Content uploaded by João Carlos Gonçalves dos Reis
Author content
All content in this area was uploaded by João Carlos Gonçalves dos Reis on Jul 23, 2024
Content may be subject to copyright.
Article Not peer-reviewed version
Edge Intelligence in Enhancing Last-
Mile Delivery Logistics
João Reis *
Posted Date: 23 July 2024
doi: 10.20944/preprints202407.1777.v1
Keywords: artificial intelligence; autonomous delivery vehicles; customer satisfaction; Delphi method;
distribution centers; edge intelligence; last-mile delivery logistics; real-time decision-making; supply chain
management
Preprints.org is a free multidiscipline platform providing preprint service that
is dedicated to making early versions of research outputs permanently
available and citable. Preprints posted at Preprints.org appear in Web of
Science, Crossref, Google Scholar, Scilit, Europe PMC.
Copyright: This is an open access article distributed under the Creative Commons
Attribution License which permits unrestricted use, distribution, and reproduction in any
medium, provided the original work is properly cited.
Article
Edge Intelligence in Enhancing Last-Mile
Delivery Logistics
João Reis
joao.reis@ulusofona.pt
Abstract: Background: The last-mile delivery phase, the final stage where goods move from a
distribution center to customers, is pivotal but faces significant inefficiencies and high costs due to
its complexity. Recent advancements in Edge AI or Edge Intelligence (EI) offer promising solutions
to these challenges. Methods: This study explores how AI-driven technologies and real-time data
processing, combined with EI, can enhance last-mile delivery operations. A thorough literature
review was conducted to assess technological advancements and a Delphi method was used to
systematically and empirically assess the impact of EI solutions on both operational efficiency and
customer satisfaction. Results: Although EI technologies offer substantial benefits, EU companies are
hesitant to adopt these innovations due to high implementation costs. However, firms that have
embraced these technologies report significant improvements, including better route optimization,
reduced delivery times, and enhanced service reliability. These findings highlight the need for a
culture of innovation and the recruitment of experts with advanced qualifications to drive
technological advancement in last-mile logistics. Conclusions: The integration of EI represents a
significant step towards more efficient, cost-effective, and customer-focused last-mile delivery
solutions. Future research should aim to refine these technologies and explore their long-term
impacts on the logistics industry.
Keywords: artificial intelligence; autonomous delivery vehicles; customer satisfaction; Delphi
method; distribution centers; edge intelligence; last-mile delivery logistics; real-time decision-
making; supply chain management
1. Introduction
The last-mile delivery phase, the final stage in the supply chain where goods are transported
from a distribution center to the end customer [1,2], is a pivotal component of the logistics ecosystem
[3]. This phase is characterized by its complexity and operational challenges [4], historically
contributing to significant inefficiencies and high costs. However, recent advancements in Edge AI
or Edge Intelligence (EI) have demonstrated substantial potential for addressing these challenges and
optimizing last-mile logistics processes [5]. Moreover, as e-commerce grows, the challenges
associated with last-mile delivery—such as high costs, complex logistics, and increasing consumer
demands—are becoming even more pronounced [6]. Traditional delivery methods often struggle
with issues such as route inefficiency, high operational costs, and customer dissatisfaction [7].
Hence, in recent years, technological advancements have introduced innovative solutions aimed
at overcoming these challenges. Among these, EI has emerged as a promising tool to enhance last-
mile delivery logistics [8,9]. AI technologies, including machine learning algorithms [10,11],
predictive analytics [12], and autonomous systems [13], have demonstrated potential in optimizing
delivery routes and forecasting demand.
Concurrently, EI, which refers to the processing of data at or near the source of data generation,
offers real-time decision-making capabilities essential for dynamic last-mile operations [5]. In their
2019 study, Zhou et al. [5] highlights that EI is gaining significant interest despite being in its nascent
stages. The authors conducted a thorough survey of research efforts in EI, covering the history and
motivation behind running AI at the network edge. They provided a comprehensive overview of the
Disclaimer/Publisher’s Note: The statements, opinions, and data contained in all publications are solely those of the individual author(s) and
contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting
from any ideas, methods, instructions, or products referred to in the content.
Preprints.org (www.preprints.org) | NOT PEER-REVIEWED | Posted: 23 July 2024 doi:10.20944/preprints202407.1777.v1
© 2024 by the author(s). Distributed under a Creative Commons CC BY license.
2
architectures, frameworks, and emerging key technologies enabling deep learning models for
training and inference at the network edge. Additionally, they discussed potential future research
directions in EI. Building on Zhou et al.'s [5] seminal work, our contribution aims to expand on the
future research perspectives they identified. Specifically, we focus on issues related to EI ecosystems
and the close collaboration and integration among different service providers. This collaboration is
crucial for broad resource sharing and one-piece service transfer. For instance, in an EI service model,
a user can simultaneously be a service consumer and a data generator. This dual role necessitates a
new smart pricing scheme that accounts for both the user's service consumption and the value of
their data contribution.
To explore this phenomenon, the role of AI and EI in revolutionizing last-mile delivery logistics,
we started by studying the application of AI-driven technologies and real-time data processing. To
do so, we conducted a comprehensive literature review to evaluate current technological
advancements [14]. Moreover, we supplemented the literature review with the Delphi method to
systematically and empirically assess the impact of AI-driven solutions on operational efficiency and
customer satisfaction. Therefore, we employed a multi-method approach [15], integrating two
qualitative methods to complement each other. Specifically, the literature review will establish the
theoretical foundation, while the empirical analysis will serve to validate or challenge the initial
theoretical findings. To conduct this research, we formulated the following research question: How
can Edge Intelligence contribute to the development of more sustainable last-mile delivery solutions?
Preliminary results reveal that despite the recognized advantages of EI, EU companies remain
hesitant to fully embrace these technologies. This reluctance is primarily driven by concerns
regarding the high implementation and maintenance costs. This barrier is particularly pronounced
among SMEs, which often face greater challenges in adopting these innovations due to limited
financial resources and expertise. Nevertheless, companies that have successfully integrated these
technologies report notable improvements in operational outcomes, including enhanced route
optimization, reduced delivery times, and increased service reliability. These observations show the
necessity of fostering a culture of innovation and development, which includes recruiting experts
with advanced academic qualifications and a strong focus on technological advancement in the field.
This article follows the IMRaD structure [16], organizing its sections into Introduction, Methods,
Results, and Discussion. In the Introduction, we discuss the problem to be studied, define the research
question, and present preliminary results to provide the reader with the necessary background
knowledge. The Methods section offers a detailed description of the methodological process we
followed. In the Results section, we organize our findings into three primary subsections. First, we
provide a conceptualization of Edge Intelligence, addressing the diverse interpretations of the term
and defining it within the context of business and management, including logistics. Following this,
we review the literature to clarify and establish the key concepts crucial to our study. The final
subsection is dedicated to the empirical validation of the conceptual framework, where we analyze
how our data aligns with or challenges the proposed theories and end with a summary that integrates
the main findings and discusses their implications. In the Discussion, we highlight the originality of
our work, its contributions to existing theory, and the most significant managerial implications for
practitioners. The final section addresses the limitations of our study and suggests avenues for future
research.
2. Methods
This research employs a multi-method approach [15], utilizing two qualitative typology
methods. The first method is a systematic literature review (SLR) [17], leveraging two renowned
scientific databases: the EU Elsevier Scopus and the US Clarivate Web of Science (WoS) Core
Collection. These databases were chosen for their source-neutral, abstract, and citation collections,
curated by independent subject matter experts who are recognized leaders in their fields [18].
Academic search engines like Google Scholar were excluded due to their lack of guaranteed blind
peer review [19]. For our searches, conducted on July 7, 2024, we used the terms "Edge Intelligence"
or “Edge Artificial Intelligence” and "Logistics" in the manuscript Title, Abstract, and Keywords
Preprints.org (www.preprints.org) | NOT PEER-REVIEWED | Posted: 23 July 2024 doi:10.20944/preprints202407.1777.v1
3
(Topic in WoS). From a temporal perspective, this research commenced in September 2023. Before
submitting the article, we updated the entire bibliography to ensure the most current information. In
Scopus, we identified eleven relevant art icles in English. I n WoS, we found six relevant articles, which
overlapped with those found in Scopus, as detailed in Table 1.
Table 1. State-of-the-art of EI in Logistics.
Database Document
Type Document Title Authors Source Year
Scopus
WoS
Journal
Article
Edge intelligence empowered
delivery route planning for
handling changes in uncertain
supply chain environment
Peng et al.
[20]
Journal of Cloud
Computing 2024
Scopus
WoS
Journal
Article
Securing clustered edge
intelligence with blockchain
Dehury et
al. [21]
IEEE Consumer
Electronics Magazine 2022
Scopus
WoS
Journal
Article
KeepEdge: A knowledge
distillation empowered edge
intelligence framework for visual
assisted positioning in UAV
delivery
Luo et al.
[22]
IEEE Transactions on
Mobile Computing 2022
Scopus Conference
Paper
A Comparison of Temporal
Encoders for Neuromorphic
Keyword Spotting with Few
Neurons
Nilsson et
al. [23]
International Joint
Conference on Neural
Networks
2023
Scopus Conference
Paper
Research on Fast Adaptive
Transmission Models for
International Inland Port Based on
Edge Intelligence
Yiwen [24]
International
Conference on Cyber
Security and Cloud
Computing (CSCloud)
2023
Scopus Journal
Article
Effective methods based on
distinct learning principles for the
analysis of hyperspectral images
to detect black sigatoka disease
Ugarte
Fajardo et
al. [25]
Plants 2022
Scopus
WoS
Journal
Article
Allocation of applications to Fog
resources via semantic clustering
techniques: With scenarios from
intelligent transportation systems
Xhafa [26] Computing 2021
Scopus
WoS
Conference
Paper
An edge based federated learning
framework for person re-
identification in UAV delivery
service
Zhang et al.
[27]
IEEE International
Conference on Web
Services
2021
Scopus
WoS
Journal
Article
Edge computing in industrial
Internet of Things: Architecture,
advances and challenges
Qiu et al.
[28]
IEEE Communications
Surveys & Tutorials 2020
Scopus Journal
Article
Secure and privacy-preserving
automated machine learning
operations into end-to-end
integrated IoT-edge-artificial
intelligence-blockchain
monitoring system for diabetes
mellitus prediction
Hennebelle
et al. [29]
Computational and
Structural
Biotechnology Journal
2024
Scopus Conference
Paper
IoT-Empowered Drones: Smart
Cyber security Framework with
Machine Learning Perspective
Mahamkali
et al. [30]
International
Conference on New
Frontiers in
2023
Preprints.org (www.preprints.org) | NOT PEER-REVIEWED | Posted: 23 July 2024 doi:10.20944/preprints202407.1777.v1
4
Communication,
Automation,
Management and
Security
A preliminary analysis reveals the scarcity of articles on this topic in the two most widely used
and internationally recognized scientific databases. Particularly, the most recent article dates to 2020,
indicating that the subject is quite new. Among the identified manuscripts, 64% are scientific articles
and 36% are conference papers, suggesting a significant interest in journals and a need for more
extensive academic and scientific discussion on this topic. Geographically, the Popular Republic of
China (PRC) leads with five manuscripts, followed by Australia and India with three each, the USA
with two, and Austria with one. This distribution highlights the global interest and varying levels of
research activity in different regions. In terms of subject areas, computer science accounts for 39%,
engineering for 26%, and decision sciences for 9%. Interestingly, Business and Management comprise
only 4% of the identified manuscripts, which is far below our expectations and underlines the
relevance and necessity of our article in filling this gap. A qualitative analysis framework, specifically
the content analysis technique [31], was employed to evaluate the effectiveness of cutting-edge AI
and intelligence technologies. This framework involved categorizing these technologies based on
their applications in last-mile delivery.
The second method we used is the Delphi method. This method was employed primarily to
validate the technological advancements identified in the literature review. This method involved
consulting a pre-selected group of experts. Data collection was extensive, encompassing a sample of
various companies operating in a European Union (EU) country (Portugal) and multinationals from
diverse sectors such as the automotive industry, electrical grid, healthcare, multinational technology
conglomerates, and retail. It was also targeted, aiming to gather insights from highly specialized
individuals with expertise in both IT and management logistics. A total of 28 experts were invited to
participate, with 7 responding and agreeing to take part in the study. For more details, see Table 2.
Table 2. Elements of the Delphi Study.
ID Job Title Company Rounds
P1 IT Support Specialist Multinational Automotive Company
(Company A)
2
P2 Director of Logistics 2
P3 IT Director/IT Manager National Electric Grid Company
(Company B) 2
P4 Chief Technology Officer National Health Company
(Company C) 2
P5 Head of IT Multinational technology conglomerate
(Company D)
2
P6 Director of Logistics 2
P7 Director of
Operations/Logistics
Multinational Retailer
(Company E) 2
The main challenge in securing expert participation was their availability. To encourage
participation, we committed to sharing detailed information on the investigation's results via email,
extending beyond the findings presented in this article. For confidentiality reasons, only summarized
information is provided here, not the full interviews. All participants signed a declaration of
informed consent under the Helsinki Treaty. The study involved two rounds of consultation, during
which experts were encouraged to converge on a unified position while allowing for the
representation of minority views that diverge from mainstream opinions. Additional refinements
were made to the proposed statements between rounds to increase support. Moreover, face-to-face
meetings were occasionally convened to facilitate the consensus-building process. Despite the
collected insights, the study has limitations too, mostly regarding the generalization of results. A
comprehensive discussion of these limitations is provided in Section 4.3 of this article. We are
Preprints.org (www.preprints.org) | NOT PEER-REVIEWED | Posted: 23 July 2024 doi:10.20944/preprints202407.1777.v1
5
confident that combining the two methods (i.e., SLR and Delphi) enhances the robustness of the
results, as each method complements and reinforces the other. Therefore, employing a multi-method
approach was the most appropriate choice for this article.
The content analysis technique was also employed for qualitative research. Initially, we
organized all the collected data and conducted a thorough review to identify discrepancies between
various sources. To enhance the visibility of these discrepancies, we implemented a scoring system.
This system also aided in determining codes and thematic categories. Furthermore, NVivo 14
facilitated the process, particularly in the more advanced phases, by expediting the integration,
coding, and analysis of the dataset. The insights gathered from the content analysis are presented in
Tables 4 and 5.
3. Results
As previously described, this section is structured into three key parts. We first define EI in the
specific context of business and management. After establishing this definition, we move on to
investigate the impact of EI on Last-Mile Delivery Logistics, analyzing how EI can improve these
logistics processes and addressing areas of theoretical debate. The section wraps up with a summary
that brings together the main findings and implications of our analysis.
3.1. EI Broad Conceptualization
Before proceeding with the core analysis of the results, we recognized the importance of briefly
conceptualizing EI. To do so, we conducted a search on Scopus on July 7, 2024, using the terms "Edge
Intelligence" or "Edge Artificial Intelligence" in the title, abstract, and keywords of the manuscripts.
Overall, we observed an exponential growth in publications on EI from 2017 to 2024. The PRC leads
with 915 manuscripts, followed by the USA with 357 and India with 217. The PRC significantly differs
from the United States. Meanwhile, EU countries exhibit numbers similar to the US, though EU
research tends to focus on the national interests and specific priorities relevant to each member state.
Most of the identified manuscripts are in the field of computer science (41%), with significant
contributions from engineering (26%) and mathematics (7%). This indicates a strong focus on applied
sciences, with a relatively modest interest in the business and management sector (1%). Most
publications are journal articles (52%), while conference papers account for 35%. To conceptualize
the EI term, we used scientific articles from journals written in English, specifically from the
"Business, Management, and Accounting" domain, as this area is most relevant to our article. This
search yielded 14 scientific articles.
Table 3. EI conceptualization for business and management areas.
Author(s) Definition(s)
Sinha et al. [32]
“Edge Intelligence is a methodology where the prediction by the AI algorithm is
processed within the embedded processor connected to the actuator and sensors
of the device for faster response by the architecture” (p. 6)
Himeur et al. [33]
“Edge AI refers to the local processing of AI algorithms on edge device (…) Edge
AI brings processing and computational tasks closer to the point of interaction
with the end-user, whether that be a smartphone, single board computer (SBC),
domestic appliance, IoT device, or edge serve” (p. 2)
Da et al. [34]
“Fog computing (or Edge computing) is a paradigm that has recently been put
forward to provide real-time/low latency services and decrease the bandwidth
requirement. The fog nodes, extend the cloud to be closer to the edge by
enabling computations to be carried out at the sensors/devices that produce and
act on IoT data” (p. 210)
Amadeo et al. [35] “Edge computing allows caching and processing services directly at the edge of
the network, close to where data is produced and consumed” (p. 2)
Preprints.org (www.preprints.org) | NOT PEER-REVIEWED | Posted: 23 July 2024 doi:10.20944/preprints202407.1777.v1
6
Pradhan et al. [36]
“Edge computing (EC) is a distributed computing paradigm that brings
computing capabilities closer to the end-users and improves the quality of
service (QoS) and user experience”. (p. 1)
Alrashdi et al. [37]
“Edge intelligence has developed as a favorable paradigm to enable effective
and instantaneous processing of data at the network’s edge (…) Edge
Intelligence emerged as a decisive computational paradigm, dedicated to
redefining and reshaping the boundaries of data analytics as well as decision-
making” (pp.1-2)
Dalabehera et al.
[38]
“Fog computing, an innovative paradigm, extends the capabilities of cloud
computing to the edge of the network, bringing computing resources closer to
end-users” (p. 2)
Huang et al. [39]
"Intelligent edge has accelerated the Internet of Things (IoT) revolution towards
next-generation operational efficiency and massive connectivity (...) The
deployment of machine learning algorithms to the edge is made possible by
edge intelligence (EI), which integrates artificial intelligence (AI) and edge
computing technologies" (pp. 1-2)
In Table 3, we present eight definitions of EI, as not all 14 articles provided explicit definitions.
Our analysis revealed that not all authors use the terms "Edge Intelligence" [32,37,39] or "Edge
Artificial Intelligence" [33]; some prefer competing terms such as "Fog Computing" [34,38] and "Edge
Computing" [34–36]. We conducted a careful analysis of the four most relevant definitions of Edge
Intelligence [32,33,37,39] and developed an integrated definition tailored to the business and
management sectors:
Edge Intelligence (EI) is a transformative computational paradigm that integrates artificial
intelligence and edge computing technologies, such as machine learning algorithms, for the real-time
processing of data at the network's edge, closer to the point of interaction with the end-user. This
paradigm is redefining the boundaries of data analytics and decision-making.
In contrast, Edge Computing (EC) is a distributed computing paradigm that brings computing
capabilities closer to the end-users and improves QoS [36], enabling caching and processing services
directly at the edge of the network, near where data is produced and consumed [35]. While both EI
and EC enhance data processing efficiency and effectiveness by leveraging the network edge, EI
specifically incorporates AI to provide advanced, real-time analytical capabilities, whereas EC
focuses on localized data handling and processing to improve overall system performance. In
practical terms, EI is relevant in scenarios where real-time data analysis and decision-making are
critical [37]. Examples include autonomous vehicles, smart manufacturing, and predictive
maintenance, where instantaneous decisions can significantly impact performance and safety. The
objective of EI is to enhance decision-making capabilities at the edge, making systems more
responsive, intelligent, and autonomous. On the other hand, EC is used to improve the efficiency and
performance of a wide range of applications by minimizing the need to transmit data to distant cloud
servers for processing. Use cases include content delivery networks (CDNs), remote monitoring
systems, and IoT deployments where bandwidth and latency considerations are paramount. The
objective of EC is to reduce latency, increase processing efficiency, and improve the reliability of
services by decentralizing computational tasks.
Fog Computing (FC) is another paradigm that extends cloud computing capabilities to the edge
of the network. While it shares similarities with Edge Computing in bringing computing resources
closer to end users [38], FC also includes intermediary devices such as gateways and routers, which
provide additional layers of computation and storage between the cloud and the edge devices. FC
aims to provide real-time, low-latency services and reduce bandwidth requirements by distributing
computing tasks across various points in the network infrastructure [34]. This makes FC particularly
suitable for complex, large-scale IoT environments where data processing needs to be distributed
efficiently across different network layers. In summary, while EI, EC, and FC aim to enhance the
efficiency and responsiveness of data processing at the network's edge, they differ in their specific
Preprints.org (www.preprints.org) | NOT PEER-REVIEWED | Posted: 23 July 2024 doi:10.20944/preprints202407.1777.v1
7
focus and implementation. EI integrates AI for advanced real-time analytics, EC focuses on localized
processing to reduce latency and improve performance, and FC adds an intermediary layer to further
distribute computing resources and manage large-scale network demands.
3.2. EI in Enhancing Last-Mile Delivery Logistics
Out of the eleven articles selected, we extracted relevant information from seven, enabling us to
analyze EI technologies and their impacts on Last-Mile Delivery Logistics. From Table 4, it is evident
that the convergence of emerging technologies is revolutionizing last-mile delivery. This
transformation is marked by advancements in prediction [29], speed [27], and risk reduction [30]. By
analyzing the literature, we can see that technologies embedded with sensors [30], sophisticated
software, and algorithms [20], along with physical networks of objects such as UAVs [22,30], are
driving these changes. One of the major advances has been the integration of IoT, EC, AI, and
blockchain [29]. Together, these technologies enable highly efficient predictions that aid in real-time
decision-making. As a result, integrated systems combining these technologies and physical objects
have greatly enhanced precision and reliability in last-mile delivery [22,30].
Table 4. EI Technologies and their Impacts on the Last Mile – insights from the SLR.
Author Technology Impact on Last-Mile Delivery
Hennebelle et al.
[29]
IoT-edge-Artificial Intelligence
(AI)-blockchain system
“Diabetes prediction based on risk factors. The
results show that the proposed system predicts
diabetes using RF with 4.57% more accuracy on
average in comparison with the other models LR
and SVM, with 2.87 times more execution time” (p.
212)
Zhang et al. [27] Fed-UAV (Federal-Unmanned
Aerial Vehicle)
“Solve the person re-identification problem in the
UAV delivery service which is a typical AI
application in smart logistics (…) This framework
enables the UAV to efficiently locate the target
receivers, and effectively reduce the data
transmission between the UAV and the cloud
server to improve the response time and protect the
data privacy” (p. 500)
Qiu et al. [28] Edge computing in IIoT
(Industrial Internet of Things)
“Allows improved health management (PHM),
smart grids, manufacturing coordination,
intelligent connected vehicles (ICV), and smart
logistics” (p. 2462)
Mahamkali et al.
[30] Internet of Drone Things (IDT)
“Method that reduces the risk of cyber-attacks by
shoring up the foundation of a NoD (network of
drones) with cutting-edge artificial intelligence-
inspired approaches” (p. 1)
Luo et al. [22] KeepEdge
“UAV delivery is being increasingly used in the
field of logistics. It is highly challenging for a UAV
to precisely identify the position for parcel delivery
if it is only aided by the GPS. KeepEdge achieves
visual information-assisted positioning for the last
mile UAV delivery services” (p. 4729)
Peng et al. [20]
Mixed-integer programming
model & Cloud-edge
collaborative mode
“The cloud server comprehensively considers
customer demand and road condition changes and
employs adaptive genetic algorithms and A-star
algorithms to adjust the delivery routes
dynamically” (p. 1)
Preprints.org (www.preprints.org) | NOT PEER-REVIEWED | Posted: 23 July 2024 doi:10.20944/preprints202407.1777.v1
8
Dehury et al.
[21]
Blockchain-based solution for
Clustered Edge Intelligence
(CEI)
“CEI allows the devices to share their knowledge
and events with other devices and the remote fog
or cloud servers” (p. 22)
From the list of articles analyzed from the table above, PRC prioritizes EC in the IIoT [28] and
the use of UAVs in logistics and supply chain management [20,22]. Similarly, Australia stands out in
the areas of IoT-edge-artificial intelligence-blockchain integration [29] and UAV delivery services
[27]. One of the most significant studies in this domain is by Hennebelle et al. [29], which investigates
the capability of monitoring and predicting diabetes incidence using EI. The study proposes an
advanced IoT-edge-Artificial Intelligence (AI)-blockchain system designed to predict diabetes based
on identified risk factors. This innovative system employs blockchain technology to consolidate
patient risk factor data from multiple hospitals, thereby ensuring both comprehensive data
integration and the security and privacy of user information. Additionally, the study presents a
comparative analysis of various sensors, devices, and medical methodologies utilized to measure and
collect risk factor values within the proposed system. Another noteworthy study is by Qiu et al. [28],
which extensively examines the role of EC in the IIoT. The authors propose a forward-looking
architecture for IIoT, emphasizing the contributions of EC. They analyze the technical advancements
in routing, task scheduling, data storage and analysis, security, and standardization. Additionally,
the study explores the opportunities and challenges of integrating edge computing in IIoT,
particularly focusing on 5G-based edge communication, load balancing, data transfer, edge
intelligence, and data sharing security. The authors conclude by discussing several key application
scenarios for edge computing in IIoT, including prognostics and health management, smart grids,
manufacturing coordination, intelligent connected vehicles, and smart logistics. One final article we
would like to highlight for stimulating discussion is by Peng et al. [20], which presents an innovative
two-phase delivery route planning method incorporating advanced intelligence technology. The
distinctive feature of this approach is the use of EC devices to monitor real-time changes in road
conditions and dynamically adjust delivery routes accordingly. This method provides an effective
solution for improving efficiency and flexibility in logistics operations.
The convergence of emerging technologies, such as IoT, edge computing, AI and blockchain, has
been significantly influencing last-mile delivery. These technologies have been enabling efficient and
robust forecasts, better decision-making in real time, and also greater precision and reliability. As
demonstrated by the studies analyzed [20–22,27–30], these advances are paving the way for more
effective and adaptable logistics solutions in most sectors of activity.
3.3. Empirical Validation
In the Delphi Study, we analyzed emerging technologies identified in the literature and explored
their implications for the final phase of logistics. Due to the limited number of companies in a single
EU country, we were unable to comprehensively examine the entire technological spectrum
presented in Table 4 and are currently under investigation in the PRC and USA. Even if we had
attempted a broader analysis, it is unlikely we would have approached the level of advancements
seen in the PRC and USA. There are two main reasons for this. First, the volume of research in this
field within the EU is significantly lower compared to these two countries. Secondly, the EU strategy
might not necessarily follow the same path as that of the US and the PRC. This limitation represents
a challenge to the generalizability of our research findings. Nevertheless, our study concentrated on
several technologies that are actively transforming industrial and commercial processes in Portugal.
Our findings reveal a consensus among companies regarding the significant impact of the below
technologies on the logistics sector (see Table 5), which further justifies the relevance of the topic
addressed in this article. However, despite the literature review highlighting potential applications
for UAVs, our research did not uncover any specific industrial or commercial initiatives involving
UAVs, nor did we find any plans for their implementation within the five companies surveyed.
Preprints.org (www.preprints.org) | NOT PEER-REVIEWED | Posted: 23 July 2024 doi:10.20944/preprints202407.1777.v1
9
Table 5. EI Technologies and their Impacts on the Last Mile – insights from the Delphi Study.
ID–
Company Technologies Consensus Participant comments (Sample)
P1–A IoT-edge-AI-
blockchain 85%
"In my perspective, the IoT-edge-AI-blockchain system
can significantly enhance predictive capabilities and
runtime efficiency, thereby improving overall
logistics". According to the IT specialist, Company A
uses IoT sensors in connected vehicles to collect real-
time data. Edge Intelligence assists by providing real-
time data. The integration of blockchain is in progress
to ensure vehicle data security and manage
transactions between vehicles and infrastructure.
P2–A IoT-edge-AI-
blockchain 81%
"We use a device that connects to our company's
application, installed on our customers' cell phones.
This allows us to use our customers' internet to receive
data from their vehicles. EI analyzes the data at the
source, while we make decisions downstream. In
practical terms, the EI Improved demand forecasting
and resource allocation. Although there is an ongoing
b
lockchain project, which I find very interesting, we
are still in the preliminary phase - so, there is a lot to
do in that regard".
P3–B IoT-edge-AI-
blockchain 78%
"IoT-edge-AI integration has allowed us to process
data at the source, optimizing power generation and
predicting failures. We use sensors to monitor wind
turbines and solar panels". Although Company B only
plans to integrate blockchain, they recognize that this
technology can create a decentralized energy
management system, where energy production and
consumption are recorded securely and transparently.
P4–C IoT-edge-AI-
blockchain 92%
"The integrated system significantly enhances logistics
accuracy and efficiency, particularly in our field. Some
of our colleagues conduct scientific research to
improve the systems we use. Practically, we remotely
monitor our patients using IoT devices. Edge
Intelligence provides real-time analytics and alerts,
while blockchain secures sensitive data and manages
access". Several examples illustrated the advancement
of technology in this company. One example is the use
of Edge Intelligence, where data is processed at the
source rather than being sent to a central server. This
approach reduces latency, enabling faster decision-
making and real-time alerts for healthcare providers
through the Internet of Things (IoT). For instance, a
smart insulin pump can continuously analyze glucose
levels and adjust insulin delivery in real-time, thanks
to sophisticated AI algorithms. Additionally,
b
lockchain technology plays a crucial role in
maintaining data integrity and access control by
tracking and verifying patient records.
Preprints.org (www.preprints.org) | NOT PEER-REVIEWED | Posted: 23 July 2024 doi:10.20944/preprints202407.1777.v1
10
P5–D Edge computing in
IIoT 93%
“In our company, Edge Computing (EC) in the
Industrial Internet of Things (IIoT) enables us to collect
and process data from industrial machines and devices
on-site. This approach significantly improves
efficiency by facilitating predictive maintenance,
which is a more advanced method compared to the
preventative maintenance practices used a few years
ago”.
P6–D Edge computing in
IIoT 89%
“A few years ago, our supported preventive
maintenance, but this approach involved recurring
downtime and frequent maintenance costs. One of the
b
iggest paradigm shifts in our company's EC strategy
for the IIoT was the automation of maintenance
procedures. Early on, we launched projects to
implement predictive maintenance, which proved
successful and delivered financial benefits within the
first few months”. As the participant explained, EC
enabled the complete automation of maintenance
procedures for IIoT devices by utilizing local data
analysis to determine the appropriate actions. The
participant further elaborated that CE in IIoT goes
b
eyond collecting data from industrial structures. He
said that this technology not only continuously
monitors but also takes proactive actions. Through
ongoing monitoring, the maintenance team can receive
detailed diagnostics and reports. In addition to these
insights, CE in IIoT can recommend to his office
(logistics) component purchases or suggest the
replacement of devices.
P7–E
Mixed-integer
programming model &
Cloud-edge
collaborative mode
75%
“Our company is one of the largest retailers in
Portugal, operating a diverse chain of supermarkets,
clothing stores, and shopping centers. The EI
application enables us to analyze real-time data from
various sources, including traffic, weather conditions,
and stock availability. This allows to dynamically
adjust delivery routes, enhancing efficiency. We also
face challenges such as unexpected changes in product
demand across different stores, but these are highly
specific and manageable”.
P1–A IoT-edge-AI-
blockchain 93%
"Our integrated IoT-edge-AI system has significantly
improved our logistics and operational efficiency. The
b
lockchain component is still in development but
shows great promise for enhancing data security and
transaction management between devices. However,
from a practical standpoint, technology has
significantly enhanced our operational and logistical
capabilities, enabling true just-in-time efficiency.".
P2–A IoT-edge-AI-
blockchain 91%
"By using IoT and edge AI, we can process data
locally, reducing latency and improving real-time
decision-making. Blockchain will further enhance our
security measures once fully integrated. From a
Preprints.org (www.preprints.org) | NOT PEER-REVIEWED | Posted: 23 July 2024 doi:10.20944/preprints202407.1777.v1
11
logistical perspective, we now produce only what is
necessary while maintaining a safe stock of products".
Between the P1 and P2-A employees at Company A,
there is a consensus that EI has brought disruptive
changes to the organization and significantly
improved downstream logistics management.
P3–B IoT-edge-AI-
blockchain 92%
"Our implementation of IoT and edge AI in
monitoring energy systems has optimized
performance and predicted equipment failures more
accurately. Blockchain is the next step for securing and
decentralizing our energy management systems". In
the companies analyzed, we found that blockchain is
an area that still needs further exploration. However,
there is a consensus that IoT-edge-AI has brought
disruptive and widespread changes across most
companies. Both Company A and Company B can
predict needs more easily and accurately, allowing for
greater resource allocation in record time, which
would not be possible without this technology.
P4–C IoT-edge-AI-
blockchain 96%
"The combination of IoT, edge AI, and blockchain has
revolutionized our healthcare services, providing real-
time patient monitoring and data security. Blockchain
ensures the integrity and confidentiality of patient
records". In this healthcare company, several
employees conduct scientific research, necessitating
the recruitment of highly specialized personnel. Given
the critical importance of privacy in this sector, they
invested in blockchain to protect confidential data and
manage information effectively. The integration of IoT,
edge AI, and blockchain has had significant real-world
impacts on users' lives.
P5–D Edge computing in
IIoT 94%
"Edge computing has transformed our maintenance
processes by enabling predictive maintenance and
reducing downtime. This proactive approach has
significantly cut maintenance costs and improved
operational efficiency". In this multinational
technology conglomerate, there was no significant
percentage change. The company predominantly uses
EC in IIoT and is almost entirely aligned with other
companies. It operates in a more comprehensive
sector, providing technological support to several
market-leading firms.
P6–D Edge computing in
IIoT 92%
"Automation of maintenance through edge computing
has delivered substantial logistic benefits. The
technology continuously monitors and provides
actionable insights, enhancing our maintenance
strategies". The second participant from Company D is
somewhat less optimistic than P5 but recognizes that
EC in IIoT has introduced disruptive changes to the
logistics industry. He emphasizes that this technology
offers transformative benefits, particularly through
Preprints.org (www.preprints.org) | NOT PEER-REVIEWED | Posted: 23 July 2024 doi:10.20944/preprints202407.1777.v1
12
actionable measures and recommendations, which
were not available before. While final decision-making
remains with humans, he believes that technology
plays a crucial role in supporting this process.
P7–E
Mixed-integer
programming model &
Cloud-edge
collaborative mode
93%
"Using edge computing and mixed-integer
programming models, we can dynamically adjust
delivery routes based on real-time data. This improves
efficiency and helps manage demand fluctuations
across our retail network". P7-E is among those least
aligned with the rest due to its focus on Cloud-edge
collaboration. However, after further interaction, the
participant acknowledges that there is still much to be
done but recognizes that EC offers significant benefits
in terms of efficiency, particularly in managing
delivery routes. As the Director of
Operations/Logistics, this practical application is of
particular interest to him.
As illustrated in the table above, one significant impact of the EI (IoT-edge-AI) technology on
Last-Mile Delivery at Company A is the enhancement of predictability. This advancement has led to
reduced delivery times and decreased costs associated with accumulated stock. The company
employs IoT sensors to collect real-time data, which is then processed immediately using EI
technologies. Although the blockchain initiative for vehicle data security and transaction
management is still in the exploratory phase, the integration of IoT and advanced AI technologies
has already resulted in relevant improvements in operational efficiency, cost reduction, and increased
customer satisfaction. These improvements are due to the company's ability to allocate resources
more effectively and provide just-in-time services as needed. In this context, we focus on the
challenges and opportunities associated with EI ecosystems, particularly the essential collaboration
and integration between service providers and clients. Such collaboration is critical for broad resource
sharing and the continuous transfer of services. For instance, within an EI service model, a user can
simultaneously be both a service consumer and a data generator. According to the empirical data,
this dual role necessitates the development of a new, intelligent pricing scheme that accounts for both
the user's consumption of services and the value of their data contributions. Lastly, both participants
of Company A report that despite the recognized advantages of IE, companies in the EU remain
hesitant to fully adopt these technologies. This reluctance is mainly motivated by concerns regarding
high implementation and maintenance costs. This barrier is particularly pronounced among
Portuguese SMEs, which often face greater challenges in adopting these innovations due to limited
financial resources and expertise.
Company B applied IoT-edge-AI technologies for monitoring wind turbines and solar panels,
resulting in optimized power generation, enhanced failure prediction, and improved resource
management. Although blockchain integration remains a planned future endeavor, there is
recognition of its potential for developing a decentralized energy management system. B Corp
participants agreed with A Corp on how these technologies have transformed performance
monitoring and future blockchain applications, focusing on their ability to minimize latency, reduce
costs, and make faster, data-driven decisions.
Company C operates within the comprehensive EI ecosystem, which integrates IoT, edge
computing, AI, and blockchain technologies as identified by Hennebelle et al. [29]. The specialist from
this company explains that IoT-edge-AI-blockchains have led to substantial advancements in both
precision and efficiency in the domains of healthcare and logistics. In the healthcare sector, this IT
ecosystem has significantly enhanced real-time patient monitoring and improved data security.
Blockchain technology, in particular, has played a crucial role in ensuring data integrity and privacy,
while AI-driven solutions have enabled highly personalized decision-making processes that adapt to
Preprints.org (www.preprints.org) | NOT PEER-REVIEWED | Posted: 23 July 2024 doi:10.20944/preprints202407.1777.v1
13
the unique needs of each patient and require ongoing medical supervision. This approach contrasts
sharply with the EC applications in IIoT, where technology supports decision-making based on
predefined criteria. In the IIoT context, the financial stakes of decision-making are high, but the
consequences of a wrong decision are less critical compared to healthcare, where errors can endanger
human lives. As a result, decision-makers in healthcare must have greater caution and are often more
reliant on medical expertise rather than solely on technological recommendations. Additionally, it
was noted that some healthcare organizations have recognized the need to foster a culture of
innovation and hire specialists with advanced qualifications to drive technological advancements,
especially in last-mile logistics.
Returning to the comparative analysis with Company A, another key consensus emerged – the
necessity to readjust the financial model for EI services. Specifically, the pricing scheme should
consider both the user's consumption of services and the value of their data contributions. According
to the interviewed specialist, patients already experience clear benefits. For instance, patients who
require continuous hospital care can receive it at home, leading to significant savings by reducing the
need for frequent hospital visits. This approach not only provides real-time, more comfortable service
for patients but also eases their physical and financial burden. For hospitals, the benefits include a
reduction in waiting lists and decreased foot traffic, which helps improve the return on investment.
This dual advantage shows the importance of rethinking the financial model to better reflect the value
provided to both patients and healthcare institutions. Overall, the effective integration of cutting-
edge technologies for real-time analytics, combined with blockchain for ensuring data integrity, has
demonstrated a profound impact on healthcare logistics. This innovative approach underlines the
potential of EI to revolutionize sensitive fields like patient care.
Company D's emphasis on edge computing within the Industrial Internet of Things (IIoT)
illustrated how this technology has advanced predictive maintenance and reduced downtime. By
shifting from preventive to predictive maintenance, edge computing has enabled on-site data
collection and proactive maintenance measures. The participant from Company D argued on the
transformative nature of edge computing for improving maintenance processes, reducing costs, and
enhancing logistics efficiency.
Company E utilized a mixed integer programming model combined with a cloud collaborative
mode for dynamic delivery route adjustments. This approach has improved predictive capabilities,
reduced delivery times and costs, and enabled faster route adjustments. Table 6 summarizes our
analysis, highlighting the EI technologies examined and their impacts on last-mile logistics.
Table 6. EI Technologies and their Impacts on the Last Mile – combined summary.
Technology Impact on Last-Mile Delivery
1. IoT-edge-AI-blockchain
Improves predictive capabilities and runtime efficiency.
Reduces delivery times and company costs.
Improves demand forecasting and resource allocation.
Improve customer satisfaction and reduce costs.
2. EC in IIoT
Makes decisions and actions according to pre-established
criteria.
Minimizes latency/reduced downtime.
Reduce costs.
3. Mixed-integer programming model
& Cloud-edge collaborative mode Make faster decisions and route adjustments.
Overall, our study reveals a broad consensus on the potential of IoT, edge computing, AI, and
blockchain technologies. This answer to our research question, as these innovations are widely
recognized for significantly enhancing operational efficiency, data security, resource management,
and customer satisfaction across diverse sectors, including automotive, energy, and healthcare.
Additionally, EI technologies improve predictive capabilities, reduce latency, and support better
Preprints.org (www.preprints.org) | NOT PEER-REVIEWED | Posted: 23 July 2024 doi:10.20944/preprints202407.1777.v1
14
decision-making and cost reduction for technological conglomerates through IIoT, as well as in the
retail sector via cloud-edge solutions.
4. Discussion
4.1. Theoretical Contributions
This research makes several contributions to the theoretical understanding of EI. Firstly, it
expands the conceptual framework of EI by integrating it with last-mile logistics, a field that has been
relatively underexplored within the context of EI. Secondly, the study reveals that despite substantial
advancements by the People's Republic of China and the United States of America, companies in the
European Union remain hesitant to fully embrace EI. This reluctance is primarily driven by concerns
about high implementation and maintenance costs. In particular, our research highlights that
Portuguese SMEs face pronounced challenges in adopting these innovations due to limited financial
resources and expertise. Nonetheless, some companies (e.g., healthcare) are investing in EI and
recognizing the importance of fostering a culture of innovation. They are hiring specialists, often
PhDs with advanced scientific research qualifications, to drive technological advancements,
especially in last-mile logistics. Finally, the research finds that EI enables actionable decision-making
in contexts where tasks require mechanical intelligence, significantly impacting company decisions.
However, in scenarios where decision-making affects human beings (i.e., healthcare services), human
decision-makers continue to play a more prominent role.
4.2. Managerial Contributions
From a managerial perspective, this study provides practical insights for logistics managers in
various sectors, including automotive, healthcare, and retail. The findings highlight the potential of
EI to significantly enhance last-mile delivery performance by reducing delivery times, optimizing
resource allocation, and improving customer satisfaction. Managers can leverage these results to
adopt EI technologies such as IoT-edge-AI-blockchain to streamline operations and achieve cost
reductions, despite the need for significant initial investments in financial and human resources.
Hence, the study also highlights the importance of fostering a culture of innovation within
organizations. It suggests that managers should invest in recruiting and training staff with expertise
in advanced technologies. Additionally, the research advocates for a reassessment of financial models
to better capture the value generated by EI, particularly regarding data contributions and service
consumption. This recommendation is essential for managers seeking to justify initial investments in
EI technologies and develop sustainable business models.
4.3. Limitations and Future Research Avenues
Despite its contributions, this study has several limitations that suggest avenues for future
research. One primary limitation is the scope of the sample, which, although diverse, is confined to a
specific geographic region of the European Union (Portugal) and a limited number of sectors. Future
studies could expand the sample size and include a broader range of industries and geographic
locations to enhance the generalizability of the findings. Another limitation is the reliance on
qualitative methods, which, while providing depth, could be complemented by quantitative analyses
to offer a more comprehensive perspective. Additionally, further research is needed to explore the
integration of EI with other emerging technologies and its implications for logistics and supply chain
management. These future research directions can contribute to a more holistic understanding of EI
and its transformative potential in the logistics industry.
Funding: This research received no external funding.
Data Availability Statement: Data will be available upon request to the author.
Acknowledgments: We would like to extend our gratitude to all the experts who participated in this study and
generously contributed their valuable insights.
Preprints.org (www.preprints.org) | NOT PEER-REVIEWED | Posted: 23 July 2024 doi:10.20944/preprints202407.1777.v1
15
Conflicts of Interest: The author declares no conflicts of interest.
References
1. Galkin, A.; Obolentseva, L.; Balandina, I.; Kush, E.; Karpenko, V.; Bajdor, P. Last-Mile Delivery for
Consumer Driven Logistics. Transportation Research Procedia 2019, 39, 74–83.
https://doi.org/10.1016/j.trpro.2019.06.009.
2. Correia, D.; Teixeira, L.; Marques, J.L. Last-Mile-as-a-Service (LMaaS): An Innovative Concept for the
Disruption of the Supply Chain. Sustainable Cities and Society 2021, 75, 103310.
https://doi.org/10.1016/j.scs.2021.103310.
3. Ha, N.T.; Akbari, M.; Au, B. Last Mile Delivery in Logistics and Supply Chain Management: A Bibliometric
Analysis and Future Directions. Benchmarking: An International Journal 2022, 30, 1137–1170.
https://doi.org/10.1108/BIJ-07-2021-0409.
4. Boysen, N.; Fedtke, S.; Schwerdfeger, S. Last-Mile Delivery Concepts: A Survey from an Operational
Research Perspective. OR Spectrum 2021, 43, 1–58. https://doi.org/10.1007/s00291-020-00607-8.
5. Zhou, Z.; Chen, X.; Li, E.; Zeng, L.; Luo, K.; Zhang, J. Edge Intelligence: Paving the Last Mile of Artificial
Intelligence With Edge Computing. Proceedings of the IEEE 2019, 107, 1738–1762.
https://doi.org/10.1109/JPROC.2019.2918951.
6. Muñoz-Villamizar, A.; Velazquez-Martínez, J.C.; Caballero-Caballero, S. A Large-Scale Last-Mile
Consolidation Model for e-Commerce Home Delivery. Expert Systems with Applications 2024, 235, 121200.
https://doi.org/10.1016/j.eswa.2023.121200.
7. Reis, J. Exploring Applications and Practical Examples by Streamlining Material Requirements Planning
(MRP) with Python. Logistics 2023, 7, 91. https://doi.org/10.3390/logistics7040091.
8. Akbari, M. Revolutionizing Supply Chain and Circular Economy with Edge Computing: Systematic
Review, Research Themes and Future Directions. Management Decision 2023, ahead-of-print.
https://doi.org/10.1108/MD-03-2023-0412.
9. Gordienko, Y.; Kochura, Y.; Taran, V.; Gordienko, N.; Rokovyi, O.; Alienin, O.; Stirenko, S. Chapter Nine -
“Last Mile” Optimization of Edge Computing Ecosystem with Deep Learning Models and Specialized
Tensor Processing Architectures. In Advances in Computers; Kim, S., Deka, G.C., Eds.; Hardware Accelerator
Systems for Artificial Intelligence and Machine Learning; Elsevier, 2021; Vol. 122, pp. 303–341.
10. Özarık, S.S.; da Costa, P.; Florio, A.M. Machine Learning for Data-Driven Last-Mile Delivery Optimization.
Transportation Science 2024, 58, 27–44. https://doi.org/10.1287/trsc.2022.0029.
11. Kalboussi, E.; Ndhaief, N.; Rezg, N. Last-Mile Optimization Using Neural Networks. Applied Sciences 2024,
14, 787. https://doi.org/10.3390/app14020787.
12. Aljohani, K. The Role of Last-Mile Delivery Quality and Satisfaction in Online Retail Experience: An
Empirical Analysis. Sustainability 2024, 16, 4743. https://doi.org/10.3390/su16114743.
13. De Maio, A.; Ghiani, G.; Laganà, D.; Manni, E. Sustainable Last-Mile Distribution with Autonomous
Delivery Robots and Public Transportation. Transportation Research Part C: Emerging Technologies 2024, 163,
104615. https://doi.org/10.1016/j.trc.2024.104615.
14. Reis, J.; Santo, P.E.; Melão, N. Artificial Intelligence in Government Services: A Systematic Literature
Review. In New Knowledge in Information Systems and Technologies; Rocha, Á., Adeli, H., Reis, L.P., Costanzo,
S., Eds.; Advances in Intelligent Systems and Computing; Springer International Publishing: Cham, 2019;
Vol. 930, pp. 241–252 ISBN 978-3-030-16180-4.
15. Reis, J.; Amorim, M.; Melão, N. Multichannel Service Failure and Recovery in a O2O Era: A Qualitative
Multi-Method Research in the Banking Services Industry. International Journal of Production Economics 2019,
215, 24–33. https://doi.org/10.1016/j.ijpe.2018.07.001.
16. Reis, J.C.G. dos Development of an N-Helix Innovation Model for the Portuguese Defense Industry.
doctoralThesis, 2023.
17. Lagorio, A.; Zenezini, G.; Mangano, G.; Pinto, R. A Systematic Literature Review of Innovative
Technologies Adopted in Logistics Management. International Journal of Logistics Research and Applications
2022, 25, 1043–1066. https://doi.org/10.1080/13675567.2020.1850661.
18. Elsevier Scopus | Abstract and Citation Database | Elsevier Available online:
https://www.elsevier.com/products/scopus# (accessed on 7 July 2024).
19. Jacsó, P. Google Scholar: The Pros and the Cons. Online Information Review 2005, 29, 208–214.
https://doi.org/10.1108/14684520510598066.
20. Peng, G.; Wen, Y.; Dou, W.; Li, T.; Xu, X.; Ye, Q. Edge Intelligence Empowered Delivery Route Planning
for Handling Changes in Uncertain Supply Chain Environment. J Cloud Comp 2024, 13, 69.
https://doi.org/10.1186/s13677-024-00613-z.
21. Dehury, C.K.; Srirama*, S.N.; Donta, P.K.; Dustdar, S. Securing Clustered Edge Intelligence With
Blockchain. IEEE Consumer Electronics Magazine 2024, 13, 22–29. https://doi.org/10.1109/MCE.2022.3164529.
Preprints.org (www.preprints.org) | NOT PEER-REVIEWED | Posted: 23 July 2024 doi:10.20944/preprints202407.1777.v1
16
22. Luo, H.; Chen, T.; Li, X.; Li, S.; Zhang, C.; Zhao, G.; Liu, X. KeepEdge: A Knowledge Distillation
Empowered Edge Intelligence Framework for Visual Assisted Positioning in UAV Delivery. IEEE
Transactions on Mobile Computing 2023, 22, 4729–4741. https://doi.org/10.1109/TMC.2022.3157957.
23. Nilsson, M.; Pina, T.J.; Khacef, L.; Liwicki, F.; Chicca, E.; Sandin, F. A Comparison of Temporal Encoders
for Neuromorphic Keyword Spotting with Few Neurons. In Proceedings of the 2023 International Joint
Conference on Neural Networks (IJCNN); June 2023; pp. 1–7.
24. Yiwen, L.; Zhirong, Z.; Tangyan; Wenkan, W.; Xiaoning, P.; Yuanquan, S. Research on Fast Adaptive
Transmission Models for International Inland Port Based on Edge Intelligence. In Proceedings of the 2023
IEEE 10th International Conference on Cyber Security and Cloud Computing (CSCloud)/2023 IEEE 9th
International Conference on Edge Computing and Scalable Cloud (EdgeCom); IEEE: Xiangtan, Hunan,
China, July 2023; pp. 324–329.
25. Ugarte Fajardo, J.; Maridueña-Zavala, M.; Cevallos-Cevallos, J.; Ochoa Donoso, D. Effective Methods Based
on Distinct Learning Principles for the Analysis of Hyperspectral Images to Detect Black Sigatoka Disease.
Plants 2022, 11, 2581. https://doi.org/10.3390/plants11192581.
26. Xhafa, F.; Aly, A.; Juan, A.A. Allocation of Applications to Fog Resources via Semantic Clustering
Techniques: With Scenarios from Intelligent Transportation Systems. Computing 2021, 103, 361–378.
https://doi.org/10.1007/s00607-020-00867-w.
27. Zhang, C.; Liu, X.; Xu, J.; Chen, T.; Li, G.; Jiang, F.; Li, X. An Edge Based Federated Learning Framework
for Person Re-Identification in UAV Delivery Service. In Proceedings of the 2021 IEEE International
Conference on Web Services (ICWS); IEEE: Chicago, IL, USA, September 2021; pp. 500–505.
28. Qiu, T.; Chi, J.; Zhou, X.; Ning, Z.; Atiquzzaman, M.; Wu, D.O. Edge Computing in Industrial Internet of
Things: Architecture, Advances and Challenges. IEEE Commun. Surv. Tutorials 2020, 22, 2462–2488.
https://doi.org/10.1109/COMST.2020.3009103.
29. Hennebelle, A.; Ismail, L.; Materwala, H.; Al Kaabi, J.; Ranjan, P.; Janardhanan, R. Secure and Privacy-
Preserving Automated Machine Learning Operations into End-to-End Integrated IoT-Edge-Artificial
Intelligence-Blockchain Monitoring System for Diabetes Mellitus Prediction. Computational and Structural
Biotechnology Journal 2024, 23, 212–233. https://doi.org/10.1016/j.csbj.2023.11.038.
30. Mahamkali, A.; Gali, M.; Muniyandy, E.; Sundaram, D.A. IoT-Empowered Drones: Smart Cyber Security
Framework with Machine Learning Perspective. In Proceedings of the 2023 International Conference on
New Frontiers in Communication, Automation, Management and Security (ICCAMS); IEEE: Bangalore,
India, October 27 2023; pp. 1–9.
31. Drisko, J.W.; Maschi, T. Content Analysis; Pocket guides to social work research methods; Oxford University
Press: New York, 2016; ISBN 978-0-19-021549-1.
32. Sinha, A.; Chowdhury, D.; Sharma, S.; Sherke, Y.R.; Das, D. nCare: Fault-Aware Edge Intelligence for
Rendering Viable Sensor Nodes. Internet of Things 2023, 21, 100643. https://doi.org/10.1016/j.iot.2022.100643.
33. Himeur, Y.; Sayed, A.N.; Alsalemi, A.; Bensaali, F.; Amira, A. Edge AI for Internet of Energy: Challenges
and Perspectives. Internet of Things 2024, 25, 101035. https://doi.org/10.1016/j.iot.2023.101035.
34. Da, B.; Ong, Y.-S.; Gupta, A.; Feng, L.; Liu, H. Fast Transfer Gaussian Process Regression with Large-Scale
Sources. Knowledge-Based Systems 2019, 165, 208–218. https://doi.org/10.1016/j.knosys.2018.11.029.
35. Amadeo, M.; Cicirelli, F.; Guerrieri, A.; Ruggeri, G.; Spezzano, G.; Vinci, A. When Edge Intelligence Meets
Cognitive Buildings: The COGITO Platform. Internet of Things 2023, 24, 100908.
https://doi.org/10.1016/j.iot.2023.100908.
36. Pradhan, S.; Tripathy, S.; Matam, R. Towards Optimal Edge Resource Utilization: Predictive Analytics and
Reinforcement Learning for Task Offloading. Internet of Things 2024, 26, 101147.
https://doi.org/10.1016/j.iot.2024.101147.
37. Alrashdi, I.; Sallam, K.M.; Alqazzaz, A.; Arain, B.; Hameed, I.A. A Contrastive Learning Approach for
Enhanced Robustness for Strengthening Federated Intelligence in Internet of Visual Things. Internet of
Things 2024, 26, 101206. https://doi.org/10.1016/j.iot.2024.101206.
38. Dalabehera, A.R.; Bebortta, S.; Kumar, N.; Senapati, D. Mist–Fog-Assisted Real-Time Emotion Recognition
Using Deep Transfer Learning Framework for Smart City 4.0. Internet of Things 2024, 27, 101237.
https://doi.org/10.1016/j.iot.2024.101237.
39. Huang, Z.; Zhong, W.; Li, D.; Lu, H. Delay Constrained SFC Orchestration for Edge Intelligence-Enabled
IIoT: A DRL Approach. J Netw Syst Manage 2023, 31, 53. https://doi.org/10.1007/s10922-023-09743-2.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those
of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s)
disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or
products referred to in the content.
Preprints.org (www.preprints.org) | NOT PEER-REVIEWED | Posted: 23 July 2024 doi:10.20944/preprints202407.1777.v1
