Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000.
Digital Object Identifier XX.XXXX/YYY.20XX.ZZZZZZZ
A 5G Automated Guided Vehicle SME
Testbed for Resilient Future Factories
DANIEL S. FOWLER, YUEN KWAN MO, ALEX EVANS, SON DINH-VAN, BILAL AHMAD, (Senior
Member, IEEE), MATTHEW D. HIGGINS, (Senior Member, IEEE), CARSTEN MAPLE
WMG, The University of Warwick, Coventry, CV4 7AL, UK
Corresponding author: Daniel S. Fowler (e-mail: dan.fowler@warwick.ac.uk).
This work was funded by the UK’s High Value Manufacturing Catapult.
ABSTRACT Factory automation design engineers building the Smart Factory can use wireless 5G broadband
networks for added design flexibility. 5G New Radio builds upon previous cellular communications standards to
include technology for "massive machine-type communication" and "ultra-reliable and low-latency communi-
cation". In this work, the authors augment an automated guided vehicle with 5G for additional capabilities (e.g.,
streaming high-resolution video and enabling long-distance teleoperation), increasing the mobile applications
for industrial equipment. Such use cases will provide valuable knowledge to engineers examining 5G for novel
smart manufacturing solutions. Our 5G private network testbed is a platform for wireless performance research
in industrial locations and provides a development mule for flexible smart manufacturing systems. The rival
wireless technology to 5G in industrial settings is Wi-Fi and it is included in the testbed. Furthermore, the
authors noted challenges, often unconsidered, facing the move to digital manufacturing technologies. Therefore,
the authors summarise the emerging challenges when implementing new digital factory systems, including
challenges linked to societal concerns around sustainability and supply chain resilience. The new Smart Factory
technologies, including 5G communications, will have their roles to play in alleviating these challenges and
ensuring economies have resilient future factories.
INDEX TERMS 5G mobile communication, AGV, AMR, CPPS, cyber-physical systems, industrial strategy,
integrated manufacturing systems, resilience engineering, smart manufacturing.
I. INTRODUCTION
GOVERNMENTS view industry and innovation as foun-
dational for a strong economy [1]. Innovation leads to
new products for manufacturers and improvements in man-
ufacturing efficiencies and cost reductions. The adoption of
smart manufacturing provides industrialists with new tools to
further improve their manufacturing processes. Smart manu-
facturing is a "method that improves its performance with
the integrated and intelligent use of processes and resources
in cyber, physical, and human spheres to create and deliver
products and services, while also collaborating with other
domains within an enterprise’s value chains" [2].
One of the technologies targeted at aiding the implemen-
tation of smart manufacturing is Fifth Generation New Radio
(5G NR) communications, simply known as 5G, a cellular
broadband telecommunications standard. 5G builds upon the
previous generations of cellular communications technology
to improve data transmission capabilities and define new use
cases [3], [4]. This allows industrial stakeholders to consider
novel smart manufacturing applications and new process
configurations.
5G and its industrial applications are an active area of
research [5], [6] with large industrial organisations testing its
capabilities. What is not clear is whether Small and Medium-
sized Enterprises (SMEs) will be able to benefit from 5G
in a cost-effective way. Growth in the take-up of advanced
wireless communications within factories requires knowl-
edge of what can be achieved in system implementations,
and practical evidence of 5G’s benefits to future factory
systems. That knowledge and evidence can be gained from
5G testbeds [7]. Here, the developed testbed use case is the
augmentation of an Automated Guided Vehicle (AGV) with
5G with the option to switch to Wi-Fi. The AGV utilises
the characteristics of 5G’s advanced wireless communication
technology to allow for testing improvements in wireless data
connectivity within industrial environments.
Introducing smart manufacturing technologies into facto-
ries does raise some challenges. This work re-examines some
of those challenges as it was seen that they are not always
addressed in the smart manufacturing literature. 5G commu-
nications can be used to help address those challenges for
future factories. Some of the challenges go beyond the level
of any deployed smart production systems as governments
IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX 1
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
Daniel S. Fowler et al.: A 5G Automated Guided Vehicle SME Testbed for Resilient Future Factories
are looking at industry to be resilient to global economic
and environmental pressures and contribute to sustainability
aims which are highlighted within the United Nations 17
Sustainable Development Goals (SDG) [8]. Advanced wire-
less communications will have a role to play in improving
manufacturing systems to help address those sustainability
aims.
The sections below begin (Section II) with a background
to the work and the motivations to build an AGV testbed
to study the implementation and use of 5G within man-
ufacturing. The ideas behind the Industry 4.0 concept, a
driver for smart manufacturing, are briefly revisited (II-A). A
summary of the range of smart technologies being applied to
manufacturing applications is provided (II-B), listing some
of their advantages. 5G wireless communications (II-C) is
summarised, and the new functionality that is useful to
manufacturing applications (II-D) covers some challenges in
5G manufacturing implementations. The 5G AGV testbed
use case is described in Section III, it includes a summary
of previous 5G AGVs (III-A), the WMG testbed (III-B),
some considerations on 5G vs Wi-Fi for SMEs (III-C), and
the WMG testbed in operation (III-D). The discussions in
Section IV examine the use cases for the testbed (IV-A) and
includes some challenges that implementing smart manufac-
turing technology can bring (IV-B). The renewed interest
in manufacturing resilience is examined (IV-C) along with
the role 5G may play in adding resilience to smart future
factories (IV-D). The work concludes in Section V. Table 1
provides a summary of the abbreviations and terms used in
this work.
II. MOTIVATION AND BACKGROUND
Manufacturers have always embraced technology to help
with the complexities of production, process control, deliv-
ering output and managing supply chains. Wireless commu-
nication is one of the technologies available for deployment
in factories, and manufacturers can look at the capabilities of
5G and advanced Wi-Fi (versions 6 or 6E and the upcoming
version 7) to add new abilities to their production systems
and enhance existing processes. This work is motivated by
the fact that AGVs in 5G networks have seen little use and
study. This has been highlighted in previous work [9]. Fur-
ther, there is a need to examine the claimed advantages and
understand technical considerations when deploying 5G in
smart manufacturing environments. Using the AGV testbed
will identify the challenges that automation engineers will
face in 5G system implementations and whether the claims
made for the technology are reliable. This is particularly
important for resource-limited SMEs whose growth could be
affected if investments in systems do not meet expectations.
Moving beyond the technical challenges of deploying new
smart manufacturing technology, there are risks around the
safety, security, and life cycle management of a deployed sys-
tem. These risks can be viewed as a variety of potential dis-
ruptors that may weaken production output despite the smart
technology deployment. To mitigate the effect of disruptors
TABLE 1. Abbreviations and terms in this work
Abbrev./Term Meaning
4G Fourth-generation cellular telecommunications
5G Fifth-generation cellular telecommunications
AI Artificial Intelligence
AMPS Advanced Mobile Phone System
AMR Autonomous Mobile Robot
AGV Automated Guided Vehicle
AR Augmented Reality
BCMS Business Continuity Management Systems
C&C Command and Control
CAD Computer Aided Design
CAM Computer Aided Manufacturer
CDMA Code Division Multiple Access
CN Core Network (or simply Core)
CNI Critical National Infrastructure
CPPS Cyber-Physical Production System
CPS Cyber-Physical System
D-AMPS Digital AMPS
DT Digital Twin
eMBB Enhanced Mobile Broadband
eNodeB or eNB Evolved Node Base station (4G)
EPC Evolved Packet Core (packet core for data and voice)
eSIM Electronic SIM
FMS Flexible Manufacturing Systems
gNodeB or gNB Generalised Node Base station (5G)
GSM Global System for Mobile
HSPA High Speed Packet Access
I4.0 Industrie 4.0 (Industry 4.0)
IIoT Industrial Internet of Things
IS-95 Interim Standard 95
IMSI International Mobile Subscriber Identity
IMT International Mobile Telecommunications
ITU International Telecommunications Union
IT Information Technology
KPI Key Performance Indicator
LARG Lean, Agile, Resilient, and Green
LiDAR Light Detection and Ranging
LoRaWAN Long Range Wide Area Network
LTE Long Term Evolution (data packet cellular comms)
MES Manufacturing Execution System
MiR Mobile Industrial Robots (a company)
ML Machine Learning
mMTC Massive Machine Type Communications
mmWave Millimetre wave
NV Network Virtualisation
NR New Radio, i.e., 5G NR
NSA Non-standalone
NTC Nordic Mobile Telephony
OT Operation Technology
PDC Personal Digital Cellular
PLC Programmable Logic Controller
QoS Quality of Service
RAMI4.0 Reference Architecture Model Industry 4.0
RE Resilience Engineering
RF Radio Frequency
SA Standalone
SIM Subscriber Identity Module
SIMPLE Smart Information Platform and Ecosystem
SDG Sustainable Development Goals
SDN Software Defined Network
SME Small and Medium-sized Enterprises
SMS Short Message Service
TACS Total Access Communication System
UE User Equipment (a wireless device)
OPC UA Open Platform Communications Unified Architecture
URLLC Ultra-Reliable and Low Latency Communications
UMTS Universal Mobile Telecommunications System
USB a Universal Serial Bus
VR Virtual Reality
WCDMA Wideband Code Division Multiple Access
Wi-Fi Wireless Fidelity (the IEEE 802.11 family)
2IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
Daniel S. Fowler et al.: A 5G Automated Guided Vehicle SME Testbed for Resilient Future Factories
strategies can be used to add resilience to any form of system.
How smart manufacturing technologies intersect with aspects
of systems resilience is an area of research that testbeds
can support. Since there is a link between the resilience
of deployed smart manufacturing systems and the overall
industrial domain resilience (mentioned in the Introduction),
the work examines the aspects of resilience applicable to
future factories and their deployed smart technologies.
A. REVISITING THE OBJECTIVES OF INDUSTRY 4.0
The Industrie 4.0 initiative, Industry 4.0 or simply I4.0, was
a recognition by Germany that the technologies enabling the
emergence of smart factories could threaten their position
as a leading manufacturing country [10]. In I4.0, the man-
ufacturing process is integrated into a network of connected
Cyber-Physical Systems (CPS) used to optimise the entire
supply chain, from initial order to final delivery. Every step
that enables a product or service to be designed, produced,
and delivered is instrumented to enable large amounts of
data to be analysed and acted upon. This enables fine-tuning
of these Cyber-Physical Production Systems (CPPSs) [11]
for maximum efficiency, minimum cost, flexibility, and zero
downtime. Online ordering and delivery to consumers via
sophisticated logistics systems are well established, having
benefited from advances in technology and software. The
deployment of the Industrial Internet of Things (IIoT) is seen
as a similar enabler for I4.0. Some of the facets that I4.0 is
seen to provide include:
•Data feeds from high-density embedded sensor net-
works.
•Granular identification of products, their location and
history.
•Autonomous machine-to-machine communication and
control.
•High levels of integration between all processes and
management information systems.
•Application of data analytics, machine learning and
other artificial intelligence techniques.
•Efficiency gains in material usage, resources, and en-
ergy consumption.
•Improving the lifecycle management of assets, includ-
ing improvements in preventive maintenance.
•Real-time management of dispersed processes and their
environments, reducing time lags between processes
and events.
•Support for on-demand highly customisable products
and services.
•Reducing environmental impact.
•Reduced use of humans for routine tasks, improving
their working environment.
A variety of the above facets of I4.0 are seen in the new
smart manufacturing systems that new digital technologies
are enabling.
B. A BRIEF OVERVIEW OF NEW SMART
MANUFACTURING TECHNOLOGIES
There is a diverse range of technologies that can be encap-
sulated within the realm of smart manufacturing [13]. These
include:
•Advanced digital-based manufacturing
•Industrial Internet of Things (IIoT) and smart sensors
•Big data analytics and visualisation
•Machine Learning and Artificial Intelligence
•Containerisation and virtualisation of apps and services
•Digital Twins (DT)
•Advanced wireless communications (5G/Wi-Fi 6/Lo-
RaWAN)
•Software Defined Networks (SDN)
•Autonomous robots, cobots, and teleoperation
•Augmented Reality (AR) and Virtual Reality (VR)
•Cloud computing
•Additive Manufacturing
This list of digital technologies undoubtedly brings many
possibilities and advances to manufacturing systems, how-
ever, it raises new challenges for the industrialists that want
to implement these technologies. They need to be integrated
into the Operational Technology (OT), Information Tech-
nology (IT), and management systems of the manufacturing
organisations that wish to utilise these technologies. Figure 1,
derived and expanded from [12], illustrates the complexities
of implementing smart manufacturing systems. There are
various interconnecting layers and systems, different types of
devices and machinery, and legacy equipment and systems.
The data and communication demands of this pyramid of
technologies and entities could benefit from the improved
capabilities that today’s high-speed digital wireless commu-
nications could offer.
To maintain control of their factories, the stakeholders
within smart manufacturing systems need access to tools and
architectures that can aid the organisation and design of a
complex CPPS. For example, the Smart InforMation PLat-
form and Ecosystem (SIMPLE), see Figure 2, is designed as
scaffolding or a framework for smart manufacturing: "The
SIMPLE platform development...objective is to stimulate the
development of manufacturing-specific but cross-sector and
cross-industry digital capabilities" [12]. The SIMPLE modu-
lar and software-driven architecture is reliant upon connectiv-
ity, not only in terms of the physical connections but how data
is connected, translated and moved between entities. Physical
communication, whether wired or wireless, is the conduit for
data connectivity. The capabilities of 5G technologies adds
more options to the design of smart manufacturing systems
that implement SIMPLE or similar frameworks.
C. NEW RADIO TECHNOLOGIES, 5G AND WI-FI 6
The success of mass-market global telecommunications re-
sults from the collaboration of multiple organisations and
stakeholders. Universally implemented technical standards
have enabled the rapid growth and worldwide spread of cellu-
IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX 3
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
Daniel S. Fowler et al.: A 5G Automated Guided Vehicle SME Testbed for Resilient Future Factories
FIGURE 1. A pyramid of manufacturing technologies that illustrates the potential complexity within smart manufacturing systems [12]
.
FIGURE 2. The SIMPLE framework to support smar t manufacturing
technologies, which is derived from [12].
lar communications. Each generation of telecommunications
technology, Table 2, increases the capabilities of communi-
cations, which in turn drives demand for more bandwidth
and features as new applications and use cases emerge. 5G
is the newest release of cellular communications deployed.
It is another step up in capabilities over the previous gener-
ations. This includes the introduction of features to enable
additional use cases for International Mobile Telecommuni-
cations (IMT) as stated by the International Telecommunica-
tions Union (ITU) in 2015. These new usage scenarios were
stated as [14]:
•Enhanced Mobile Broadband (eMBB) - This is targeted
at the human use of multimedia content, services and
data. A higher data capability drives the take-up of
higher-density data services. Enhancing mobile broad-
band capability at high user density events (i.e., crowded
locations) and improving data rates for roaming users.
This is beneficial for the users, service providers and
businesses that can use those services.
•Ultra-Reliable and Low Latency Communications
(URLLC) - Improving data throughput, latency, and
availability are beneficial for several domains that want
to engineer new advanced systems, these domains in-
clude industry, medical systems, power grids, smart city
automation, and transportation.
•Massive Machine Type Communications (mMTC) -
Many multiples of devices are deployed to provide
data, e.g., from sensor networks. These devices may be
required to operate for many years on battery power.
Applications can include remote monitoring of instal-
lations, building monitoring, providing agriculturally
data, and smart cities.
Each of the generations of cellular communications has
needed to support the older generations to enable the tech-
nology to transition over several years across the global mar-
4IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
Daniel S. Fowler et al.: A 5G Automated Guided Vehicle SME Testbed for Resilient Future Factories
TABLE 2. Fifth Generation New Radio (5G NR), simply known as 5G, is a recently deployed global release of mobile cellular communications
Generation of mobile
communications (nG) The primary decade of
deployment and take up Principle technologies used in cellular communications across the world
1G 1980s Analogue transmission, Total Access Communication System (TACS), Advanced Mobile Phone
System (AMPS), Nordic Mobile Telephony (NTC)
2G 1990s Switch to digital transmission, support for Short Message Service (SMS), Global System for Mobile
(GSM) became a dominant technology, Code Division Multiple Access (CDMA) was specified in
Interim Standard 95 (IS-95), others included Digital AMPS (D-AMPS) and Personal Digital Cellular
(PDC)
3G 2000s Introduction of mobile data services, High Speed Packet Access (HSPA), Wideband Code Divi-
sion Multiple Access (WCDMA) used in Universal Mobile Telecommunications System (UMTS),
CDMA2000 (successor to 2G’s CDMA)
4G 2010s Long Term Evolution (LTE), and revisions, supporting packet switching on an IP network
5G 2020s New Radio (NR) with frequencies up to 6GHz and at millimetre wave (mmWave) over 24GHz,
supporting defined use cases are Enhanced Mobile Broadband (eMBB), Massive Machine Type
Communication (mMTC) and Ultra-Reliable and Low-Latency Communication (URLLC)
FIGURE 3. 5G Non-Standalone vs Standalone.
ket [15], [16]. This has resulted in two primary methods for
a 5G device to connect to a 5G network. A 5G device, in the
United States and China, will initiate a connection directly
with a 5G base station. This is called standalone or SA. In the
rest of the world, including Europe, a 5G device will initiate
a connection with a 4G base station and then be passed to the
5G network. This is called non-standalone or NSA [17]. See
Figure 3. In both cases, if no 5G network is available then
4G or lower is used. Devices that are programmed to operate
in only NSA mode may present a problem to private 5G
networks that only deploy 5G only base stations. Overtime
devices will need to be programmed to support standalone as
5G base station nodes and core networks become the most
prevalent.
D. 5G IN MANUFACTURING
Manufacturing plants typically consist of several fixed ma-
chines connected via wired Ethernet networks. The intro-
duction of new technologies to the manufacturing space has
highlighted a requirement for reliable wireless communica-
tion in industrial environments. AMRs (Autonomous Mobile
Robots), augmented reality and virtual reality are transform-
ing manufacturing and IIoT projects are driving an increasing
demand for sensor data, sometimes in locations that do not
lend themselves to being wired. Note, the term AMR covers
all forms of mobile robots, including wheeled, e.g. AGVs,
and legged systems, e.g. the infamous Boston Dynamics Spot
"dog" robot.
Typical enterprise Wi-Fi networks are vulnerable to
changes in the number of connections and the quantity of
data being exchanged. Wi-Fi-based components will employ
strategies to limit the load on the network. An AMR will
use onboard maps, missions, and obstacle avoidance routines,
requiring a minimal exchange of commands and status data
over Wi-Fi to complete a task. Video-based devices may
employ compression/decompression routines that introduce
additional processing time.
5G offers the potential to expand the capabilities of these
technologies and present new opportunities and potential use
cases for flexible connected manufacturing. Features such as
dedicated channels and the lower latency promised in the
specification should lead to more deterministic communica-
tions, allowing for some aspects of control to be handled on
the edge. The network can also be private, dedicated to a
particular site, which would allow the mix of channels and
features to be tailored to the company’s requirements.
This work has looked at some of the use cases for advanced
wireless communications, Table 3. The use cases have been
placed under vision, IIoT, logistics, flexible manufacturing,
and safety categories. The advantages of the use case to
manufacturing are summarised, as are the advantages over
a wired solution. The challenge of implementing the given
use case on established Wi-Fi technology is summarised,
and the advantages of using 5G are listed. The final column
provides a suggestion of the kind of 5G equipment that would
be required within the marketplace to realise the use case.
At the time of this work, it was observed that there are a
limited number of dedicated industrial devices with 5G built-
in. The availability of manufacturing-specific 5G equipment
is mainly concentrated on the provision of a router to send
data collected by other equipment over a 5G network.
The availability of industrial equipment and sensors with
IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX 5
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
Daniel S. Fowler et al.: A 5G Automated Guided Vehicle SME Testbed for Resilient Future Factories
TABLE 3. Use cases for industrial 5G
Category Use case Use case advantages Versus wired comms Wi-Fi challenges 5G advantages Realising 5G
Vision Augmented
Reality Enhanced work
instructions; machine
status; visually guided
maintenance
Freedom of
movement in and
around machines,
indoors or outdoors
Slow network due to
edge-based image
processing data
volumes, impacting
other network traffic
High data rates;
dedicated channels 5G headsets with
built-in 5G or adapter
and good uplink rates;
engagement with AR
hardware companies
Virtual
Reality Virtual models
inspection/testing; training
away from the factory floor;
training in controlled
environments
Freedom of
movement Large amounts of
video data slowing
networks
Ditto 5G routers to replace
existing wired and
wireless connections
Inspection Powerful edge computing
image analysis; remote
inspection with mobile
robots
Mobility; indoors or
outdoors; elevated
structures or
equipment
Providing a high
capacity and reliable
edge connection
Ditto 5G routers to replace
wired connections
IIoT Predictive
mainte-
nance
Retrofit sensors, e.g.,
vibration or temperature;
edge computing analytics
Greater freedom in
sensor positioning
with less disruption
Reliability A large number and
lower power sensor
and compute nodes
Sensors development;
engagement with
sensor firms; solving
power routing
Logistics Enhanced
AGV
operations
Shared obstruction and
traffic information for route
optimisation
n/a Channel connection
limitations and
latency, channel
availability at
transmission time
Dedicated
connections, low
latency
5G routers fitted to
AGVs, software
development
Identifying obstructions n/a Ditto Dedicated,
high-capacity
channels
5G routers, suitable
cameras, and image
recognition software
are available
Factory map or 3D plan
updates n/a Ditto Ditto 5G routers available;
use of point clouds
from laser scanners
possible; software
development
Edge control of vehicles;
flexibility with
edge-controlled missions
and course adjustments;
edge-to-center coordination
n/a Ditto Dedicated
connections; low
latency
Low-cost 5G vehicles,
control software for
routing
Coordinated
delivery and
shipping
Optimised and coordinated
deliveries and collections
via public 5G networks;
real-time estimated arrivals;
loading bay locations
n/a Local router limits Utilise widespread
public networks;
small data
requirements can be
achieved with 4G
5G router availability
Flexible
manufacture Flexible
distributed
control
Simpler and lower cost
stations for flexibility, quick
reconfiguration and
enhancement; rapid product
changes and demand
matching; central or edge
control; different controller
code developed and tested
via DTs or virtual
commissioning
Increased flexibility
to move, add, and
change stations and
their operation
A limited number of
connections,
connection latency,
network availability
at time of
transmission
A large number of
connections, low
latency, dedicated
communication
channels
Currently available soft
Programmable Logic
Controllers (PLCs) or
standard PLCs used as
centralised controllers;
5G router availability;
testing of automation
protocols over 5G;
engagement with end
users on suitable
processes
Movable
stations Local control of
general-purpose automated
stations using wireless
communications;
coordinated centrally or at
the edge; quick
augmentation and
rearrangement to meet
demand or process changes;
spare capacity hire
Improves flexibility
to move stations Ditto A large number of
connections;
dedicated
communication
channels
5G routers available to
replace wired
connections; existing
production planning
software may need
additional development
Safety Define safe
areas Scanning for hazards;
distinguishing between
humans and mobile robots
and their proximity to
machines; data and image
processing at the edge
Additional options in
sensor positioning
with less disruption
Connection latency
could lead to timeouts
and false trips or
delayed cut-offs
Low latency with
dedicated connections Justifying investment
vs wired systems;
engagement with
safety manufacturers;
development of 5G
enabled safety systems;
approvals processes
6IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
Daniel S. Fowler et al.: A 5G Automated Guided Vehicle SME Testbed for Resilient Future Factories
FIGURE 4. The Ventus 5G enabled industrial sensor (microphone and
vibration pickup) developed at the Advanced Manufacturing Research Centre
North West, image courtesy of The University of Sheffield.
built-in native 5G communications ability is required to aid
5G industrial take-up. An example of a 5G sensor is shown
in Figure 4. It is in use at The University of Sheffield’s
Advance Manufacturing Centre North West. The sensor uses
a microphone and vibration pickup to monitor the machining
operation on a part. It sends data to a cloud compute server
that can apply a Machine Learning (ML) model to fine-
tune the machining speed. The machining speed adjustment
caters for minute variations in the parts material and the
cutting head to improve the quality of the machined surface.
Whilst, ML can be run at the edge, more demanding ML
may need cloud computation resources and the low latency
communications provided by 5G.
Manufacturers will need to evaluate the benefits of using
5G in their factory networks against the additional costs of
installing and running a 5G network. A 5G network instal-
lation will require specialist equipment and antennas. The
equipment would require experts in mobile communication
to install and configure and maintain it. An operating licence
or subscription to a mobile operator is required. There is
a need for manufacturers to gain or bring in 5G technical
knowledge and for 5G equipment to become as straight-
forward to install as more familiar networking equipment.
This issue may be particularly relevant when considering
the needs of SMEs that may lack time and expertise in 5G
technology.
The obvious competing technology for 5G in the factory
is Wi-Fi, commonly seen in enterprise IT systems. The
latest version of Wi-Fi is Wi-Fi 6 (IEEE 802.11ax) with an
increment in capabilities over the previous standards (IEEE
802.11ac and IEEE 802.11n). Wi-Fi networks require a sim-
pler installation. Once a Wi-Fi router has been purchased,
most organisations would have all the expertise, necessary
to install and maintain the network, in-house. This coupled
with the improved performance of Wi-Fi 6 could challenge
some use cases for 5G. One use of the 5G AGV testbed will
be to test the limitations of 5G and Wi-Fi 6 technologies.
III. 5G AGV MANUFACTURING USE CASE
The choice of an AGV as a 5G platform supports several
of the categories and use cases listed in Table 3. AGVs
are traditionally used to aid in logistics, moving parts and
goods around factories and onward to loading bays for fi-
nal dispatch. AGVs, and AMRs in general, are being used
FIGURE 5. A robot arm on an AGV can be used as an automated aid for
production and human operatives, i.e., as a cobot.
in Flexible Manufacturing Systems (FMS), and in cobot
applications to integrate automation into human-orientated
manufacturing tasks. In Figure 5 a robot arm mounted on
a MiR100 AGV is designed to support human operators in
specialised manufacturing operations.
A 5G AGV will support safety and surveying operations,
autonomously monitoring factory spaces for hazards and
supporting factory operations. As new use cases for manu-
facturing DTs emerge a variety of methods to keep the data
within DTs updated and accurate will be required. AGVs and
AMRs as autonomous surveying systems are one of those
methods. Furthermore, surveying functionality is useful for
wireless site surveys to help determine potential sources of
RF interference within factories.
A. EXAMPLES OF DEPLOYED 5G-ENABLED AGVS
The argument for deploying 5G within factories has been
provided for several years [18]. A summary of AGVs and
AMRs and their relation to 5G and relevant applications can
be found in [9], though it finds only one realised AGV, a
robot lawnmower teleoperated over a 5G connection utilising
Huawei equipment [19]. An AGV is controlled over 4G/5G
Ericsson equipment in [20]. They choose a couple of Key
Performance Indicators (KPIs), the guidance error and the
current consumption. The argument provided is that a lower
latency connection should provide improved control of posi-
tioning and less power consumed on course corrections. The
5G connection saw a 3.1mm average guidance correction and
2.78A average current consumption compared to 4.8mm and
2.47A over 4G. The 5G control improved guidance accuracy
whilst reducing current consumption by 11%. They further
showed that increased link latency and packet loss in a 5G
link would increase the AGVs guidance error and hence its
current consumption. This limited experiment does start to
IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX 7
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
Daniel S. Fowler et al.: A 5G Automated Guided Vehicle SME Testbed for Resilient Future Factories
gather evidence for the advantages of 5G communications
for use in smart factories.
In [21], a remotely located ML model is used to anticipate
AGV guidance errors and provide corrective actions when
a 5G link experiences perturbations. Wireless links can be
subject to degradation and if the guidance of an AGV is not
computed locally then such predictive methods could help
maintain performance levels.
5G is chosen as the communications link in [22] because of
its performance advantage over 4G and Wi-Fi. In that work,
two AGVs communicate over 5G to enable the transport of
a large load. The 5G link enables one AGV to track the
movement of another AGV with a small margin of error,
enabling the conveying of a shared load. This demonstration
of AGV cooperation shows the value of low-latency wireless
communication links. A similar leader-follower configura-
tion of AGVs is presented in [23], where a Kalman filter is
used to reduce the control delay in the 5G link to improve
the tracking accuracy of the following AGV. Indeed, time-
sensitive applications are a reason to choose a 5G link over
earlier wireless technologies.
Ensuring a consistent Quality of Service (QoS) in an
environment where 5G is deployed is the focus of [24]. They
argue that factories and logistics environments change over
time and those changes can impact the quality of the wireless
connections and the required configuration of equipment. An
AGV is used to monitor communications KPIs in a 10,000m2
area. The AGV is seen as an efficient solution compared to
human walking surveys. The data collected is relayed to an
aggregating KPI analysis dashboard. The solution is stated
as providing cost savings between 30% to 70% compared
to previous testing regimes. The use of an AGV, AMR, or
drone for automating signal surveys to aid the determina-
tion of antenna locations and wireless equipment parameter
configuration makes sense, particularly for larger industrial
installations. Our own research on 5G NR signal character-
istics within industrial environments [25] confirms that the
theoretical maximum settings for configuration parameters
will vary, hence the need for easy-to-perform wireless site
surveys within industrial spaces.
These examples of 5G-enabled AGVs have been built for a
specific experimental purpose. The authors’ AGV platform is
designed to support our research on the smart manufacturing
use cases listed in Table 3. Further, it will interface with other
manufacturing cells within WMG’s industrial laboratories.
B. THE WMG 5G AGV TESTBED
To evaluate the use of 5G within industrial applications and
provide a platform to trial the benefits, and challenges, of
using 5G communications in smart manufacturing use cases,
particularly for SMEs, a testbed has been created. The testbed
chosen combines an industrial AGV with a 5G router. The
AGV is from Mobile Industrial Robots (MiR), their MiR100
AGV provides a mobile platform for industrial applications
and logistics, see Figure 6. The standard communications of
the MiR100 is via traditional Wi-Fi (IEEE 802.11a/n/g/ac).
FIGURE 6. An industrial mobile guided platform vehicle is used in the 5G test
case (image courtesy of Mobile Industrial Robots).
FIGURE 7. Overview of the 5G-enabled guided vehicle.
The 5G router increases the communications capabilities
of the AGV. The addition of 5G provides the ability to
support high-density data applications, e.g., high-resolution
digital video without the need to share the Wi-Fi network
with other office and factory devices. This has advantages
for both the AGV and the Wi-Fi network. For the AGV, it
enables improved communications bandwidth and latency.
Improved latency is beneficial for applications that include
teleoperation and remote feedback control. The benefit for
the office/factory network is it enables the network to func-
tion without the bandwidth being consumed by high data vol-
ume smart manufacturing applications. This can contribute to
the resilience of operations. Running the factory systems on
separated communications channels can contribute to perfor-
mance and security goals (discussed further in Section IV-A).
A schematic of the 5G AGV is shown in Figure 7. An
equipment battery is used to supplement the AGV’s drive-
train battery, enabling the AGV to operate for longer. An Intel
RealSense digital camera, incorporating Light Detection and
Ranging (LiDAR) is linked to a video processing workstation
via a Universal Serial Bus (USB) connection. The processing
computer runs LiDAR software. The 5G router is linked to
the AGV and the computer workstation over Ethernet via
a pass-through gateway. The AGV is configurable with a
Siemens SCALANCE MUM856-1 industrial 5G router or a
Netgear 5G router. Testing a variety of different router makes
is one of the testbed’s uses.
The 5G router connects via a SA 5G connection to a Gen-
eralised Node Base (gNB) base station. The gNB links via a
5G switch to the 5G core providing a 5G SA private network.
The private network is licensed by the British regulator the
8IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
Daniel S. Fowler et al.: A 5G Automated Guided Vehicle SME Testbed for Resilient Future Factories
FIGURE 8. Overview of the Wi-Fi-enabled guided vehicle.
Office of Communications (Ofcom). The AGV’s 5G router
exposes the web interface to the private 5G network and port
forwarding is used to expose the workstation.
The command and control (C&C) monitoring workstation
for the AGV testbed connects to the same 5G private network.
The connection is through a 5G SA mobile phone that is
acting as a 5G modem. This modem is connected to the
C&C workstation via USB. The C&C workstation provides
Mobile Edge Computing (MEC), a.k.a. Multi-access Edge
Computing capability. The MEC functionality allows for a
connection to the AGV’s exposed web interface, via a ROS
(Robotic Operating system) bridge. Furthermore, there is the
ability to remote desktop onto the AGV and to access the
LiDAR streaming port.
The 5G AGV can be used for Wi-Fi communications ap-
plications and Wi-Fi testing. A schematic of the AGV in the
Wi-Fi configuration is shown in Figure 8. The configuration
is similar to the 5G configuration with the same software,
however, Wi-Fi routers are used instead of 5G routers and
modems. Then the AGV connects via Wi-Fi to an access
point (AP). The C&C workstation for the AGV connects to
the AP via USB. Figure 9 shows a picture of the completed
AGV in use.
C. 5G AND WI-FI 6/6E CONSIDERATIONS FOR SMES
Manufacturing SMEs needing to implement smart manufac-
turing use cases will require advanced wireless communica-
tions (see Table 3). They would consider deploying Wi-Fi
since organisations have experience with Wi-Fi from their
IT operations. However, 5G communication is seen as an
important technology for industrial economies [26]. SMEs
will need assistance in determining whether 5G or Wi-Fi
6/6E, or both, will meet their requirement.
Note that our 5G configured AGV system has the ad-
ditional equipment that is associated with the 5G private
network, and that this additional equipment sits between
the AGV and the C&C workstation. This means the 5G
FIGURE 9. The completed 5G MiR100 AGV carrying a sub-assembly.
communications path has extra steps compared to the Wi-Fi
path. Despite this added complexity, the 5G communications
performance is considered superior to the Wi-Fi performance
when correctly configured and managed, particularly under
heavy use, where QoS becomes important. QoS is critical
for manufacturing applications. The higher 5G performance
compared to Wi-Fi is due to the different nature of their
design:
•Wi-Fi uses Carrier Sensing Multiple Access with Colli-
sion Avoidance. Each wireless node in the system will
listen for idle radio transmissions before sending. The
transmitting node will send a Request to Send message
and wait for a Clear to Send message before transmis-
sion. Not having the ability to sense the channel while
transmitting results in inefficiency [27].
•In cellular communications time slots are allocated on
the downlink and uplink and assigned to nodes as and
when they are needed over a fixed period, although there
is some wastage in the system collisions are generally
avoided.
•Wi-Fi has the Control Plane inside the Data Plane,
whilst 5G (and 4G) have a separate Control Plane and
Data Plane.
•Wi-Fi’s unlicensed spectrum may result in access con-
tention issues from unneeded devices if the Wi-Fi net-
work is not correctly configured and managed.
It can be considered that 5G is a proactive approach [28]
to handling message transmission and Wi-Fi is a reactive
approach. Wi-Fi has a bandwidth drop when the probability
of collision is high [29]. This means that for large numbers
of nodes, high data bandwidth, and low latency applications
5G communications will provide the required performance
for future flexible manufacturing applications [9]. It was
shown in [30] that Wi-Fi 6 can outperform 5G in certain
situations where the distance to the antenna is short, whilst
IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX 9
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
Daniel S. Fowler et al.: A 5G Automated Guided Vehicle SME Testbed for Resilient Future Factories
FIGURE 10. 5G AGV link performance.
5G’s performance can hold up as distance increases. Further,
they state that achieving theoretical maximum data through-
puts is unlikely, which our tests in industrial environments
show [25]. Additional evidence is required to support all
of these claims, and our 5G AGV testbed will support the
experimentation required to build up the evidence knowledge
base and provide the industrial benchmarks that have been
stated as lacking [31].
D. 5G AGV TESTBED IN OPERATION
Validating the performance of 5G within manufacturing is
one aim of the research project. The 5G link performance for
the AGV testbed is shown in Figure 10. The link is providing
a download/upload of 163/54 Mbps with a latency of 19±6.6
ms. When testing the streaming of video from the AGV it
is providing no noticeable lag when compared to the lab’s
existing Wi-Fi which does demonstrate lag under the same
streaming conditions. One aspect of the link performance of
interest for manufacturing applications is the latency figure.
Latency is affected by several hardware and software factors,
having a reliable low-latency connection with a target in
the single-digit millisecond range aids the implementation of
time-sensitive applications. This was a reason 5G was chosen
for the wireless link in some of the AGV research discussed
in Section III-A.
Another application of the 5G AGV testbed is to test AGV
teleoperation over wide area networks. An operator located
in Hamburg, Germany, was able to use the AGV located
in a laboratory on the University of Warwick campus in
England. Usage included negotiating obstacles within the
laboratory. The AGV control link was over public 4G/5G
cellular networks with the final hop over the testbed’s 5G
private network for a total distance of 1109 kilometres.
These types of high bandwidth and low latency applica-
tions and reasons to deploy faster wireless infrastructure. An
automated assembly manufacturing cell may not generate
large amounts of data. Figure 11 shows a plot of the amount
of configuration and monitoring data to/from a production
cell. The Open Platform Communications Unified Architec-
ture (OPC UA) format is used to exchange data between the
monitoring and configuration server and the cell. During the
normal operational phases of the cell, the data packets are a
few hundred bytes or between 3000 and 4000 bytes in size,
with a few peaks higher during initiation phases.
Furthermore, the rate of the data packets is modest. Fig-
ure 12 is displaying mainly 10 or 40 packets per second, again
with a few higher peaks at phase transitions. This shows that
SMEs do not need 5G communications as a replacement for
hardwired existing functional systems. SMEs can assess 5G
for when they want to add new features (see Table 3) to their
manufacturing processes.
IV. 5G FOR RESILIENT FUTURE FACTORIES
During this research, it has been noted that the use of 5G
within factories for smart manufacturing applications is in
its infancy. 5G’s specific goals for the URLLC and mMTC
use cases will require readily available hardware and sup-
porting software, and the knowledge to implement 5G within
industrial locations. The above 5G AGV testbed is designed
to accelerate the knowledge of practical 5G applications.
Furthermore, the research has found other considerations
around the challenges in the move to digital manufacturing.
A. 5G AGV TESTBED USE CASES
In addition to supporting some of the industrial use cases
listed in Table 3, the 5G AGV provides an adaptable platform
for use as a testbed for advanced industrial wireless commu-
nications in a variety of manufacturing and logistic situations.
The use cases include:
•as a research tool and 5G demonstrator within factory
settings;
•for testing the transmission of high-definition video and
other high-density data over 5G and Wi-Fi within a
factory application;
•for performance testing of advanced wireless communi-
cations and communications equipment targeted at man-
ufacturing applications in an industrial environment;
•testing autonomous 5G and RF surveying techniques in
industrial spaces;
•as a platform to test cloud versus mobile edge process-
ing in manufacturing CPSs;
•to investigate cybersecurity issues and mitigation tech-
niques in manufacturing CPSs;
•the platform’s private 5G network is used for the testing
of 5G’s designed-in usage scenarios (see Section II-C);
•to analyse the beneficial claims for 5G within factory
locations;
•to research new uses of AGVs and AMRs in factory
spaces;
•testing the interaction of AGVs and AMRs with FMS
cells and cobot operations (Figure 13 shows the 5G
AGV delivering an assembly to a robot station).
Several facets contribute to the overall performance of end-
to-end communications within 5G systems. As stated in the
list above, one of the goals of the industrial 5G AGV testbed
is to research the performance of 5G within industrial set-
tings. Knowledge of 5G performance factors within industrial
spaces will be beneficial to industrial systems designers and
automation engineers. The factors for research include:
10 IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
Daniel S. Fowler et al.: A 5G Automated Guided Vehicle SME Testbed for Resilient Future Factories
FIGURE 11. OPC UA data size in a manufacturing cell.
FIGURE 12. OPC UA data packet density.
FIGURE 13. 5G AGV delivering an assembly to a robotic FMS station.
•the capability of data processing and sensor systems,
including the performance of the processor, system
memory considerations, and network card or interface
performance;
•data channel issues, including the number of accept-
able corrupt packets, packet overflow, packet underflow,
packet checksums, and packet latency;
•RF issues related to the position of equipment and multi-
path reflections;
•the systems required to instrument, monitor and provide
reports on the wireless communication channels;
•separation of communications between OT and IT for
operational and cybersecurity considerations.
The design and implementation of factory 5G communi-
cations systems must consider the day-to-day operation and
ongoing maintenance. Just as wired networks are monitored
for issues, wireless "situational awareness" is important.
This is done through the extension of existing network and
plant monitoring systems to include support for advanced
wireless communications, alternatively, the implementation
of separate monitoring dashboards with the ability to feed
IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX 11
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
Daniel S. Fowler et al.: A 5G Automated Guided Vehicle SME Testbed for Resilient Future Factories
into existing systems if necessary, as with WMG’s OPC UA
server dashboards. Physical security and cybersecurity are
additional considerations.
The physical security of communications hardware is
likely to fall under the security policies of other plant and
factory systems. The wireless communications equipment
will require protection from physical tampering and potential
vandalism. This requires consideration of equipment posi-
tion, physical protection mechanisms, and access control
policies.
Cybersecurity and the split between OT and IT systems,
and their sub-systems, have system design implications. Sep-
aration of communications channels is a likely requirement
to support cybersecurity and system design goals. There is a
variety of ways to achieve separate communication channels.
Separate physical hardware is one option. Another option
is to have virtual communications channels on an existing
network, using network virtualisation or a virtual private
network. 5G uses network slicing to separate applications.
Operations can dedicate a slice of the spectrum and network
services to applications to guarantee QoS requirements.
In the design and maintenance phases management of the
communications system assets is a consideration. For large
installations, there are potentially hundreds or thousands of
individual pieces of communication equipment. The unique
International Mobile Subscriber Identity (IMSI) number for
the endpoints will require management systems, along with
the associated Subscriber Identity Module (SIM) cards, or the
"embedded" or "electronic" version (eSIM). The requirement
for SIM programming or provision by a service or equipment
supplier is a consideration. Industrial locations that require
many endpoints may rely on system suppliers for SIM man-
agement. Security issues related to IMSI numbers and SIMs
(the cloning and stealing of SIM cards) have been reported in
the past, thus IMSI and SIM security is a consideration.
B. CHALLENGES WITHIN SMART MANUFACTURING
It could be argued that the implementation of advanced
wireless communications within manufacturing facilities is
another example of the industrial domain deploying new
technology as it emerges. Manufacturers have often taken
new technology and applied it to the shop floor to aid
productivity. Technology advances allowed industrialists to
develop the PLC, the production line robot, Computer-Aided
Design (CAD), and Computer-Aided Manufacturing (CAM),
through a host of emerging new digital manufacturing tech-
nologies (see the list in section II-B). However, manufactur-
ers must take credit for accelerating or inventing advances in
some technologies. 5G is an example as the design includes
the capabilities to support the I4.0 vision. Yet in this rush to
new digital manufacturing and advanced wireless communi-
cations technology, some often unconsidered challenges are
present [12]. These challenges include:
•the increase in the heterogeneity of the systems, the
number of system elements and the number of manage-
ment points;
•the handling of the natural division between enterprise
IT and factory OT;
•the potential plethora of different data formats and de-
sign models between system elements;
•obtaining good quality data from the new systems and
sensors;
•the appropriate formatting and analysis of data;
•making data available, understandable, and useful;
•the integration and communication between disparate
systems;
•the need for commonality in machine-to-machine com-
munication and understanding;
•improving connectivity, which can be fragmented, be-
tween systems and sub-systems;
•the resilience to handle changes and absorb the impact
of external events;
•the consideration and catering for cybersecurity;
•efficiently handling the lifecycle of this digital manufac-
turing super-system and the myriad of its elements.
The resources of multi-million or multi-billion pound in-
dustrial companies will allow them to address these chal-
lenges in the deployment of digital-based smart manufac-
turing and 5G technology. However, resource-limited SMEs
may be looking at suppliers and other knowledge sources to
address implementation challenges. This 5G AGV project is
part of the process of building and disseminating relevant
knowledge, particularly the role 5G can play in being an
enabler in overcoming some of the challenges raised.
Other ways to build the requisite knowledge include the
implementation of standards for I4.0 technologies. Standards
can help alleviate some of the challenges, e.g., around data
formatting and diverse system integration. The standards on
I4.0 would be available to manufacturing system designers
to help overcome any barriers to smart manufacturing imple-
mentation. Such I4.0 standards would build upon the existing
well-established manufacturing technology standards long
established through national and international trade bodies
and standards organisations. This means standards are not
always open and accessible, requiring membership of organ-
isations or access via paywalls.
There is the Reference Architecture Model Industry 4.0
(RAMI4.0) [32] that came from the I4.0 vision. RAM4.0
is a conceptual standard that recognises the importance of a
common description of all the manufacturing entities that can
exist in a smart factory. However, it is adding another layer
on top of the existing body of standards. Maybe it would have
been more productive to build an open software tool to help
engineers productively understand how to apply the existing
standards to smart manufacturing designs, systematically ad-
dressing the complexity involved in such highly data-driven
systems.
There is the potential for digital twinning techniques to
aid the management of the complexity within large-scale
smart manufacturing systems. DTs can ensure that changes
to CPPSs can be made that are beneficial to the system
throughout its life cycle [33]. To enable effective I4.0 systems
12 IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
Daniel S. Fowler et al.: A 5G Automated Guided Vehicle SME Testbed for Resilient Future Factories
and their DTs the supporting communication systems must
be capable of handling the data loads, latency, and response
times. This provides a case for 5G as it is designed to cater
for these use cases. However, some manufacturers, possibly
SMEs, may view 5G seen as too challenging for deploying
an industrial wireless solution. This is when they may look
to the familiarity of Wi-Fi, and the newer Wi-Fi 6 capabil-
ities to provide the communication link. Wi-Fi 6 boosts the
performance of the long-established Wi-Fi technology, but
similarly to 5G, the complete abilities of Wi-Fi 6 within an
industrial setting need to be fully understood.
C. ASPECTS OF MANUFACTURING RESILIENCE
The manufacturing sector was affected by global events from
2019 onward that have reawakened interest in industrial re-
silience [34]. Events that tested industrial resilience included
the coronavirus pandemic, the Suez Canal blockage, a semi-
conductors shortage, cyber-attacks, and political upheaval.
Industrial resilience can be examined at three levels, from
the national or macro level to the local or meso [35] level,
and down to the factory or micro level, see Table 4.
Resilience at the macro level is often associated with
ensuring the continued operation of Critical National Infras-
tructure (CNI) and other societal systems. This is driven by
government policies and national and international organ-
isations. Protection of CNI and similar critically required
systems is guided by Resilience Engineering (RE) [36]. Re-
silience at the meso level is associated with regional policies
and local systems, influenced by both the macro issues and
the lowest level micro issues. It is at the micro resilience level
that factories, organisations, and individuals have the most
direct influence on resilience factors.
At the micro level resilience is often addressed under the
guise of Business Continuity Management Systems (BCMS),
and operational procedures and processes which include
monitoring, testing, maintenance (including predictive main-
tenance), the use of KPIs, health and safety procedures, phys-
ical site security, and cybersecurity. Therefore, organisations
may not be consciously performing RE, but they are aware
of potential threats to day-to-day operations. However, RE
looks beyond day-to-day operations to enable systems to
handle more extreme unexpected disruptions and threats, see
Table 5.
Smart manufacturing technologies are used to improve
production efficiency, reduce costs and improve productivity.
If the deployment of digital factory technology is further
aimed at increasing system resilience, a desirable trait, then
engineers should incorporate aspects of resilience within
system designs. However, this requires them to understand
the concept of resilience and how systems can contribute
to it. RE at the micro level is not yet widely studied with
operations focused around a BCMS for day-to-day risk
management. Furthermore, risks and resilience aspects are
mainly addressed qualitatively.
The 5G AGV platform aims to use the data it gener-
ates to investigate quantitative methods of measuring system
FIGURE 14. Desired system traits that contribute to resilience [34].
resilience, particularly around the role advanced wireless
communications have to play in improving manufacturing
systems resilience. The study of resilience for the test plat-
form is likely to be concerned with micro-level issues. Yet,
meso-level issues are of interest to the sector. For example,
the availability of supply chain data to anticipate potential
disruptions and therefore production delays, triggering the
automatic reassessment of production scheduling and its
impact on machine utilisation.
Improving the granularity of supply chain data is one
vision for 5G’s usefulness to manufacturing. The study of
supply chain resilience [37] builds upon many years of study
into green supply chains [38], [39] and forms part of manu-
facturers’ sustainability aims. Alongside green and resilient
the lean and agile paradigms have been used to study supply
chains [40]. These various paradigms combine to produce
Lean, Agile, Resilient, and Green (LARG) research [41],
[42]. Achieving LARG goals requires timely data distributed
over fast and efficient communication networks linked to
sensor systems, i.e., the requirements of the I4.0 vision [43].
Engineers that do address resilience will find that it is
multi-faceted and is achieved by ensuring systems possess
several traits [34], see Figure 14. A system design or or-
ganisation taking steps to achieve several of these traits will
raise the level of resilience. However, whilst using tech-
nology to implement these traits, technology could detract
from resilience traits if it is misconceived or incorrectly
designed and deployed. Section IV touched upon some of
the overlooked challenges associated with introducing smart
manufacturing technologies. For example, introducing new
digital technology to improve a factory’s adaptability and
flexibility could detract from reliability and security due to
increases in system complexity. System designers aware of
these possible trade-offs can mitigate against them in their
system designs.
The combined effort of engineers implementing resilience
at the micro level will contribute to industrial resilience at the
meso and macro levels. Similarly, policies and initiatives at
the meso and macro levels will affect individual organisations
and factories. An increase in the overall industrial resilience
of an economy is seen as a beneficial goal for nations and
IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX 13
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
Daniel S. Fowler et al.: A 5G Automated Guided Vehicle SME Testbed for Resilient Future Factories
TABLE 4. Macro, Meso, and Micro resilience levels, from [34]
Level Coverage Applicability Examples
Macro National and multinational Governments and large organisations with
considerations covering critical infrastructure
over multiple sites
Networks of power, communications and transportation,
supply chains and logistics, social structures, law and
order, health systems, financial systems, labour markets
Meso Limited geographical areas and buildings Local facilities and branches of organisations,
medium and small enterprises Business and industrial parks, factories and offices, pub-
lic facilities and spaces
Micro Building internals and worker groups Internal systems and subsystems, operational
systems, individuals Plant and machinery, equipment, processes and proce-
dures
TABLE 5. Disruptors impacting an organisation’s resilience, from [34]
Level Disruptors
Macro Natural disasters (e.g., volcanoes or earthquakes)
Disease outbreak (epidemic or pandemic)
Extreme or severe weather events
Market forces and new startups
Changes in laws, regulations, and standards
Changes in societal behaviour
Technology obsolescence
Meso Local power outage
Animal or insect plagues
Riots, protests and activism
Trade unionism or strike action
Raw material and parts shortages
Micro Malfunctions
Equipment fires
Security breaches (e.g. theft or burglary)
Deviation from procedures
Cyber-attack
Disgruntled, badly behaved, or ill employee
aids the related objectives of environmental sustainability and
human well-being [8], [44].
D. 5G AS A FUTURE RESILIENCE AID
The 5G AGV testbed will be used to research the subject of
resilience within smart manufacturing systems. In particular,
how cellular 5G can be deployed to increase the various
resilience traits, building upon the general aim to research
5G’s use in building more capable and flexible manufacturing
systems. Indeed, 5G will have a role to play in addressing
some of the challenges described in Section IV-B, particu-
larly around its ability to improve connectivity options, aid
mobility and wireless options for equipment, improve com-
munication latency compared to older wireless technology,
and improve wireless data bandwidth. Future examples of 5G
communications and other smart manufacturing technologies
aiding improvements in system resilience include:
•for some manufacturing scenarios, improvements in
system flexibility, with equipment and manufacturing
cells supporting more than one task and being able to
relocate automatically between workstations or produc-
tion lines;
•improved worker flexibility, for example, workers sup-
porting operations from home or for more than one site
from a single location;
•reduction in reliance on human operators for equipment
and the associated cost savings, e.g., internal factory
logistics using autonomous vehicles to move parts re-
ducing the number of required forklift operators;
•anticipation of materials and parts supply issues across a
supply chain and using the data to enable diversion and
re-balancing of the supply chain;
•the monitoring of human shop floor activity, work, and
assembly to identify mistakes earlier requires improved
communications to handle the vision and AI systems
data and response demands;
•improving data collection, analysis, and at-the-edge
analysis to get an increase in the granular detail on how
a factory is functioning and using resources to aid in
spotting areas for productivity improvements and cost
reductions;
•keeping DTs updated and relevant in real-time to pro-
vide management with improved operational data and
forecasts, and to support the ability to run multiple what-
if scenarios.
The 5G AGV use case developed in this work will enable
further research in 5G-connected AGVs in the following
areas:
•single and group AGV intelligence;
•AI (machine learning) augmented AGVs to enable
smarter AGVs beyond autonomous logistics and make
the AGV part of a mobile and safe cobotic system;
•group intelligence and AGV sensing for dynamic and
efficient routing in complex operations;
•centralised control and overview of large fleets of
AGVs;
•inter-AGV communication via the 5G network for co-
operative operations;
•and teleoperation of AGVs;
The use of smart manufacturing technologies, 5G commu-
nications, and the 5G-enabled AGV can aid the addition of
the various resilience traits, shown in Figure 14, to a factory
production process. However, as stated earlier, the challenge
is not to assume the application of smart technologies will
automatically improve resilience. Instead, systems designers,
who are aware of the concept of resilience, are proactive in
considering which traits can be improved without a negative
overall impact. For example, a mobile cobot may be useful to
support more than one process or workstation, but it needs to
be robust enough to support multiple scenarios. If utilisation
is not correct, one location could be starved of the cobot’s
time, detracting from overall efficiency and resilience.
The application of 5G within factories will enable and
require new monitoring systems. The production process
14 IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
Daniel S. Fowler et al.: A 5G Automated Guided Vehicle SME Testbed for Resilient Future Factories
FIGURE 15. Using sensors to derive additional value from manufacturing
data. Collecting additional data beyond traditional production figures can lead
to new manufacturing insights, address threat and sustainability issues and
improve overall resilience.
relies on KPIs to ensure manufacturing output and quality is
being maintained. The world of smart manufacturing will be
adding new monitoring points, KPIs, additional cybersecurity
requirements, and further interactions between IT, OT, and
supply chain systems. Sustainability and green agenda goals
are likely to see research and investment in granular moni-
toring of power consumption, material, and parts usage, see
Figure 15. 5G communications and sensors will be playing
their parts in providing data to industrial information man-
agement systems. The mobility and cybersecurity use cases
for enhanced monitoring are likely to see initial deployments.
One area that the research platform is targeted to improve
is the dissemination of knowledge. 5G and all new smart
manufacturing technologies will need to be understood by
future automation engineers, system designers, and factory
technicians. The requisite skills have been highlighted as
a barrier to smart manufacturing adoption [45]. Education
and training on digital manufacturing technologies will be
a requirement for all types of industries. The 5G AGV will
be a learning platform for advanced industrial communica-
tion technologies and future 5G-based smart manufacturing
solutions. Furthermore, new technology should be intuitive
to use and if complexity is evident in implementing 5G
manufacturing solutions it will be highlighted.
V. CONCLUSION
The UK Government sees 5G as a key contributor to a
modern economy, stating: "Our evidence is clear that the
most significant economic benefits from 5G will come from
widespread adoption of advanced 5G by industrial sectors,
including manufacturing and logistics, and by public ser-
vices." [26]. The incentive for this research was, and is, to
enable SMEs to benefit from this vision.
The authors’ aim was to implement a manufacturing-based
system that can be used to perform a variety of tasks with
5G (and Wi-Fi 6) as the communications technology. It will
allow SMEs, who may lack the resources to experiment with
5G communications, to see how 5G could be applied within
their organisations. It has been achieved by augmenting an
FIGURE 16. The 5G AGV testbed is operational and is being extended to
other types of industrial mobile robots and manufacturing cells.
AGV with advanced wireless capability as a flexible and
reusable testbed for mobile and adaptable manufacturing
systems. The testbed is in use within WMG’s industrial
laboratories, see Figure 16, and the 5G communications can
be applied to other types of AMRs, e.g., WMG’s walking
robot.
The testbed is being used to research the application of
5G’s designed-in capabilities (URLLC, mMTC) to industrial
environments and factory laboratories, i.e., not simply as
a replacement for wired Ethernet connections, or the com-
monly used Wi-Fi. Existing networking technologies (wired
or wireless) provide sufficient bandwidth for typical mon-
itoring and configuration requirements. It will be new use
cases, see Table 3, that will benefit from advances in wireless
communications provided by 5G and Wi-Fi 6E.
Whilst we have seen examples of 5G AGV experiments
that required low latency capability (see Section III-A), there
is a lack of time-sensitive applications in the industrial 5G
space. This was noted in [46], where they identified the need
for more time-sensitive networking applications utilising 5G
as a data link. This provides another research opportunity
for the testbed. Furthermore, the research will investigate in
detail 5G performance and RF issues in industrial locations,
including additional performance comparisons between 5G
and the latest versions of Wi-Fi (6/6E and the upcoming Wi-
Fi 7).
Advanced wireless communication is only part of the
overall picture of the digital technologies that are and will
be part of the future factory, see Figure 1. The testbed will
aid WMG and SMEs understand if 5G and advanced Wi-Fi
will meet the claims made for them for many of the I4.0 vi-
sions. However, the authors found new smart manufacturing
technologies come with challenges that may be overlooked
in the literature and that need consideration. The discussion
of those challenges in this work was to raise their awareness
and ensure automation designers consider those aspects.
IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX 15
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
Daniel S. Fowler et al.: A 5G Automated Guided Vehicle SME Testbed for Resilient Future Factories
Manufacturing technologies are no longer viewed in terms
of only internal factory systems. There are issues outside of
the day-to-day factory operation that needs awareness, not
only to ensure the resilience of operations but to satisfy wider
society’s resilience. A simple example on sustainability was
noted in Section III-A, where a finer control of an AGV can
reduce power requirements. Another example can be pro-
vided by the development of our testbed, which experienced
delays due to issues with supply chains. How 5G can be
used to gain insights into supply chains is another relevant
area in research. SMEs using the 5G testbed to address smart
manufacturing challenges will contribute to the operation of
resilient future factories.
REFERENCES
[1] HM Government, “Industrial Strategy,” Department for Business, Energy
& Industrial Strategy, Tech. Rep., 2017.
[2] S. Wei, Y. Ma, R. Li, and L. Hu, “Toward smart manufacturing: Key
technologies and trends driving standardization,” IEEE Computer, vol. 53,
DOI 10.1109/MC.2020.2970821, no. 4, pp. 46–50, 2020.
[3] 3GPP Organizational Partners, “3GPP TR 21.916 V16.1.0 Summary of
Rel-16 Work Items (Release 16),” 3rd Generation Partnership Project,
Sophia Antipolis, Tech. Rep., 2022.
[4] X. Wang and L. Gao, When 5G meets industry 4.0. Singapore: Springer,
2020.
[5] E. O’Connell, D. Moore, and T. Newe, “Challenges associated with
implementing 5g in manufacturing,” Telecom, vol. 1, DOI 10.3390/tele-
com1010005, no. 1, p. 48–67, Jun. 2020.
[6] J. Cheng, Y. Yang, X. Zou, and Y. Zuo, “5g in manufacturing: a litera-
ture review and future research,” The International Journal of Advanced
Manufacturing Technology, DOI 10.1007/s00170-022-08990-y, 2022.
[7] P. Skarin, W. Tärneberg, K.-E. Årzen, and M. Kihl, “Towards mission-
critical control at the edge and over 5g,” in 2018 IEEE International Con-
ference on Edge Computing (EDGE), DOI 10.1109/EDGE.2018.00014,
pp. 50–57, 2018.
[8] United Nations, “Sustainable Development Goals.” [Online]. Available:
https://sdgs.un.org/goals
[9] E. A. Oyekanlu, A. C. Smith, W. P. Thomas, G. Mulroy, D. Hitesh,
M. Ramsey, D. J. Kuhn, J. D. Mcghinnis, S. C. Buonavita, N. A. Looper,
M. Ng, A. Ng’oma, W. Liu, P. G. Mcbride, M. G. Shultz, C. Cerasi,
and D. Sun, “A review of recent advances in automated guided vehicle
technologies: Integration challenges and research areas for 5g-based smart
manufacturing applications,” IEEE Access, vol. 8, DOI 10.1109/AC-
CESS.2020.3035729, pp. 202 312–202353, 2020.
[10] H. Kagermann, W. Wahlster, and J. Helbig, “Securing the future of
German manufacturing industry: Recommendations for implementing the
strategic initiative INDUSTRIE 4.0,” National Academy of Science and
Engineering, Tech. Rep., 2013.
[11] Acatech, “Cyber-Physical Systems, Driving force for innovation in mo-
bility, health, energy and production,” National Academy of Science and
Engineering, Berlin, Heidelberg, Tech. Rep., 2011.
[12] R. Harrison, D. A. Vera, and B. Ahmad, “A Connective Framework
to Support the Lifecycle of Cyber–Physical Production Systems,” Proc.
IEEE, vol. 109, DOI 10.1109/JPROC.2020.3046525, no. 4, pp. 568–581,
Apr. 2021.
[13] A. Kusiak, “Smart manufacturing,” International Journal of Production
Research, vol. 56, DOI 10.1080/00207543.2017.1351644, no. 1-2, pp.
508–517, 2018.
[14] I. T. Union, “Imt vision - framework and overall objectives of the future
development of imt for 2020 and beyond,” ITU, Geneva, Tech. Rep., 2015.
[15] E. Dahlman, S. Parkvall, and J. Skold, 5G NR: The Next Generation
Wireless Access Technology, 2nd ed. Elsevier Science, 2020.
[16] 3rd Generation Partnership Project, “TR 38.801 Study on new radio access
technology: Radio access architecture and interfaces (Release 14),” 3GPP,
Sophia Antipolis, Tech. Rep., 2017.
[17] G. Liu, Y. Huang, Z. Chen, L. Liu, Q. Wang, and N. Li, “5g deploy-
ment: Standalone vs. non-standalone from the operator perspective,” IEEE
Communications Magazine, vol. 58, DOI 10.1109/MCOM.001.2000230,
no. 11, pp. 83–89, 2020.
[18] J.-S. Bedo, E. C. Strinati, S. Castellvi, T. Cherif, V. Frascolla, W. Haerick,
I. Korthals, O. Lazaro, E. Sutedjo, L. Usatorre, and M. Wollschlaeger,
“5G and the Factories of the Future,” 5G Infrastructure Public Private
Partnership, Tech. Rep., 2015.
[19] K. Yaovaja, P. Bamrungthai, and P. Ketsarapong, “Design of an au-
tonomous tracked mower robot using vision-based remote control,” in
2019 IEEE Eurasia Conference on IOT, Communication and Engineering
(ECICE), DOI 10.1109/ECICE47484.2019.8942741, pp. 324–327, 2019.
[20] W. Nakimuli, J. Garcia-Reinoso, J. E. Sierra-Garcia, P. Serrano, and
I. Q. Fernández, “Deployment and evaluation of an industry 4.0
use case over 5g,” IEEE Communications Magazine, vol. 59, DOI
10.1109/MCOM.001.2001104, no. 7, pp. 14–20, 2021.
[21] S. Vakaruk, J. E. Sierra-García, A. Mozo, and A. Pastor, “Forecasting
automated guided vehicle malfunctioning with deep learning in a 5g-based
industry 4.0 scenario,” IEEE Communications Magazine, vol. 59, DOI
10.1109/MCOM.221.2001079, no. 11, pp. 102–108, 2021.
[22] X. Fu, D. Wang, J. Hu, J. Wei, and C.-B. Yan, “Leader-follower based two-
agv cooperative transportation system in 5g environment,” in 2022 IEEE
18th International Conference on Automation Science and Engineering
(CASE), DOI 10.1109/CASE49997.2022.9926664, pp. 67–72, 2022.
[23] L. Wang, Q. Liu, C. Zang, S. Zhu, C. Gan, and Y. Liu, “Formation
control of dual auto guided vehicles based on compensation method in
5g networks,” Machines, vol. 9, DOI 10.3390/machines9120318, no. 12,
Nov. 2021.
[24] C. Di Martino and A. Walid, “Continuous testing and sla management
of 5g networks for industrial automation,” in 2021 IEEE International
Symposium on Software Reliability Engineering Workshops (ISSREW),
DOI 10.1109/ISSREW53611.2021.00105, 2021.
[25] B. B. Cebecioglu, Y. K. Mo, S. Dinh-Van, D. S. Fowler, A. Evans,
A. Sivanathan, E. Kampert, B. Ahmad, and M. D. Higgins, “Sub-6
ghz channel modeling and evaluation in indoor industrial environments,”
IEEE Access, vol. 10, DOI 10.1109/ACCESS.2022.3227052, pp. 127 742–
127 753, 2022.
[26] Department for Science, Innovation and Technology, “Uk wireless infras-
tructure strategy,” 2023.
[27] J. Hu, Y. Liao, L. Song, and Z. Han, “Fairness-throughput tradeoff in full-
duplex wifi networks,” in 2016 IEEE Global Communications Conference
(GLOBECOM), DOI 10.1109/GLOCOM.2016.7841837, pp. 1–6, 2016.
[28] V. Yazıcı, U. C. Kozat, and M. O. Sunay, “A new control plane for 5g
network architecture with a case study on unified handoff, mobility, and
routing management,” IEEE Communications Magazine, vol. 52, DOI
10.1109/MCOM.2014.6957146, no. 11, pp. 76–85, 2014.
[29] L. Song, Y. Liao, K. Bian, L. Song, and Z. Han, “Cross-layer protocol
design for csma/cd in full-duplex wifi networks,” IEEE Communications
Letters, vol. 20, DOI 10.1109/LCOMM.2016.2519518, no. 4, pp. 792–
795, 2016.
[30] M. Hoppari, M. Uitto, J. Mäkelä, I. Harjula, and S. Rantala, “Performance
of the 5th generation indoor wireless technologies-empirical study,” Future
Internet, vol. 13, DOI 10.3390/fi13070180, no. 7, Jul. 2021.
[31] J. Ansari, C. Andersson, P. de Bruin, J. Farkas, L. Grosjean, J. Sachs,
J. Torsner, B. Varga, D. Harutyunyan, N. König, and R. H. Schmitt,
“Performance of 5g trials for industrial automation,” Electronics, vol. 11,
DOI 10.3390/electronics11030412, no. 3, Jan. 2022.
[32] IEC, “Smart manufacturing - Reference architecture model industry 4.0
(RAMI4.0),” 2017.
[33] Technical Committee ISO/TC 184, “ISO 23247 Automation systems and
integration - Digital Twin framework for manufacturing,” Geneva, 2021.
[34] D. S. Fowler, G. Epiphaniou, M. D. Higgins, and C. Maple, “Aspects of
resilience for smart manufacturing systems,” Strategic Change: Briefings
in Entrepreneurial Finance, 2023, forthcoming.
[35] C. Baumann, M. Cherry, and W. Chu, “Competitive Productivity (CP)
at macro–meso–micro levels,” Cross Cultural & Strategic Management,
vol. 26, DOI 10.1108/CCSM-08-2018-0118, no. 2, pp. 118–144, 1 2019.
[36] K. Thoma, “Resilien-tech. resilience by design: A strategy for the tech-
nology issues of the future,” acatech - National Academy of Science and
Engineering, 2014.
[37] R. Maharjan and H. Kato, “Resilient supply chain network de-
sign: a systematic literature review,” Transport Reviews, vol. 0, DOI
10.1080/01441647.2022.2080773, no. 0, pp. 1–23, 2022.
[38] S. K. Srivastava, “Green supply-chain management: A state-of-the-art
literature review,” International Journal of Management Reviews, vol. 9,
DOI 10.1111/j.1468-2370.2007.00202.x, no. 1, pp. 53–80, 2007.
[39] B. Fahimnia, J. Sarkis, and H. Davarzani, “Green supply chain man-
agement: A review and bibliometric analysis,” International Journal of
16 IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
Daniel S. Fowler et al.: A 5G Automated Guided Vehicle SME Testbed for Resilient Future Factories
Production Economics, vol. 162, DOI doi.org/10.1016/j.ijpe.2015.01.003,
pp. 101–114, 2015.
[40] D. Oliveira-Dias, J. Moyano-Fuentes, and J. M. Maqueira-Marín, “Under-
standing the relationships between information technology and lean and
agile supply chain strategies: a systematic literature review,” Annals of
Operations Research, DOI 10.1007/s10479-022-04520-x, pp. 973–1005,
2022.
[41] M. do Rosário Cabrita, S. Duarte, H. Carvalho, and V. Cruz-Machado,
“Integration of lean, agile, resilient and green paradigms in a business
model perspective: Theoretical foundations,” IFAC-PapersOnLine, vol. 49,
DOI 10.1016/j.ifacol.2016.07.704, no. 12, pp. 1306–1311, 2016, 8th IFAC
Conference on Manufacturing Modelling, Management and Control MIM
2016.
[42] V. Sharma, R. D. Raut, S. K. Mangla, B. E. Narkhede, S. Luthra, and
R. Gokhale, “A systematic literature review to integrate lean, agile, re-
silient, green and sustainable paradigms in the supply chain management,”
Business Strategy and the Environment, vol. 30, DOI 10.1002/bse.2679,
no. 2, pp. 1191–1212, 2021.
[43] M. S. Amjad, M. Z. Rafique, S. Hussain, and M. A. Khan, “A
new vision of larg manufacturing — a trail towards industry 4.0,”
CIRP Journal of Manufacturing Science and Technology, vol. 31, DOI
10.1016/j.cirpj.2020.06.012, pp. 377–393, 2020.
[44] European Commission Directorate-General for Research and Innovation,
“Industry 5.0 : Towards a sustainable, human-centric and resilient Euro-
pean industry,” European Union, Luxembourg, Tech. Rep., 2021.
[45] F. Azmat, B. Ahmed, W. Colombo, and R. Harrison, “Closing the
skills gap in the era of industrial digitalisation,” in 2020 IEEE Con-
ference on Industrial Cyberphysical Systems (ICPS), vol. 1, DOI
10.1109/ICPS48405.2020.9274788, pp. 365–370, 2020.
[46] Z. Satka, M. Ashjaei, H. Fotouhi, M. Daneshtalab, M. Sjödin, and
S. Mubeen, “A comprehensive systematic review of integration of time
sensitive networking and 5g communication,” Journal of Systems Archi-
tecture, vol. 138, DOI 10.1016/j.sysarc.2023.102852, 2023.
DANIEL S. FOWLER has a B.Eng. (Hons.) in
computer and control system engineering, an M.S.
in forensic computing, and a Ph.D. in automotive
cybersecurity, all degrees from Coventry Univer-
sity, UK. He is a Research Fellow in the Secure
Cyber Systems Research Group at WMG in the
field of secure and resilient system design. He
has aided the engineering and delivery of several
Innovate UK-funded projects. He contributes to
the open source community through software and
has written over 300 articles. He is a Chartered Engineer, and a member of
the IET and ACM.
YUEN KWAN MO has a Ph.D. in communica-
tions and network engineering from The Univer-
sity of Warwick, Coventry, UK. He is a Project
Engineer with the Connectivity and Communi-
cations Technology Research Group, WMG, The
University of Warwick. His specialisms include:
5G and cellular communications for Industry 4.0,
connected and autonomous vehicles, millimeter-
wave communications, massive MIMO, precoding
techniques and optimization algorithms.
ALEX EVANS joined WMG as a Project Engineer
in 2017 and is now a Lead Engineer. During this
time he has helped to deliver a number of indus-
try collaborative advanced technology automation,
digital manufacturing and data collection projects.
Having received a BSc in Cybernetics and Control
Engineering from the University of Reading in
1993, the early years of his career were spent
designing, building and programming embedded
microcontroller projects for a variety of customers
in scientific and automation industries. A further 10+ years working for a
leading HMI manufacturer and 9 years as a self-employed HMI and PLC
software engineer enabled him to gain experience with all the leading PLC
technologies, across a variety of sectors including automotive, textiles, food,
utilities, building automation and energy monitoring.
SON DINH-VAN received a B.S. degree from
Hanoi University of Science and Technology,
Vietnam, in 2013, the M.S. degree from Soongsil
University, Seoul, South Korea, in 2015, and the
Ph.D. degree from Queen’s University of Belfast,
Belfast, UK, in 2019, all in electrical engineering.
He is currently a Research Fellow with the Con-
nectivity and Communications Technology Re-
search Group, WMG, The University of Warwick,
UK. He was a Data Scientist with Frequenz Gmbh,
Germany, and a Visiting Researcher with Middlesex University, London,
UK, in 2020 and 2021. His current research interests include 5G-and-beyond
communications for manufacturing, wireless security, millimeter-wave, and
machine learning.
BILAL AHMAD (Senior Member, IEEE) received
the M.Sc. degree in mechatronics and the Ph.D.
degree in automation systems from Loughborough
University, Loughborough, UK, in 2007 and 2014,
respectively. He was a Research Fellow and a
Senior Research Fellow with WMG, The Univer-
sity of Warwick, Coventry, UK, and a Research
Associate at the Wolfson School of Mechanical
and Manufacturing Engineering, Loughborough
University. He is currently an Associate Professor
with the WMG, The University of Warwick. He has worked on a range
of high-profile national and international research projects. He is named
as PI and Co-I on a number of Innovate UK and HVM Catapult projects.
He has published his research findings in over 50 peer-reviewed journal
articles and conference papers. His research interests are in the areas of
manufacturing digitalization and lifecycle engineering of cyber–physical
production systems, especially focusing on the design and deployment of
real-time control systems and their connectivity with IT systems.
IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX 17
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
Daniel S. Fowler et al.: A 5G Automated Guided Vehicle SME Testbed for Resilient Future Factories
MATTHEW D. HIGGINS (Senior Member,
IEEE) is a Reader at the University of Warwick,
where he leads WMG’s Connectivity and Com-
munications Technology Research Group within
its Intelligent Vehicles Directorate. His research
interests span 5G and Beyond, Core Networking,
IEEE 802.3xx, GNSS, and Timing, with applica-
tions to both the Automotive and Manufacturing
domains. Coupled with an overarching motivation
to ensure ongoing resilience of the domain is
considered, Matthew leads many high-value collaborative projects funded
through EPSRC, Innovate UK, and HVMC, as well as also leading multiple
projects funded directly by industry.
CARSTEN MAPLE is a professor of cyber sys-
tems engineering at the Cyber Security Centre at
the University of Warwick, where he leads the
GCHQ-EPSRC Academic Centre of Excellence
in Cyber Security Research. He has published
over 200 peer-reviewed papers. He has provided
evidence and advice to governments and organi-
zations across the world, including being a high-
level scientific advisor for cyber security to the
European Commission. He is principal or coinves-
tigator on a number of projects in cyber security. He is Immediate Past Chair
of the Council of Professors and Heads of Computing in the UK and a Fellow
of the Alan Turing Institute.
18 IEEE OPEN JOURNAL OF THE INDUSTRIAL ELECTRONICS SOCIETY. VOLUME XX, 20XX
This article has been accepted for publication in IEEE Open Journal of the Industrial Electronics Society. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJIES.2023.3291234
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
