to studying machine vision systems that can solve safety prob-
lems and simplify the work with collaborative robots, endowing them with basic cogni-
tive functions, like parts recognition. The theoretical part of the article will be devoted
to a brief overview of machine vision systems, their role in modern collaborative cells,
and practical examples of their application will be considered. As a practical part, a case
study of using a machine vision system in a collaborative cell to increase its flexibility
will be considered.
2 Literature Review
Machine vision systems in collaborative applications perform many tasks, such as
object detection, sorting, robot control, visual inspection, barcode identification, con-
trol of employee movement, supervision identifying defects, tracking, reporting, etc.
[8]. Speaking of using machine vision systems to increase the flexibility of production
cells, the authors [9] proposed a deep reinforcement learning algorithm for grasping
using machine vision. The perspective of studies in this field is that such algorithms
will allow capturing various objects with high efficiency and minimal human involve-
ment, including objects that did not participate in the training of the neural network.
Moreover, the scientific developments in data exchange between robots will allow the
exchange of experience in grasping various objects. In the future, work in this area can
significantly increase robotic systems’ flexibility and performance [7]. The authors [10]
proposed and verified a complex for grasping objects in a cluttered environment, which
consists of a novel gripper with controlled rigidity, a depth camera, and a computer that
can identify a suitable position for capturing correctly. Moreover, big companies are
Machine Vision Systems for Collaborative Assembly Applications 15
for the robot’s end effector. Also, the system can consider potential changes in the
workspace [23], such as the movement of other people and machines, if there is data for
training the neural network. As a future work, the authors consider the algorithm fitting
for the work of several people with a robot and a person with several robots.
As a result, scientific research in machine vision systems aims to improve existing
image processing algorithms, develop new ones to optimize task performance or expand
the technological capabilities of machine vision systems.
3 Research Methodology
There are 1D, 2D, and 3D machine vision systems available on the market. 1D machine
vision systems analyze a linear region instead of an area analysis since these systems
have one row of photodiodes. Based on their properties, they are most often used to detect
objects and control them continuously (e.g., while moving on a conveyor). Information
processing algorithms (which provide cropping, de-distorting, and joining images in the
correct order) make it possible to form a 2D image, and the image width can be unlimited.
In addition, providing sufficient and more uniform light will be necessary due to the need
to reduce the exposure time when shooting in dynamic conditions, which is much easier
when using 1D machine vision systems [25]. Also, 1D machine vision systems are used
for high-resolution visual inspection. These systems can scan moving objects at high
frequency, while the weight resolution of 1D machine vision systems can be from 512 to
Machine Vision Systems for Collaborative Assembly Applications 17
12888 pixels [26]. Among the industrial applications of these systems is the continuous
control and analysis of sheet materials on the conveyor, such as paper, plastic [27],
polymers [28, 29], and steel sheets.
2D machine vision systems are the most common because of their universality. They
have a photosensitive matrix consisting of many rows of photodiodes. Therefore, they
are optimal for working with static objects and without needing to work with height
information. Also, they are the easiest to use. Among the limitations are sensitivity
to light intensity and poor contrast, which in turn is solved by installing additional
filters or using image processing algorithms, as well as the presence of parallax. Typical
examples of using 2D machine vision systems are determining objects’ coordinates for
further capture by an industrial robot, quality and size control, reading characters from
packages, monitoring, ensuring worker safety, etc. [30].
Matrices of 2D machine vision systems can be designed to produce monochrome
and color images. The main difference between monochrome and color matrices is that a
monochrome matrix captures the entire incident light beam on each photodiode, regard-
less of the length of the wave. In contrast, in color matrices, light is captured by groups
of pixels with additional color filters that transmit only a specific color (green, blue,
or red). Then the missing color components are calculated by interpolation algorithms
based on data from surrounding photodiodes [31]. Based on this, monochrome matrices
work faster, have a more significant resolution and greater detail level, and work better
in poor lighting conditions [32]. As a result, monochrome cameras remain the preferred
choice, while color cameras are recommended when information about the color of an
object is needed.
In addition, matrices of 2D machine vision systems can be with global or rolling
shutters. In a rolling shutter matrix, individual photodiodes or groups of photodiodes
begin and end the absorption of a light beam. In contrast, in the global shutter matrix, all
photodiodes begin to receive light simultaneously but end sequentially. Rolling shutter
sensors are cheaper but do not work well in dynamic environments [33, 34], distorting
moving objects’ images. As a result, global shutter systems are recommended, especially
in dynamic scenes [35].
3D machine vision systems provide a 3D digital model of the area or object being
scanned and can optionally overlay a 2D color image for a complete object representation.
Also, this type of camera usually is less sensitive to environmental conditions. But due to
the complexity of the design and the high cost of components, this is the most expensive
type of machine vision system.
Typical practical applications of 3D machine vision systems include object size
control with height inclusive, object scanning and digitization (reverse engineering),
defect detection, and safety control of collaborative cells.
The importance of choosing the proper resolution of the machine vision system
should be noted. A study [36] during the practical test of the proposed safety solution
revealed a relationship between the quality of the application execution and the camera
resolution. But cameras that offer higher resolution come at a higher cost. As a result,
paying attention to the application responsibility, the difference between the minimum
and maximum controlled element that needs to be controlled simultaneously, and the
budget is recommended. Based on this, choose the machine vision system. Also, as
18 V. Andrusyshyn et al.
mentioned above, the maximum resolution of a camera sensor depends on its type. So,
in some cases, it will be necessary to change the machine vision system with another
type of camera or use several machine vision systems to control different parameters.
An example of choosing machine vision systems for the SmartTechLab laboratory will
be considered as a practical verification of the obtained methodology. SmartTechLab
laboratory is located at the Faculty of Manufacturing Technology with a seat in Prešov
of the Technical University of Košice, Slovak Republic. A more detailed description of
the laboratory is available in previous work [39].
The laboratory is dedicated to teaching students the key technologies of the Industry
4.0 concept, research, and development work. The key equipment of the laboratory is the
collaborative assembly line (Fig. 1). In this example, it is planned to configure a line for
the assembly of switches (Fig. 2a). Since the switch consists of many parts (Fig. 2b), there
are problems with tool accessibility, small part sizes, and complex assembly steps (for
example, installing a spring). Collaborative assembly is preferable to full automation.
The assembly of the switcher consists of the following stages: incoming inspection,
assembly, and final inspection. During the incoming inspection, it is necessary to ensure
that the base part of the switcher is correctly installed on the holder and that there are no
external defects. The holder enters the field of view of the camera on the conveyor at low
speed. In addition, it is allowed to stop the conveyor to perform the inspection operation
since the bottleneck of the system is the assembly operation. Based on this, the system
can be considered static. In addition, when choosing equipment for the laboratory, the
focus should be more on flexibility and versatility than on performance. The fact that the
base part is monotonous and dimpled makes the verification process more complicated.
The 3D machine vision system is the optimal choice for these production conditions.
Because the system is static, 1D machine vision systems are not the optimal choice. In
addition, for a 3D machine vision system, the monotony of a part will not be an issue,
unlike 2D machine vision systems. In addition, information about the height of the part
will help determine the correct installation of the part on the holder. A machine vision
system with a structured light sensor was chosen, as it can create a high-resolution model,
works well with small parts, is relatively cheap, and works well with monotonous parts.
Based on the analysis of available solutions, it was decided to choose the Keyence CA-
H048CX machine vision system (Fig. 3a). A 3D model comparison tool and a profile
measurement tool were used to check the geometric parameters and the correct height
of the profile (Fig. 3b).
During collaborative assembly, the robot must be able to determine the exact coordi-
nates of parts and pre-assemblies since a person can accidentally or deliberately change
their position. Because there are small parts that are involved in the assembly, it can be
critical to capture them. Since the system is static, and there is no need to determine the
colors of the workpieces, a monochrome 2D machine vision system is the optimal choice.
Since ABB YuMi is used as the collaborative robot, ABB offers a force-controlled col-
laborative gripper with an integrated Cognex AE3 machine vision system. Compared to
an installation in a fixed place, the machine vision system built into the gripper has more
flexibility since it is possible to change the inspection location freely, and there are no
issues with dead zones. The process of controlling the position of the base part is shown
in Fig. 4a. The circle finder is used as a tool for the controller (Fig. 4b).
Fig. 4. Machine vision system of the collaborative robot: a – general view; b –application.
After the assembly, it is necessary to perform a visual inspection of the final product.
All remarks about the statics of the system during the incoming inspection are applicable
to the final inspection control. However, unlike the input control, the appearance of the
assembly will be checked, so there is no need for measurements in the three-dimensional
plane. Moreover, since the assembly elements are only black and white, there will be
high contrast for even monochrome cameras. Based on the description of the inspection
conditions, it is optimal to use a monochrome 2D machine vision system. Keyence CA-
H048MX was chosen for the final inspection (Fig. 5a). The process of carrying out the
final inspection of the assembly is shown in Fig. 5b, which compares the original pattern
with the current one.
22 V. Andrusyshyn et al.
Regarding research in this direction, the authors [8] presented an overview of the
application of machine vision systems in smart factories, including considerations of
collaborative applications. Still, they did not consider the choice of hardware. The authors
[40] provided information about the features of the choice of agricultural machine vision
systems. As a result, this study has a scientific novelty and describes the features of the
choice of machine vision systems for smart factories with collaborative production cells.
5 Conclusions
This article examined the main areas of application of machine vision systems for col-
laborative assembly cells and their place in the Industry 4.0 concept. The experience
of using machine vision systems in collaborative cells was also systematized, and the
market of machine vision systems was analyzed. The scientific novelty of this work lies
in the fact that recommendations on the optimal choice of machine vision systems for
collaborative areas were received. These recommendations were tested in a case study
when choosing machine vision systems for the robotic collaborative assembly cell of the
SmartTechLab laboratory (Faculty of Manufacturing Technologies with a seat in Presov,
Technical University in Kosice). The practical value of this study lies in the fact that the
received recommendations will allow researchers and engineers to simplify and speed
up the decision-making process on the optimal choice of machine vision systems for lab-
oratories and production lines, which may include collaborative robots. The limitations
of this article include focusing only on the main elements of machine vision systems
(image sensors and vision processing devices), as well as on the main parameters of the
elements (for example, parameters such as aperture, image storage method, etc. were
not considered). In future work, the authors consider a more detailed analysis of the
application of machine vision systems in security and an analysis of other elements of
machine vision systems, such as lighting and lenses.
Machine Vision Systems for Collaborative Assembly Applications 23
Acknowledgment. This work was supported by the Slovak Research and Development
Agency under contract No. APVV-19-0590, and also by the projects VEGA 1/0700/20
and KEGA 022TUKE-4/2023 granted by the Ministry of Education, Science, Research and
Sport of the Slovak Republic, and The NAWA Ulam Programme under the grant number
BPN/ULM/2022/1/00045. The results were partially obtained within the Joint Ukrainian-Slovak
R&D Project for the period of 2022–2023 granted by the Slovak Research and Development
Agency (SK-UA-21-0060) and Ministry of Education and Science of Ukraine (State Reg. No.
0122U002657). The scientific results were partially obtained within the research project “Ful-
fillment of tasks of the perspective plan of development of a scientific direction “Technical sci-
ences” Sumy State University” ordered by the Ministry of Education and Science of Ukraine
(State Reg. No. 0121U112684). This research was partially supported by the International
Association for Technological Development and Innovations.
References
1. Dillinger, F., Bernhard, O., Kagerer, M., Reinhart, G.: Industry 4.0 implementation sequence
for manufacturing companies. Prod. Eng. 16, 705–718 (2022). https://doi.org/10.1007/s11
740-022-01110-5
2. Sherwani, F., Asad, M.M., Ibrahim, B.S.K.K.: Collaborative robots and industrial revolution
4.0 (IR 4.0). In: 2020 International Conference on Emerging Trends in Smart Technologies
(ICETST), pp. 1–5 (2020). https://doi.org/10.1109/ICETST49965.2020.9080724
3. Bill, M., Müller, C., Kraus, W., Bieller, S.: World Robotics 2022 Report. Frankfurt, Germany
(2022)
4. Saenz, J., Elkmann, N., Gibaru, O., Neto, P.: Survey of methods for design of collaborative
robotics applications- why safety is a barrier to more widespread robotics uptake. In: ICMRE
2018: Proceedings of the 2018 4th International Conference on Mechatronics and Robotics
Engineering, pp. 95−101. ACM (2018). https://doi.org/10.1145/3191477.3191507
5. Villani, V., Pini, F., Leali, F., Secchi, C., Fantuzzi, C.: Survey on human-robot interaction
for robot programming in industrial applications. IFAC-PapersOnLine 51(11), 66–71 (2018).
https://doi.org/10.1016/j.ifacol.2018.08.236
6. Nimawat, D., Gidwani, B.D.: Challenges facing by manufacturing industries towards imple-
mentation of industry 4.0: an empirical research. Int. J. Interact. Des. Manuf. (IJIDeM) 16(4),
1371–1383 (2022). https://doi.org/10.1007/s12008-022-00961-7
7. Gao, Z., Wanyama, T., Singh, I., Gadhrri, A., Schmidt, R.: From industry 4.0 to robotics
4.0 - a conceptual framework for collaborative and intelligent robotic systems. Procedia
Manufacturing 46, 591–599 (2020). https://doi.org/10.1016/j.promfg.2020.03.085
8. Javaid, M., Haleem, A., Singh, R.P., Rab, S., Suman, R.: Exploring impact and features of
machine vision for progressive industry 4.0 culture. Sensors Int. 3, 100132 (2022). https://
doi.org/10.1016/j.sintl.2021.100132
9. Quillen, D., Jang, E., Nachum, O., Finn, C., Ibarz, J., Levine, S.: Deep reinforcement learning
for vision-based robotic grasping: a simulated comparative evaluation of off-policy methods.
In: 2018 IEEE International Conference on Robotics and Automation (ICRA), pp. 6284–6291.
IEEE (2018). https://doi.org/10.1109/ICRA.2018.8461039
10. D’Avella, S., Tripicchio, P., Avizzano, C.A.: A study on picking objects in cluttered envi-
ronments: exploiting depth features for a custom low-cost universal jamming gripper. Robot.
Comput.-Integr. Manuf. 63, 101888 (2020). https://doi.org/10.1016/j.rcim.2019.101888
11. Eppner, C., et al.: Four aspects of building robotic systems: lessons from the amazon picking
challenge 2015. Auton. Robot. 42(7), 1459–1475 (2018). https://doi.org/10.1007/s10514-
018-9761-2
24 V. Andrusyshyn et al.
12. Lin, H.-Y., Liang, S.-C., Chen, Y.-K.: Robotic grasping with multi-view image acquisition
and model-based pose estimation. IEEE Sens. J. 21(10), 11870–11878 (2021). https://doi.
org/10.1109/JSEN.2020.3030791
13. Yu, Y., Cao, Z., Liang, S., Geng, W., Yu, J.: A novel vision-based grasping method under
occlusion for manipulating robotic system. IEEE Sens. J. 20(18), 10996–11006 (2020). https://
doi.org/10.1109/JSEN.2020.2995395
14. Paxton, C., Hundt, A., Jonathan, F., Guerin, K., Hager, G.D.: CoSTAR: instructing collabora-
tive robots with behavior trees and vision. In: 2017 IEEE International Conference on Robotics
and Automation (ICRA), pp. 564–571. IEEE (2017). https://doi.org/10.1109/ICRA.2017.798
9070
15. Liang, X., Zhu, H., Chen, Y., Yamakawa, Y.: Tracking and catching of an in-flight ring using
a high-speed vision system and a robot arm. In: IECON 2021 – 47th Annual Conference of
the IEEE Industrial Electronics Society, pp. 1–7. IEEE (2021). https://doi.org/10.1109/IEC
ON48115.2021.9589565
16. Scime, L., Beuth, J.: Using machine learning to identify in-situ melt pool signatures indicative
of flaw formation in a laser powder bed fusion additive manufacturing process. Addit. Manuf.
25, 151–165 (2019). https://doi.org/10.1016/j.addma.2018.11.010
17. Cha, Y.-J., You, K., Choi, W.: Vision-based detection of loosened bolts using the Hough
transform and support vector machines. Autom. Constr. 71, 181–188 (2016). https://doi.org/
10.1016/j.autcon.2016.06.008
18. Wang, J., Fu, P., Gao, R.X.: Machine vision intelligence for product defect inspection based
on deep learning and Hough transform. J. Manuf. Syst. 51, 52–60 (2019). https://doi.org/10.
1016/j.jmsy.2019.03.002
19. Cherubini, A., Navarro-Alarcon, D.: Sensor-based control for collaborative robots: funda-
mentals, challenges, and opportunities. Front. Neurorobot. 14, 113 (2021). https://doi.org/10.
3389/fnbot.2020.576846
20. Mazhar, O., Navarro, B., Ramdani, S., Passama, R., Cherubini, A.: A real-time human-
robot interaction framework with robust background invariant hand gesture detection. Robot.
Comput.-Integr. Manuf. 60, 34–48 (2019). https://doi.org/10.1016/j.rcim.2019.05.008
21. Ferraguti, F., et al.: Safety barrier functions and multi-camera tracking for human–robot shared
environment. Robot. Auton. Syst. 124, 103388 (2020). https://doi.org/10.1016/j.robot.2019.
103388
22. Makris, S., Aivaliotis, P.: AI-based vision system for collision detection in HRC applications.
Procedia CIRP 106, 156–161 (2022). https://doi.org/10.1016/j.procir.2022.02.171
23. Besedin, M., Popowska, M., Ivanov, V., Trojanowska, J.: Digital model and assembling of a
lathe. J. Eng. Sci. 9(1), A1–A8 (2022). https://doi.org/10.21272/jes.2022.9(1).a1
24. Components of Machine Vision | Cognex. https://www.cognex.com/what-is/machine-vision/
components. last accessed 2022/10/03
25. Boucher-Genesse, A.: 1D, 2D, 3D... What Kind of Vision System Should I Use for My
Application? https://blog.robotiq.com/1d-2d-3d...-what-kind-of-vision-system-should-i-use-
for-my-application. last accessed 2022/10/08
26. Line scan cameras - Quality right down the line. https://www.stemmer-imaging.com/en-gb/
technical-tips/line-scan-cameras/. last accessed 2022/10/08
27. Krol, O., Sokolov, V.: Research of toothed belt transmission with arched teeth. Diagnostyka
21(4), 15–22 (2020). https://doi.org/10.29354/diag/127193
28. Berladir, K., Gusak, O., Demianenko, M., Zajac, J., Ruban, A.: Functional properties of
PTFE-composites produced by mechanical activation. In: Ivanov, V., et al. (eds.) DSMIE 2019.
LNME, pp. 391–401. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-22365-6_39
29. Budnik, A.F., Rudenko, P.V., Berladir, KV., Budnik, O.A.: Structured nanoobjects of
polytetrafluoroethylene composites. J. Nano- Electron. Phys. 7(2), 02022 (2015)
Machine Vision Systems for Collaborative Assembly Applications 25
30. NT, B.: Machine vision systems: 1D, 2D, and 3D, https://roboticsbiz.com/machine-vision-
systems-1d-2d-and-3d/. last accessed 2022/11/08
31. Monochrome VS color sensors - Opto Engineering Basics. https://www.opto-e.com/en/bas
ics/monochrome-vs-color-sensors. last accessed 2022/10/09
32. A short guide to why monochrome cameras have the edge over color cameras –
e-con Systems. https://www.e-consystems.com/blog/camera/technology/a-short-guide-to-
why-monochrome-cameras-have-the-edge-over-color-cameras/. last accessed 2022/11/09
33. Pavlenko, I., Ivanov, V., Gusak, O., Liaposhchenko, O., Sklabinskyi, V.: Parameter identi-
fication of technological equipment for ensuring the reliability of the vibration separation
process. In: Knapcikova, L., Balog, M., Perakovic, D., Perisa, M. (eds.) 4th EAI International
Conference on Management of Manufacturing Systems. EICC, pp. 261–272. Springer, Cham
(2020). https://doi.org/10.1007/978-3-030-34272-2_24
34. Monkova, K., et al.: Condition monitoring of Kaplan turbine bearings using vibro-diagnostics.
Int. J. Mech. Eng. Robot. Res. 9(8), 1182–1188 (2020). https://doi.org/10.18178/ijmerr.9.8.
1182-1188
35. Rolling and global shutter (CMOS sensors). https://www.stemmer-imaging.com/en/knowle
dge-base/rolling-shutter/. last accessed 2022/10/09
36. Ozkahraman, M., Yilmaz, C., Livatyali, H.: Design and validation of a camera-based safety
system for fenceless robotic work cells. Appl. Sci. 11(24), 11679 (2021). https://doi.org/10.
3390/app112411679
37. Introduction to Vision Sensors - White Papers | Cognex. https://www.cognex.com/resources/
white-papers-articles/introduction-to-vision-sensors. last accessed 2022/10/07
38. Sun, T., Cao, J.: Research on machine vision system design based on deep learning neural
network. Wirel. Commun. Mob. Comput. 2022, 16 (2022). https://doi.org/10.1155/2022/480
8652
39. Demčák, J., Lishchenko, N., Pavlenko, I., Pitel’, J., Židek, K.: The experimental SMART
manufacturing system in SmartTechLab. In: Trojanowska, J., Kujawińska, A., Machado, J.,
Pavlenko, I. (eds.) MANUFACTURING 2022. LNME, pp. 228–238. Springer, Cham (2022).
https://doi.org/10.1007/978-3-030-99310-8_18
40. Syed, T.N., Lakhiar, I.A., Chandio, F.A.: Machine vision technology in agriculture: a review
on the automatic seedling transplanters. Int. J. Multi. Res. Dev. 6(12), 79–88 (2019)
Digital Twins for Industrial Robotics:
A Comparative Study
1 Introduction
The development of Industry 4.0 ideas supports the growth of Digital Twin technol-
ogy, particularly in the manufacturing sector. It can be characterized as a two-way data
integration between a physical and a virtual machine [1]. Intelligent manufacturing
solutions use information from production processes to enhance individual and overall
industrial production. Most of the time, the most expensive part of production is the
human workforce.
A Digital Twin is a virtual model of a physical object. It models the object’s whole
lifecycle and uses real-time data sent from sensors on the object to simulate the behavior
and monitor operations [2]. This method combines traditional mechanical engineering,
mechatronics, computer science, and artificial intelligence. Engineers can evaluate the
functionality of industrial concepts without creating a prototype due to the advanced sim-
ulation of the machine. Early correction of mistakes greatly minimizes the costs needed
Fig. 1. Influence on Design and Cost in the Product Design and Process Planning [4].
Digital Twins and PLM software can also predict the product’s lifecycle before
manufacturing. Siemens is one of the producers of PLM software and offers a whole
PLM solution. However, smaller companies typically cannot afford the entire portfolio
and are limited to only essential SW (software) [5].
This paper compares SW, which could be used for simulations of the main part of the
manufacturing process. The study compares the simulation of educational and research
models of assembly lines that can serve as a benchmark because they are in the fuzzy
area between design and commissioning.
2 Literature Review
During the design of mechatronic devices, it is necessary to consider the increasing
complexity of the manufacturing machines. Since a mechatronic device comprises
mechanical, electrical, and software parts, the number of potential failures is increased
[6].
One of the most widely used development models is the “V-Model” (Fig. 2). It shows
a product’s lifecycle from the requirements analysis to the final commissioning. Figure 2
shows the full version of the V-Model for designing mechatronic devices. For purely
mechanical devices, the V-Model is simplified [7].
28 D. Fait and V. Mašek
The left part of the diagram shows the problem specification, analyses, and design
phase. The bottom part shows prototype testing and problem-solving (DTs are most
beneficial in this part). The right part shows integration tests, Hardware in the Loop
(HIL) testing, debugging, and preparation for commissioning [7].
One of the main advantages is avoiding errors in the control code. Testing a code
on a physical mechatronic system is complicated because every mistake can lead to
a crash. This is significant, especially for manufacturing machines, because they must
move various components simultaneously, making crashes sometimes hard to predict
[10]. However, the DT is not just that. Another essential part is physics simulation and
line layout optimization by a real-time two-way flow of the data between the physical
and digital models of the studied system. Correction of errors [11] during the operational
phase of the lifetime [12] is also costly [1, 4, 13].
3 Research Methodology
The educational assembly lines were chosen as practical examples because all the fea-
tures of the programs could be easily demonstrated. The machine was designed in sec-
tions. Simulation SW (Process Simulate or Mechatronic Concept Designer) that allows
users to define 3D models and all their movements, sensors, physical attributes, and
input/output signals was used for the mechatronics design [4].
This paper compares DT creation of the assembly lines for education and research
purposes, described in the following sections. Digital Twins are especially suitable for
Digital Twins for Industrial Robotics: A Comparative Study 29
education not only to understand the concept and the methodology of making DTs but
also for co-simulation purposes in Totally Integrated Automation (TIA) Portal program-
ming. So each student has access to the device. Even though it is just a digital model, it
suits the purpose well enough. Then we establish a connection with the TIA Portal, where
DT can be used as a surrogate model of the real Hardware to create robotic manipulators.
The work aims to make a functional DT and compare the possibilities of different
software not just from a functionality point of view but also with the stress on educational
purposes, user-friendliness, and the possibility of simulating real-life machine parts,
similar to the solution presented in [13]. Another study goal is to discuss areas in the
V-Model covered by combining the MCD and TPS with TIA Portal. It is also necessary
to define which areas are covered by mentioned SW and deserve increased attention.
A further aim is also to compare both programs and provide recommendations for the
further development of the system engineering process.
An assembly line MecLab (Fig. 3) prepared by the company FESTO was used for
NX Mechatronic Concept Designer. It consists of three separate stations that can be
combined and/or used separately. The first one consists of a pneumatic gripper that can
manipulate the part from one position to the second. The second part is the conveyor
with a separating actuator that can push the part from the conveyor to the slide. The
third part is the press that connects two manipulated parts. MecLab’s goal is to teach
students automation technology through realistic, practical examples that are also simple
to understand. Students can grasp new concepts or enhance their knowledge by compar-
ing the digital simulation and the actual machine and understand the fundamentals of
electrical and control design [14].
A manipulation line shown in Fig. 4 was used for the Tecnomatix Process Simulate. It
was designed at the Department of Machine Design for education in automation and fluid
30 D. Fait and V. Mašek
mechanism control using PLC. The device contains all the basic automation elements: a
conveyor, linear and angular pneumatic drive, both bistatic and monostatic valves, two
infrared position sensors and end position sensors, and two gravity conveyors. When
the part is at the end of the conveyor, it is pushed by the pneumatic drive and slides
to the position gripped by a magnetic actuator attached to the linear drive. Then it is
manipulated above the second slide and slides back on the conveyor.
Defining the physical and mechanical properties of the various machine parts is the
first step in creating a Digital Twin of a machine. It models the dynamic behavior
and interactions of several components. TIA Portal enables a control program to be
programmed. DT can be combined with the PLC SimAdvanced for the code function
assessment. Finally, it can be connected to a PLC and used in a program control [16].
In this part, the main areas of the SW comparison are covered. They can be divided into
the following categories:
Fig. 12
Fig. 14
Fig. 15
Fig. 16
ORDINARY TABLE
(i) Billiard spot 12¾ in. from face (or nearest point) of the top
cushion.
(ii) Radius of semicircle of D, 11½ in.
(iii) Baulk-line, 29 in. from face of bottom cushion.
CHAMPIONSHIP TABLE
(i) 12½ in. (ii) First 9½ in. then 10 in. (iii) 28 in.
It is essential to the true running qualities, as well as to the lasting
qualities of the table, that the frames (which few people ever see)
should be exceedingly strong, thoroughly well seasoned, and
accurately levelled. The slightest warping of these frames is fatal to
the preservation of the table, and they are made, for this country at
any rate, of the most carefully selected red deal. If the table is
intended for the tropics, mahogany or teak should be used. As soon
as the frames are bolted to the legs (which, by the way, are erected on
an absolutely level base), they are carefully trimmed over with a long
plane, and, until the straight-edge fairly meets the frames all over, in
whatever direction it may be tried, the bed is not ready for the slate.
Fig. 17
References to Diagram
B, B, B, B. Bed of table.
Cushion 1. Top cushion.
Cushion 2. Right top side cushion.
Cushion 3. Right bottom side cushion.
Cushion 4. Bottom cushion.
Cushion 5. Left bottom side cushion.
Cushion 6. Left top side cushion.
D. The D.
L, L. Baulk-line.
P1. Left top pocket.
P2. Right top pocket.
P3. Right middle pocket.
P4. Right bottom pocket.
P5. Left bottom pocket.
P6. Left middle pocket.
p, p, p, ..., Pocket plates.
S1. The spot, or the billiard spot.
S2. Pyramid spot.
S3. Centre spot.
4
S . Left spot of the D.
5
S . Centre spot of the D.
6
S . Right spot of the D.
Holes are bored in the sides of the slates and metal dowels leaded
into one side, as shown in the sketch (fig. 17), so that each slate may
fit into the next, and then large holes are drilled out on the underside
of the slates and steel nuts leaded in to take the long screws which
fasten the cushions firmly to the slates (fig. 18).
Fig. 18
Let the iron be too cold rather than too hot. It is easy to go over the
table a second time, and after a little experience the proper
temperature will become known. If the cloth be once scorched it will
never be satisfactory. Therefore:
Rule 1.—Never put an iron on the cloth until you have practically
satisfied yourself that it is not too hot.
Rule 2.—Never iron an unbrushed cloth.
UNDERSIZED TABLES
Tables can be bought 10 ft. by 5 ft., 9 ft. by 4½ ft., and so on down
to 6 ft. by 3 ft. There are also combination tables which serve as
dining-tables as well, and they seem to answer very well in small
houses.
For the earnest student a so-called ‘spot stroke’ table 6 ft. by 3 ft.,
which is really a section of the top of a full-sized table, made after the
fashion of the sketch (fig. 24), will afford an immense amount of
amusement and practice; for, in addition to the ‘spot,’ the top of the
table game, ‘rail’ cannons, &c., can be practised. The pattern of the
spot stroke table is a matter of fancy and cost; the side A, B, C need
have neither cushion nor pocket.
Fig. 24
If the small tables do not satisfy the ambition of the player, he has
yet another course open—viz. to get a French table, and seriously
study the cannon game in all its phases; and signs are not wanting
that this beautiful game, as it is becoming better known, is finding
favour with advanced players.
It may not be generally known by the vast army of people who rent
furnished houses for a few months at a time that billiard-tables can
be hired from the leading makers by the week or the month, the rate
being from, say, 2l. 2s. to 3l. 3s. a month, with a small payment in
addition to cover erecting and dismantling. Lighting and all fittings
are, of course, included in the above charges, and thus, without the
expense of purchase, a billiard-table can be brought within anyone’s
reach.
Before leaving the subject of tables, an ingenious modification of
an old arrangement whereby the balls are automatically returned
from the pockets to the baulk end of the table may be mentioned. It
has been patented by Messrs. Orme & Sons, and will be found a
convenience where no marker is kept, specially in practising losing
hazards from baulk. For spot, or top of the table play, it is
inconvenient; but the drawback of having to return to baulk for the
ball after each hazard may be remedied by placing a plug or false
bottom into each top pocket (failing anything better, a pocket
handkerchief will do), so that the ball may be removed in the usual
way. Mr. Rimington-Wilson has devised a mode of meeting this
difficulty, and Messrs. Orme have acquired the right to use his
patent.