Cooperation and Synchronization of Robotic Tasks Using a Digital Twin
Alexandre Parant
1 a
, Laurent Arcese
2 b
, Sinuh
´
e Martinez-Martinez
1 c
and Arthur Marguery
1,2
1
CESI LINEACT, UR 7527, Reims, France
2
Universit
´
e de Reims Champagne-Ardenne, CRESTIC, UR 3804, Reims, France
Keywords:
Robotics, Digital Twin, Cyber-Physical Production Systems, Automation, Flexibility.
Abstract:
Industry 4.0 marks a significant advancement in the manufacturing process by integrating advanced digital
technologies. Robotics is one of the nine pillars defining the contours of Industry 4.0. These robots must be
able to perform tasks safely, especially when working simultaneously in shared areas. However, robots only
have a partial view of the production environment and need to communicate with each other to obtain more
extensive information. To facilitate the exchange of information and ensure safety during the process, we can
use a digital twin that contains information on the layout of the production system and is tasked with converting
and transmitting part position information from one robot to the other. The communication between the robots
is realized thanks to the OPC UA communication protocol. The effectiveness of this strategy is illustrated on
a robotic platform constituted by two 6-axis Niryo Ned robots associated with their digital twin.
1 INTRODUCTION
Industry 4.0 represents a significant leap forward
across the entire manufacturing chain (Wanasinghe
et al., 2020), (Bajic et al., 2021). This new era is
characterized by the integration of advanced digital
technologies, to control cyber-physical systems for
example, to enhance operational efficiency but also
to redefine the capabilities of industrial automation.
Cooperative robots are components of Industry 4.0,
which work alongside human operators to enhance
productivity and efficiency. Robots are designed to
be highly adaptable allowing them to perform various
tasks ranging from simple repetitive actions to com-
plex, precision-oriented processes.
In the framework of Industry 4.0, robots are in-
tegrated into a smart factory environment where they
can communicate with a network of interconnected
machines and systems. This connectivity allows for
real-time data exchange and decision-making, leading
to more efficient production processes. For instance,
robots can automatically adjust their operations based
on their sensor’s data, ensuring optimal performance
and minimizing errors.
Moreover, the use of robots in manufacturing can
address the growing demand for customized prod-
a
https://orcid.org/0000-0001-8433-8671
b
https://orcid.org/0000-0003-1797-4672
c
https://orcid.org/0000-0001-9161-6393
ucts. Indeed, by rapidly reprogramming robots, man-
ufacturers can switch between different product lines
without significant downtime. Resource management
and optimization of robot arm trajectories to mini-
mize energy consumption contribute to more environ-
mentally friendly production processes (Barenji et al.,
2021), (Mohammed et al., 2014). In addition, robots
can perform repetitive or hazardous tasks for human
workers, improving workplace safety and reducing
the risk of injury.
Cooperation and synchronization among robots
hold significant potential across various industries.
By working together, robots can complete tasks more
efficiently and accurately than a single unit. This col-
laborative approach allows for task division, where
each robot can specialize in specific tasks, lead-
ing to enhanced productivity and reduced comple-
tion times. Additionally, by sharing real-time data,
robots can adapt to changing conditions and correct
errors more quickly by defining new tasks according
to this new situation. This robotics cooperation leads
to faster, more accurate results and increased produc-
tivity across various applications.
However, with several robots operating in a shared
work area, the various production tasks must be syn-
chronized to avoid collisions. This constraint com-
plicates the design of the controller and extends the
development time, which is problematic in Industry
4.0, where production changes are frequent and con-
Parant, A., Arcese, L., Martinez-Martinez, S. and Marguery, A.
Cooperation and Synchronization of Robotic Tasks Using a Digital Twin.
DOI: 10.5220/0012999900003822
Paper published under CC license (CC BY-NC-ND 4.0)
In Proceedings of the 21st International Conference on Informatics in Control, Automation and Robotics (ICINCO 2024) - Volume 2, pages 73-79
ISBN: 978-989-758-717-7; ISSN: 2184-2809
Proceedings Copyright © 2024 by SCITEPRESS – Science and Technology Publications, Lda.
73

trol configuration changes must be made as quickly as
possible to limit production downtime.
In this context, the use of a digital twin for co-
ordinating tasks among multiple robots offers many
advantages. A digital twin is a virtual replica of a
cyber-physical system that allows real-time monitor-
ing and simulation (Barricelli et al., 2019). By em-
ploying digital twins, robots can be coordinated more
effectively, as the virtual model provides a compre-
hensive overview of the entire operation. This enables
precise task allocation, ensuring each robot performs
optimally. Real-time data from the robots is fed back
into the digital twin, allowing for dynamic adjust-
ments and continuous improvement. This enhances
synchronization among robots, reducing downtime
and improving overall productivity. Furthermore, dig-
ital twins can facilitate better resource management
by optimizing robot trajectories and minimizing en-
ergy consumption. Integrating digital twins in robotic
coordination and synchronization leads to more ef-
ficient, accurate, and adaptive manufacturing pro-
cesses.
In this work, a digital twin approach is used to im-
prove synchronisation and information exchange be-
tween two robots using the same work area. A pallet
containing two types of parts arrives at the robots via
a conveyor belt to be disassembled. Each robot is re-
sponsible for one type of part and only one robot at a
time can operate in the area. In addition, there is only
one camera on one of the robots, so the other robot
must receive information about the parts’ position to
carry out its task.
To facilitate the exchange of information and en-
sure safety during the process, we use a digital twin
that contains information on the production system
layout and is tasked with converting and transmitting
part position information from one robot to the other.
The digital twin provides an overall view of the pro-
duction area, unlike robots, which only have a partial
view based on information from their sensors. The
contribution of this paper is to show how a digital twin
can be used to facilitate the cooperation and synchro-
nisation of robots thanks to its overview and the use
of the OPC UA communication protocol.
This paper is organized as follows. Section 2 pro-
vides a brief overview of the use of a digital twin in
robotics applications. Section 3 describes the robotic
platform, the digital twin and its relevance to this ap-
plication, as well as the communication protocol be-
tween the digital twin and the platform. Section 4
is devoted to experiment results, illustrating the strat-
egy’s effectiveness. Conclusions and future research
are given in the last section.
2 DIGITAL TWIN AND
ROBOTICS
The development of digital twins for robotics is one
of the areas that has seen growing interest from re-
searchers and manufacturers, particularly for mod-
elling and simulation, data management, and bring-
ing intelligence and connectivity to production sys-
tems to support the transition to Industry 4.0 (Liang
et al., 2022).
The authors in (Wu et al., 2022) propose the use
of a digital twin to compensate for the positioning er-
rors of a robotic arm. A position sensor is added to
the robotic arm to communicate to the digital twin the
actual position via Websocket. By comparing the in-
formation from the robot and the sensor, the digital
twin can adjust the position if it detects a significant
deviation.
A digital twin integrating resource information,
sensor information and layout information is pro-
posed in (Kousi et al., 2019) to adapt the behavior of
mobile robots in real time. The authors have defined
a data model to reconstruct the 3D environment and
enable real-time trajectory adaptation. However, the
components exchange information via the Robot Op-
erating System (ROS) interface, making it more com-
plex to integrate non-robotic components into the dig-
ital twin.
Digitilization and simulating a robot in a software
environment makes it possible to use artificial intel-
ligence (AI) techniques to train or optimize before or
during task execution. Reinforcement learning is used
in (Matulis and Harvey, 2021) to train a robotic arm
with six degrees of freedom for a pick-and-place ap-
plication. In (Bansal et al., 2019), the authors have
developed an ant colony algorithm to avoid collisions
in an assembly task performed by an industrial robot.
A genetic algorithm is used in (Liu et al., 2022) to
optimize the trajectory of a mobile robot using real
data from the physical robot. Several partial models
are integrated into a common model in (Erkoyuncu
et al., 2018) to provide information for a learning
model. The model embedded in the digital twin en-
ables the exploration of various scenarios and real-
time decision-making.
A digital twin of a 6-axis robotic arm has been
developed in previous work to perform virtual com-
missioning and improve robotics learning (Sow et al.,
2023). This paper extends the functionality by includ-
ing the digital twin in real-time production for syn-
chronisation and information transmission. The new
architecture of the digital twin is described in the next
section.
ICINCO 2024 - 21st International Conference on Informatics in Control, Automation and Robotics
74

3 METHODOLOGY
In this section, we briefly present the robotic platform,
the digital twin and the network protocol, as well as
the Environment Layer required for robotic tasks.
3.1 Robotic Platform
The robotic platform consists of two 6-axis Niryo
Ned robots (Niryo, 2024) and a conveyor belt (see
Figure 1). Each robot has a suction cup at the end
of its arm, enabling it to pick up an object. A cam-
era is placed on just one robot to observe its environ-
ment. Using image processing, it is then possible to
detect objects and determine their characteristics. The
robots have different tasks to perform depending on
the type of object.
Figure 1: Flexible manufacturing system.
Table 1 shows the ranges and speed limits of the
six Niryo Ned joints. In the case study presented in
Section 4, the speed is voluntarily limited to 60% of
the maximum value to reduce mechanical wear, par-
ticularly on the belts, which is one of the main causes
of failure.
3.2 Digital Twin and Protocol Network
Figure 2 shows the architecture of the digital twin,
which consists of three layers:
• The Simulation Layer acts as an interface with the
user, who can visualise robot movements in a sim-
ulated environment. It receives the position of the
physical robot joints and copies the trajectories in
real time.
• The Environment Layer contains information
about the layout of the flexible manufacturing sys-
tem. By receiving information about the position
Table 1: Parameters of the robotic arm.
Joint (type : rotation)
N
◦
Range Speed limit
1 −170
◦
⩽ θ
1
⩽ 170
◦
˙
θ
1
⩽ 150
◦
.s
−1
2 −120
◦
⩽ θ
2
⩽ 35
◦
˙
θ
2
⩽ 115
◦
.s
−1
3 −77
◦
⩽ θ
3
⩽ 90
◦
˙
θ
3
⩽ 140
◦
.s
−1
4 −120
◦
⩽ θ
4
⩽ 120
◦
˙
θ
4
⩽ 180
◦
.s
−1
5 −100
◦
⩽ θ
5
⩽ 55
◦
˙
θ
5
⩽ 180
◦
.s
−1
6 −145
◦
⩽ θ
6
⩽ 145
◦
˙
θ
6
⩽ 180
◦
.s
−1
of the parts on the pallet, it updates this informa-
tion for all the robots according to their reference
frame.
• The Communication Layer provides the interface
between the robots and the digital twin’s layers
thanks to an OPC UA server.
Figure 2: Global architecture.
Unlike many articles that used ROS services to
communicate between the physical robot and the dig-
ital twin, we use the OPC UA protocol, which is in-
creasingly used in industrial automation. This proto-
col uses the client-server and publish-subscribe prin-
ciples and offers a high level of interoperability for
communication between systems from different sup-
pliers. In our case study, the production system is
composed entirely of robots, but a cyber-physical pro-
duction system is composed of heterogeneous compo-
nents which may have different communication pro-
tocols. OPC UA’s advantage over ROS is that it is eas-
ier to integrate non-robotic components such as con-
veyors or cylinders into the digital twin, which bet-
ter reflects a real production system. In addition, one
of the major problems in integrating digital twins is
Cooperation and Synchronization of Robotic Tasks Using a Digital Twin
75

designing the data model to enable communication.
OPC UA offers the possibility of creating informa-
tion models, and working groups are developing them
to create standardisation. In robotics, there is the OPC
UA Robotics Companion Specification, which defines
a data model for describing the motion device system,
so that information from a robot can be passed up to
the upper layers of a manufacturing system, such as
the cloud.
Finally, the authors in (Profanter et al., 2019) find
that the OPC UA protocol is faster in real-time com-
munication than ROS, which is a significant advan-
tage in our use of the digital twin for information
transmission.
3.3 Environment Layer
The Environment Layer is fundamental to our ap-
proach to synchronizing and transferring information
between robots, thanks to its global vision of the sys-
tem. The robots’ perception is restricted by their sen-
sors: they only know their arm position thanks to the
joints’ value, but they do not know where they are
located concerning the other components of the sys-
tem. The controller authorises all movements within
the limits of the joint angles, although there may be
common working areas where there is a risk of colli-
sion if there is no synchronization.
Moreover, in our example, one of the robots is
under-equipped with a sensor, as it does not have a
camera that can locate parts for its pick-and-place
task. Therefore, cooperation is not only limited to
authorizing access to the work area but also to trans-
mitting the position of parts from the robot with the
camera. However, each robot has its reference frame,
which means that the position of a part sent by one
robot to another requires a change of reference frame.
The Environment Layer overcomes these prob-
lems by incorporating robot positions with production
system layouts, making it easier to design the control
system and quickly identify conflict areas. Figure 3
shows the Environment Layer flowchart. Initially, a
file containing the location of all system components
is imported into the application when the system is
designed, to calculate the changes in robot reference
points. During production, the Environment Layer re-
ceives the parts position from the robot with the cam-
era and converts it into the frame of the robots that
need it. Then, the layer synchronizes the robots’ tasks
according to their states.
Control adaptability is enhanced by the automatic
conversion of part position from one robot to another.
When a robot’s location is modified, the Environ-
ment Layer must be updated to reflect the new sys-
Figure 3: Environment layer flowchart.
tem configuration. Knowing the relative position be-
tween the robots, the layer then adjusts the reference
frame change to the new robot position, which re-
sumes its pick-and-place task without having under-
gone any code modifications.
4 RESULTS AND DISCUSSION
In this section, we present different scenarios to il-
lustrate the gains in production time and energy con-
sumption of the real system that can be achieved by
using the digital twin.
4.1 Case Study
This section shows the results obtained using the
robotic platform presented in the previous section.
Figure 4 details the hardware and software compo-
nents for implementing the cyber-physical production
system:
• The flexible manufacturing system consists of two
ICINCO 2024 - 21st International Conference on Informatics in Control, Automation and Robotics
76

Niryo Ned robots, a conveyor and a presence sen-
sor. Each robot is controlled by a Python program
developed using the PyNiryo library and executed
by their Raspberry Pi;
• The digital twin’s Simulation Layer is imple-
mented on the Webots software to monitor the
robots’ movements in real-time using an OPC UA
client (Webots, 2024) ;
• The digital twin’s Communication Layer is imple-
mented using Prosys OPC UA Simulation Server
software. This layer establishes the interface
between the various elements of the system by
implementing an OPC UA server (ProsysOPC,
2024) ;
• The digital twin’s Environment Layer is imple-
mented with a Python script that converts the
robots’ positions and synchronises the robotic
tasks.
Figure 4: Architecture implementation.
Three scenarios are evaluated to illustrate the in-
terest of the approach proposed in this paper. Four
different parts are arranged on a pallet and forwarded
to the robot work area on a conveyor belt. In the
three scenarios described below, the aim is to pick-
and-place some parts according to their characteris-
tics. For all figures, the plots start when the presence
sensor detects a pallet and stops the conveyor belt. In
an inertial reference frame associated with the base
of i-th robot,
−−→
O
i
M
i
represents the position vector of
i-th robot between the origin of the reference frame
and the point associated with the suction cup tip of
i-th robot. ||
−−−→
O
1
M
1
|| and ||
−−−→
O
2
M
2
|| denote respectively
the magnitude of the position vector of Robot 1 and
Robot 2.
Scenario 1 (Sc1): In this first scenario, each robot
is equipped with a vision set to locate on the pallet
the parts to be picked up and placed onto a dedicated
platform. In this case, the two robots work indepen-
dently to avoid any risk of collision. The first robot
performs all its tasks before sending a signal to the
second robot so that it can perform its tasks (see Fig-
ure 5). Each vision set associated with a robot checks
0 5 15 25 35 45 55 65 75
Time (s)
0
50
100
150
200
250
1
a
1
b
1
c
2
a
2
b
2
c
1
d
2
d
Robot 1
Robot 2
Figure 5: (Sc1) Robot 1 : ||
−−−→
O
1
M
1
|| ; Robot 2 : ||
−−−→
O
2
M
2
||.
to see if there are any parts left on the pallet, which
means that the scenario takes a long time, with unnec-
essary movements and long image processing times.
Figure 5 shows three cycles for Robot 1, followed by
three cycles for Robot 2. Peaks 1
a
and 2
a
appear when
the robots are in a high position, above the pallet, to
locate parts. When the vision set associated with each
robot detects a part with certain established character-
istics, the suction cup at the end of the robot arm picks
the component (peaks 1
b
and 2
b
) and places it on the
dedicated platform (peaks 1
c
and 2
c
). For each robot,
after peaks 1
d
and 2
d
, the robots return to their initial
position, as no more parts are present on the pallet to
pick. One can notice that to pick-and-place the four
parts, the process time is about 75s.
Scenario 2 (Sc2): In the second scenario, cooperation
between the two robots is introduced to enable them
to work simultaneously. A shared work area in which
only one robot can operate at a time is defined to en-
sure that only one robot can pick up a part at a time
and ensure collision-free synchronisation. If the area
is occupied by the other robot, the robot must wait for
authorization from the digital twin before picking up
a part. Moreover, Robot 1 informs the digital twin of
the parts presence to be picked up by Robot 2. Figure
6 shows that the robots take it in turns to pick up parts:
while one robot places a part in the drop zone, the
other picks up a part and so on until there are no more
parts to pick up. This scenario minimizes the time
during which the work area is not used by any robot
and limits unnecessary movements, reducing process
time by 40% compared to the first scenario.
Scenario 3 (Sc3): In the final scenario, we assume that
only Robot 1 is equipped with a vision set. The level
of cooperation has been increased compared to Sce-
nario 2, as Robot 1 no longer simply informs the dig-
Cooperation and Synchronization of Robotic Tasks Using a Digital Twin
77

0 5 15 25 35 45
Time (s)
0
50
100
150
200
250
Robot 1
Robot 2
Figure 6: (Sc2) Robot 1 : ||
−−−→
O
1
M
1
|| ; Robot 2 : ||
−−−→
O
2
M
2
||.
ital twin that there are parts to be picked up by Robot
2, but also provides the position of the parts. This sce-
nario uses the Environment Layer of the digital twin,
as described in subsection 3.3, to transmit and con-
vert part positions from Robot 1 to Robot 2. Figure 7
shows that the part picking alternates as in Scenario 2.
The main difference is that the movements of Robot
2 are significantly reduced, as it does not need to scan
the pallet using its vision system. In addition to the
56% reduction in process time compared with Sce-
nario 1 and 27% reduction compared with Scenario 2,
we can notice that Robot 2 consumes less energy than
the previous scenarios by making fewer movements
in this scenario (see Table 2).
0 5 11 22 33
Time (s)
0
50
100
150
200
Robot 1
Robot 2
Figure 7: (Sc3) Robot 1 : ||
−−−→
O
1
M
1
|| ; Robot 2 : ||
−−−→
O
2
M
2
||.
Table 2: Distance comparison.
.
Scenario 1 2 3
Robot 1 (m) 1.1024 1.0881 1.0624
Robot 2 (m) 1.3984 1.0724 0.556
Robots 1 + 2 (m) 2.5008 2.1605 1.6184
Improvement (%) - -13.61 -35.28
4.2 Discussion
The sum of cooperation, the global vision provided
by the Environment Layer and the data conversion
carried out by the digital twin have improved produc-
tivity and reduced energy consumption in our pick-
and-place application. Table 2 compares the distance
covered by the end effector of the two robotic arms
in the three scenarios presented in the previous sec-
tion. The first two rows of the table provide the dis-
tance covered by each robot arm, and the third row
shows the sum of these two distances. Finally, the
last row shows the improvement in terms of distance
covered in scenarios 2 and 3 compared to scenario 1.
One can notice a decrease in the total distance cov-
ered by the two robot arms of 13.61% in scenario 2,
and a decrease of 35.28% in scenario 3 compared to
scenario 1. These results illustrate the effectiveness
of the proposed approach.
In the customised production context, production
systems are subject to frequent configuration changes,
which may include layout rearrangements. These
modifications lead to the development of new robot
controllers by updating the new pick-and-place posi-
tions. In our approach, simply updating the locations
of the robots in the Environment Layer is enough to
adapt the control without modifying the robot code.
More generally, the aim is to link information from
the production environment to the control systems to
simplify modifications using the Environment Layer.
The Environment Layer provides additional infor-
mation to which robots have no access via their sen-
sors. This information gives the robots a broader
view, making them more robust to failures. There are
two advantages to transferring decision-making to the
digital twin: on the one hand, the robot can devote its
computing capacity to optimizing its trajectory, and
on the other, the digital twin can optimize tasks be-
tween the various robots in the system thanks to its
higher computing capacity.
5 CONCLUSIONS
In this paper, we have proposed a methodology show-
ing the benefits of using a digital twin for the coop-
ICINCO 2024 - 21st International Conference on Informatics in Control, Automation and Robotics
78

eration and synchronization of robotic tasks. Our ap-
proach is fully in line with Industry 4.0, using tech-
nologies such as the OPC UA protocol, which enables
interoperability and using machine vision with image
processing. In fact, by centralizing all the data re-
quired to perform the tasks of each robot, the digital
twin can reconstruct the working area and thus con-
trol each robot even if this last has a faulty sensor.
Moreover, thanks to the knowledge of this working
environment, the movement of each robot arm is min-
imized, thus reducing energy consumption and me-
chanical wear.
Future research will focus on developing a
decision-making layer in the digital twin, which will
be responsible for planning robot tasks and optimis-
ing image processing. In our case study, the robots
receive the position of the parts to be picked and au-
tonomously choose which one to pick. The decision-
making layer currently under development will in-
dicate the most efficient sequence of movements for
each robotic arm, to save even more time and energy.
ACKNOWLEDGEMENTS
The project was supported in part by Appel
`
a manifes-
tation d’int
´
er
ˆ
et (AMI) du R
´
eseau d’ESR du site Cham-
pardennais.
REFERENCES
Bajic, B., Rikalovic, A., Suzic, N., and Piuri, V. (2021).
Industry 4.0 implementation challenges and opportu-
nities: A managerial perspective. IEEE Systems Jour-
nal, 15(1):546–559.
Bansal, R., Khanesar, M. A., and Branson, D. (2019).
Ant colony optimization algorithm for industrial robot
programming in a digital twin. In 2019 25th Inter-
national Conference on Automation and Computing
(ICAC), pages 1–5. IEEE.
Barenji, A. V., Liu, X., Guo, H., and Li, Z. (2021). A dig-
ital twin-driven approach towards smart manufactur-
ing: reduced energy consumption for a robotic cell.
International Journal of Computer Integrated Manu-
facturing, 34:844–859.
Barricelli, B. R., Casiraghi, E., and Fogli, D. (2019). A
survey on digital twin: Definitions, characteristics,
applications, and design implications. IEEE Access,
7:167653–167671.
Erkoyuncu, J. A., Butala, P., Roy, R., et al. (2018). Digital
twins: Understanding the added value of integrated
models for through-life engineering services. Proce-
dia Manufacturing, 16:139–146.
Kousi, N., Gkournelos, C., Aivaliotis, S., Giannoulis, C.,
Michalos, G., and Makris, S. (2019). Digital twin for
adaptation of robots’ behavior in flexible robotic as-
sembly lines. Procedia manufacturing, 28:121–126.
Liang, X., Xiao, R., and Zhang, J. (2022). A review on
digital twin for robotics in smart manufacturing. In
2022 IEEE 17th Conference on Industrial Electronics
and Applications (ICIEA), pages 1510–1515.
Liu, X., Jiang, D., Tao, B., Jiang, G., Sun, Y., Kong,
J., Tong, X., Zhao, G., and Chen, B. (2022). Ge-
netic algorithm-based trajectory optimization for dig-
ital twin robots. Frontiers in Bioengineering and
Biotechnology, 9:793782.
Matulis, M. and Harvey, C. (2021). A robot arm digital
twin utilising reinforcement learning. Computers &
Graphics, 95:106–114.
Mohammed, A., Schmidt, B., Wang, L., and Gao, L. (2014).
Minimizing energy consumption for robot arm move-
ment. Procedia CIRP, 25:400–405.
Niryo (2024). https://niryo.com/.
Profanter, S., Tekat, A., Dorofeev, K., Rickert, M., and
Knoll, A. (2019). Opc ua versus ros, dds, and mqtt:
Performance evaluation of industry 4.0 protocols. In
2019 IEEE International Conference on Industrial
Technology (ICIT), pages 955–962.
ProsysOPC (2024). https://prosysopc.com/.
Sow, M. C., Assila, A., Garcia, D., Martinez, S., Zghal,
M., and Baudry, D. (2023). Towards the development
of a digital twin for micro learning factory: A proof
of concept. In International Conference on Extended
Reality, pages 284–298. Springer.
Wanasinghe, T. R., Wroblewski, L., Petersen, B. K., Go-
sine, R. G., James, L. A., De Silva, O., Mann, G. K. I.,
and Warrian, P. J. (2020). Digital twin for the oil and
gas industry: Overview, research trends, opportuni-
ties, and challenges. IEEE Access, 8:104175–104197.
Webots (2024). https://cyberbotics.com/.
Wu, Z., Chen, S., Han, J., Zhang, S., Liang, J., and Yang,
X. (2022). A low-cost digital twin-driven positioning
error compensation method for industrial robotic arm.
IEEE Sensors Journal, 22(23):22885–22893.
Cooperation and Synchronization of Robotic Tasks Using a Digital Twin
79