325
1 INTRODUCTION
The term Artificial Intelligence (AI) can be (quite
simplistically) approached as the effective simulation
of human intelligence processes by computer systems;
without the following list being viewed as complete, a
few indicative examples of AI include expert systems,
natural language processing and speech recognition
(used by voice assistants to “understand and respond
to user commands), as well as machine vision (serving
the task of object/other vessel detection from nearfield
to horizon), among others [1]. Focusing on the financial
aspect of AI, an extended number of relevant
applications have already been employed by various
organizations around the world, with the aim to boost
their revenues by streamlining the related business
procedures, automating repetitive jobs, and improving
customer service. When the discussion is shifted to the
wider shipping industry, the so-called “Digitalization”
phenomenon -which also includes the topic of
Maritime Autonomous Surface Ships (MASS)- there is
a quite disruptive picture of how this industry may be
transformed in the near future [1-3].
The introduction into full service of a certain
number of MASS-type vessels (along with numerous
other advanced technology applications based on
digital means) provides a clear argument that shipping
has already entered a new era [3]. The rather simplistic
and at most times confusing term “Autonomous
Vessels” is often used to describe certain types of
modern ships (complex systems engaged in maritime
transport activities) that, to some extent, are able to
make decisions by themselves, requiring no human
input. This type of vessels is very heavily reliant on AI
applications for the conduct of safe navigation, obstacle
detection, and decision-making; the role of trustworthy
AI applications that can reliably serve the necessary
decision-making tasks needed is vital towards the
effective introduction of more MASS-type vessels in
full service [1-4]. In any case, safely and effectively
operating any ship under the MASS framework -
Artificial Intelligence (AI) Applications
and the Shipping Industry
D. Dalaklis
World Maritime University, Malmö, Sweden
ABSTRACT: Artificial Intelligence (AI) can be simply approached as the (effective) simulation of human
intelligence processes by computer systems. The issue of Maritime Autonomous Surface Ships (MASS), based on
support by numerous AI applications, is providing a quite disruptive picture of how the shipping industry may
be transformed in the future. After the necessary clarification of terms, a summary of certain important legal
developments in relation to the on-going introduction of MASS type vessels into full service is provided. The role
of trustworthy AI applications that can reliably serve the associated decision-making tasks is also discussed. In
the near future, the vast majority of maritime transport needs will continue to be served by those vessels termed
as “conventional” (regularly manned ships); the shipping industry is well known for its risk adverse behaviour
and a slow pace of adaptation towards this new operating paradigm is the most probable path of adoption.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 19
Number 1
March 2025
DOI: 10.12716/1001.19.01.38
326
whether it is controlled remotely, or, under a “fully
autonomous” classification- requires robust regulation
to ensure that nothing will go wrong therefore
protecting human lives, as well as cargo on board and
the vessel itself.
Over the course of time, the relationship between
humans and technology has always been associated
with continuous transformation [5]; however, history
also testifies that the relevant pace of change clearly
increased with the spread of steam power that set off
the notorious “Industrial Revolution” (IR), which is
related with wide-ranging technological discoveries,
particularly in the areas of agriculture, manufacturing,
mining, metallurgy, and transport, including off course
the very widespread application of the factory system
[6]. The IR (is a still on-going transformative process,
which) corresponds to a period of global transition of
the economy towards more widespread, efficient and
stable manufacturing. At its very beginning, water and
steam power made possible the expansion of factories,
replacing manual labour with machines. Furthermore,
a number of quite important developments followed
next (Fig. 1). Especially for shipping, it should be
highlighted that has benefitted via three main ways
during that early period: a) the introduction of steam
power on merchant ships (mechanization); b) an
enlarged volume of goods to be traded and demand for
raw materials, because of the increased factory output;
and c) certain metallurgical innovations that also
improved shipbuilding techniques [3, 6].
Figure 1. Different Stages of the Industrial Revolution and
Applications related to Industry 4.0 [1].
In summary, the first stage of the industrial
revolution involved a change from agrarian societies to
greater industrialization as a result of the steam engine;
the second stage of this evolutionary process (or,
second industrial revolution) was driven by electricity
and involved very significant expansion of industrial
production, because of the introduction of the
assembly line concept, as well as an extended number
of related technological advances (spanning over the
18th and 19th centuries, with a spill over to the early
20th century). Important innovators like Thomas
Edison and Henry Ford played crucial roles during
that era, introducing electricity to power factories and
revolutionizing mass production techniques [1,3].
Subsequently, the third stage of industrial revolution
(which is quite often termed as “digital revolution”),
involved the development of computers and
information technology (IT) (including the creation of
the Internet) during the second half of the 20th century.
A very important issue to highlight is the widespread
introduction of automation in various industrial
sectors (the shipping one also included) during this
period, which improved precision, efficiency, and the
ability to handle complex tasks, marking therefore a
significant leap forward in capabilities. Finally, the
current fourth stage of the industrial revolution is
considered as a “new era” rather than a continuation of
the third stage, because of fast and rapid developments
and the disruptiveness of associated technologies [3].
In other words, the term “fourth (stage of the)
industrial revolution” (4IR, or Industry 4.0) describes
the current and foreseeable environment in which
disruptive technologies and trends such as Cloud
Computing and the Internet of Things (IoT), Artificial
Intelligence (AI) and Big Data Analytics, Robotics,
Virtual and Augmented Reality (VR/AR) and/or
Simulation applications are all changing the way
humans live and work. [3, 7] All these technologies
(summarized with the help of Fig. 1) are working in the
same collaborative environment (interconnected, most
commonly via the Internet); this in turn allows various
computers, or even very complex systems (often
described as “machines”, or “system of systems” in the
wider literature) to easily communicate, analyse data
in real-time, and even make informed decisions (by
utilizing some form of AI application). For example, a
report under the title “Transport 2040: Automation,
Technology, Employment - The Future of Work”,
which was released a few years ago by the World
Maritime University (WMU), puts forward the notion
that: “Technological progress and innovation have
occurred throughout history and changed its course,
for example the Industrial Revolution in the eighteenth
and nineteenth centuries. Currently, we are about to
embrace what is now termed the Fourth Industrial
Revolution, which is characterized by the introduction
of artificial intelligence, robotics, more and more
interconnection, among other innovations” [8].
Especially for shipping, the multi-level impacts of
4IR have the potential to completely change the
contemporary framework of operations and associated
activities. A number of reports and academic research
efforts imply that the tasks of seafarers are expected to
be transformed into more digital ones, such as system
management and monitoring of operations, as well as
that their operational work may be decreased; statistics
also clearly indicate that more than 1,000 maritime
autonomous surface ships (MASS of varying
capabilities) are already operated by (more than) 50
organizations/companies throughout the world [4]. A
very indicative example of MASS-related vessel is the
construction project of MV Yara Birkeland, (already in
full operation) in Norway. This (zero-emission) vessel
transports mineral fertilizer from Yara's production
plant in Porsgrunn to the regional export port in
Brevik; it was put into commercial operation during
the spring of 2022. During the first two years of
operation, the vessel went through a gradual transition
towards a “full autonomous” mode of sailing [9]. This
initiative facilitated to substitute cargo transport
operated by trucks with a MASS-type container ship
between specific places for environmental and
327
economic reasons. Therefore, it does not prevent other
traditional ships operating on the exact same route. At
the same time, it must be noted that the capacity of that
vessel is quite limited (3000-4000 TEU) as compared
with traditional container ships (20,000 TEU, or much
more). Finally, it is worth highlighting that this ship is
owned by the shipper itself, instead of a traditional
ship-owner.
Another “similar” example is the Autonomous
Spaceport Drone Ship (ASDS) that is utilized by
SpaceX company. The overall engineering approach is
to facilitate cost-efficient launching of space ships by
collecting and reusing the first stages of rockets after
launching, and ASDA captures the exhausted first
stage. Cutting a long way short, ASDSs can sustain
their positions autonomously, or by remote control
from another support ship when the rockets will be
landing. There is also another (quite similar) ship
(same concept), under the name A Shortfall of Gravitas
(ASOG), which does not require support by tug boats
anymore (Fig. 2) [3]. Like the previously mentioned
Yara Birkeland, ASDS has been developed to meet a
new need, which is to collect exhausted rockets, and it
will never replace conventional (or, as often called
“traditional”) ships. Thus, those MASS type ships that
were previously presented can be viewed as something
totally different from the vast majority of ships in
service today; for the time being they cannot easily
replace traditional ships (except, maybe for the tasks of
short-range domestic passenger ferries, which is
looking as a very appropriate business model by
considering that human intervention (in case
something goes wrong) is quite easy because of the
vicinity of the troubled vessel with the coastline).
However, it is also useful to highlight that work at
sea/seafaring is (very probably) expected to change in
the not-so-distant future; roles, organizational
structure and responsibilities will probably shift from
today’s normal way of working at sea (that involves
very long periods away from home) to an office type
profile that will mainly involve monitoring, managing
and supervising systems (from ashore).
Figure 2. The vessel “A Shortfall of Gravitas” [3].
Combining computer vision and deep learning
capabilities so that MASS type vessels can effectively
understand” their surrounding environment (therefore
achieving a complete situational awareness picture)
and safely navigate in close proximity with regularly
manned” vessels is a very probable future. But, before
moving to a different direction, it is necessary to note
that certain MASS-type vessels are already working
alongside ordinary manned vessels with minimal
autonomous-specific regulation. With this lack of
guidance in mind, the International Maritime
Organization (IMO) has already presented the results
of its latest autonomous scoping exercise and
established a road map for autonomous regulation.
The IMO aims to integrate new and advancing
technologies in its regulatory framework -balancing
the benefits derived from new and advancing
technologies against safety and security concerns, the
impact on the environment and on international trade
facilitation, the potential costs to the industry, and their
impact on personnel, both on board and ashore. The
Organization wants to ensure that the regulatory
framework for Maritime Autonomous Surface Ships
(MASS) keeps pace with technological developments
that are rapidly evolving. Items under consideration
include remote control station related regulation;
determination of remote operators as mariners;
autonomous shipping SOLAS equipment
requirements; and regional-specific regulations [10,11].
It is a rather self-explanatory fact that automation and
the on-going digital transformation will fundamentally
change the shipping industry; the skills needed for the
next generation of deck and engineering officers will be
different than those required until now.
2 MASS TERMINOLOGY AND CLARIFICATIONS
RELATED TO AI APPLICATIONS
Etymologically, the term technology” is derived by
the combination of the ancient Greek words “τέχνη”
(tékhnē), which meant “knowledge of how to make
things” and “λόγος” (logos) that described concepts
like “reason”, or “consideration” [5]. Much can be said
about “autonomy”, but perhaps etymology provides
the most appropriate method to approach the specific
term. It is also useful to remember that not only
technology, but also language is transformed as time is
passing by. In the older versions of the Greek language,
the work “αὐτονομία” (autonomia) is derived by
combining together the words “αὐτός” (auto) -by
having as a starting point in Ancient Greek ἑαυτο
(heautoû), changed into Byzantine Greek towards
“ἑαυτός” (heautós) and then simplified even further-
which means “self” and the word “νόμος” (nomos),
meaning in turn “law” and resulting into “αὐτόνομος”
(autonomos); the later can be (combined) understood
to mean “one who gives oneself one's own law” [1].
This is why in developmental psychology and
moral, political, and bioethical philosophy, autonomy
is viewed as is the capacity to make informed and
uncoerced decisions. Shifting the discussion back
towards shipping, a certain number of publications
and reports utilize concepts like “autonomous vessels”
(or, “autonomous ships”), along with uncrewed
vessels” and “unmanned ships” interchangeably. To
avoid confusion, it is clarified that an unmanned (or,
uncrewed) ship is simply a vessel without crew on-
board. Then, it is emphasized that the content of this
paper is relying solely on the IMO’s terminology in
relation to MASS. According to IMO, a Maritime
Autonomous Surface Ship (MASS) is defined as a ship
which, to a varying degree, can operate independently
of human interaction (and making AI applications a
necessity); computer and software can help to carry out
328
various ship functions such as navigation, propulsion,
steering, and control of machinery. It is perhaps better
to use the term MASS type vessel and stay away from
the most times confusing term “Autonomous Vessel”
[1,3].
The wider maritime transport industry, which is
comprised by shipping companies and a very extended
number of related vessels (serving a literally huge
portfolio of maritime transport needs), as well as of a
really huge number of ports (with varying sizes and
capabilities), along with ship’ building companies
(shipyards, etc.) and various entities ashore (like ship
management commercial entities, maritime
administration designated authorities, port authorities,
among others) is already under the strong influence of
a digital transformation and is kind of self-fulfilling
prophecy that the utilisation of MASS type vessels will
continue to grow in the years to come. Technological
advances in sensors, electromechanical drives,
cameras, and satellite technology can help this type of
ships move across the seas and oceans of our planet
without human intervention. In the contemporary and
“digital-reliant” era, the correct combination of
hardware and software can effectively handle all tasks
related to ship operations. Indicative examples include
equipment status monitoring, engine control, cargo
control/loading, as well as docking and undocking,
which all can be performed by systems without a
human in the loop [1].
It has already been clearly emphasised that the
shipping industry’s operating paradigm is under the
influence of a digital transformation and it expected
that MASS-type vessels will expand their operations
along/in close vicinity of regular (or “regularly
manned”) ships (mixed traffic, or hybrid type
scenario). There is also no doubt that MASS level 4
ships will utilise advanced AI algorithms to safely
navigate the seas and oceans of our planet. More MASS
type vessels will be introduced into full service in the
near future of our planet and it is therefore not a
coincidence that according to a relevant Fortune
Business Insights report [12]: The global autonomous
ships market size was valued at USD 5.61 billion in
2023 and is projected to grow [from USD 6.11 billion in
2024] to USD 12.25 billion by 2032, exhibiting a CAGR
of 9.1% during the forecast period. Asia Pacific
dominated the autonomous ship market with a market
share of 37.79% in 2023. Additionally, the related
projections for the years to come are summarized
below (Fig. 3).
Figure 3. Asia Pacific region (projected) MASS type vessels
market size [12].
Steering the discussion towards to topic of AI, this
term (AI) is very commonly utilised by the IT industry
and deals with the design and implementation of
computer systems that mimic elements of human
behaviour that imply (even elementary) intelligence:
learning, adaptability, drawing conclusions, contextual
understanding, problem-solving, etc. AI is a crossroads
between multiple sciences, such as computer science,
psychology, philosophy, neurology, linguistics, and
engineering, to synthesize intelligent behaviour, with
elements of reasoning, learning, and adaptation to the
environment, while usually applied on specially
designed computers [13]. Its ultimate scope is to create
a fully functional “thinking machine” that is
intelligent, has consciousness, has the ability to learn,
has free will and is ethical. The term is often be credited
to John McCarthy (1956); Alan Turing had a few years
earlier devised the notorious Turing test, as a way to
test the intelligent behaviour of a machine [1]. It is a
rather common knowledge that AI applications have
already helped numerous organizations around the
world to boost their revenues by streamlining the
related business procedures, automating repetitive
jobs, and improving customer service. Therefore, it is
not a coincidence that the financial impact of AI
utilisation and rather pompous terms like “How
generative AI will reshape the enterprise?” have
recently dominated the public sphere. It is very clear
that AI has the potential to lead to a massive
productivity boom but, unfortunately that impact
(most likely) won’t be shared equally across economies
around the world.
AI applications can be divided into those relating to
symbolic artificial intelligence. This domain attempts
to simulate human intelligence algorithmically by
using high-level symbols and logical rules. There is
also sub-symbolic artificial intelligence that seeks to
reproduce human intelligence by using elementary
numerical models that synthesize inductive intelligent
behaviours with the sequential self-organization of
simpler structural components (“Behavioural artificial
intelligence”) and therefore simulating brain function
(“Computational intelligence”) or are simply based on
the application of statistical methodologies [13, 14]. On
the other hand, the so-called “conventional” AI
involves most commonly machine learning methods,
which are characterized by rigorous mathematical
algorithms and statistical methods of analysis and
divided into [15]: a) Experienced or specialized
systems (Expert systems), which implement
programmed logic routines, designed exclusively for a
specific task, to give a conclusion. To this end, large
amounts of known information are processed; b) Case-
based reasoning. The solution to a problem is based on
the previous solution of similar problems; c) Bayesian
networks. They are based on statistical analysis for
decision-making and are useful tools with knowledge
discovery since directed acyclic graphs allow
representing causal relations between variables; d)
Behaviour-based AI. This describes a method of
shredding the logical process and then manually
constructing the result.
Facilitating the execution of AI applications via
relevant computer systems can be approached as
learning through repetitive processes (configuration).
The specific learning process can be based on empirical
data and non-symbolic methods. This type of activity
329
can be distinguished in: a) Artificial neural networks,
with very powerful pattern recognition capabilities.
These networks “simulate the function of the neurons
of living beings; b) Fuzzy logic systems. They are based
on decision-making techniques under uncertainty.
Systems as such are based on the concept of partial
truth, where the truth value may range between
completely true and completely false; c) Evolutionary
computation. Their development arose from the study
of living organisms and relate to concepts such as
population, mutation, and natural selection (survival
of the fittest) to more accurately solve a problem. These
methods can be further distinguished into
evolutionary algorithms and swarm intelligence, such
as algorithms that simulate the behaviour of an ant
community.
Focusing mainly on Machine Learning (ML), it
should be clarified that it relates towards “helping” a
machine (computer system) to develop ways of
learning, analogous to how a human being is able to
learn; this can be further divided into: a) supervised
learning, b) unsupervised learning; c) supportive
learning. The complete analysis is outside the scope of
this paper, but in order to provide a few more
clarifications/details, the term Supervised Learning
describes the process where the algorithm constructs a
function that represents given inputs (set of training) in
known desired outputs, with the ultimate goal of
generalizing this function to inputs with unknown
output. It is used in problems like classification,
prediction and interpretation. Next, Unsupervised
Learning, is describing the situation where the
algorithm constructs a model for a set of inputs in the
form of observations without knowing the desired
outputs. It is most commonly used in problems like
Association Analysis and Clustering. Additionally,
there is Reinforcement Learning, which describes the
situation where the algorithm learns an action strategy
through direct interaction with the environment and
Ensemble methods, that describe the situation of using
the together (combine) the results from multiple
learning algorithms or different initial data to obtain a
better overall performance for the relevant computer
system [16].
3 SUMMARY AND CONCLUSIONS
Today, over four fifths of all trade in the world flows
through the seas and oceans; this includes the crucial
trade of food, energy, and other essential goods. The
Covid pandemic and on-going geopolitical pressures,
put the maritime transport industry at a crossroads,
with many forces at play reshaping the sector’s roles
and operating landscape; at the same time, maritime
logistics are becoming more dependent on digitalized
processes [17]. MASS type vessels can provide
shipowners/charterers with various economic,
environmental and risk-related benefits. These benefits
can include decreased voyage times, reduced fuel
consumption, and no need to expose crews at risk.
Voyage efficiency can benefit many aspects of the
industry: in addition to saving expenses, it can help
parties with (binding and nonbinding) efforts to reduce
emissions and adhere to already established
international standards on carbon reduction. Taking
advantage of the benefits associated with the fourth
stage of the industrial revolution (also, often termed as
the “Industry 4.0in the wider literature) has already
become a priority for various shipping companies and
ports globally, as part of their further development.
AI applications are increasingly being utilised by
various stakeholders of the wider shipping industry to
improve the current state of affairs and promote
further development; indicative AI applications that
are already used in MASS type vessels, include
reinforcement learning (mainly for safe navigation
purposes), computer vision (for obstacle detection and
improving situational awareness, although in the case
of a collision/accident the question of liability remains
unclear), and last but not least predictive maintenance
algorithms (for reducing related maintenance costs)
[18, 19].
It is indicative the on-going introduction and
further expansion of automation and various cutting-
edge technologies within the wider shipping industry,
such as exploitation of Cloud Computing and the
Internet of Things (IoT) concepts, deploying Artificial
Intelligence (AI) and Big Data Analytics (the shipping
industry is constantly and very heavily dependent on
timely and accurate data to feed its overall and quite
complex logistical plans), utilizing Autonomous
Vehicles/Robotics and Access Control Sensors, as well
as integrating Virtual and Augmented Reality (VR/AR)
and/or Simulations in daily operational (or, related
training) activities, among others. All these
applications are clearly standing out as important
enablers that can substantially improve efficient
operations and related profits. It is self-explanatory fact
that the various advanced technology applications
related to 4IR (Industry 4.0) are exercising significant
influence upon the shipping industry. However, in the
next years to come the vast majority of maritime
transport needs will continue to be served by those
vessels termed as “conventional” (or, “regularly
manned ships”); the shipping industry is well known
for its risk adverse behaviour and a slow pace of
adaptation towards this new operating paradigm will
(most likely) be adopted.
It is also true that at any point of time the building
characteristics and equipment of ships are heavily
reliant on the vessels’ intended” mission, and most
importantly, upon the technology applications
available to support these quite complex activities.
Modern ships are already very heavily equipped with
numerous technologically advanced systems and
highly automated. At the same time, the on- going
improvement and integration-interconnection of
electronics systems, as well as advances in automations
and robotics have created a new operating
environment for the shipping industry, with MASS
type vessels being capable of transitioning from
research tasks/efforts into full and effective service;
relevant opportunities are just waiting to be reaped. At
the same time, it is important to factor in that AI
applications have reached a certain maturity level, and
associated software can help to truly break down any
existing limitations and create a collaborative
environment suitable for people and “machines”,
including off course remotely controlled unmanned
vessels. In any case, gradually adopting, testing, and
re-evaluating the expected technological
breakthroughs can help to build the necessary
confidence for effective “human-machine” interaction.
330
Only extremely effective and with response in real time
AI applications can pave the way towards autonomous
systems and eventually to fully unmanned (uncrewed)
ships.
Trustworthy AI applications that can reliably serve
the decision-making tasks needed for the safe conduct
of navigation (under all weather conditions and no
matter the complexity of surrounding traffic) are vital
towards the effective introduction of more MASS-
related vessels into full service. At the same time, it is
important to consider that the hardware element of the
numerous sensors onboard “conventional”
contemporary ships has already exhausted any room
of further improvement; the use of advanced software
applications and utilization of AI tools to improve
more the capabilities of the various already existing
systems used to support the conduct of navigation on-
board those vessels could be the best way forward.
Coming to an end, another conclusion clearly standing
out is that building, improving and running AI
applications requires immense computing power; a
Cloud-based architecture can offer that in a flexible and
easy “scalable” environment (at relatively low-cost and
without huge initial investments), at least at the early
stages of development; off course, when the issue of
security will become the top priority, different
modalities should be created to ensure that there is no
opportunity to mess up with the AI application(s).
Last, but not least, effective management of “Big Data”
and deploying the right analytical tools should be
approached as a prerequisite for AI; for example, the
exploitation of Big Data and the role of certain software
applications in accessing and managing large volume
of information are key factors for improving (and even
optimizing) the conduct of operations and effective
management related activities. In turn, AI applications
can provide the solution to process unstructured data
and derive useful insights from it.
REFERENCES
[1] Dalaklis, D. (2024). Future Trends in Autonomous
Shipping: Discussing the Impact of Artificial Intelligence
(AI) Applications, International Conference on
Autonomous Shipping in Asia and the Pacific (United
Nations Economic and Social Commission for Asia and
the Pacific), Bangkok-Thailand, 28 February 2024.
https://doi.org/ 10.13140/RG.2.2.35201.43369
[2] Chen, X., Ma, D., & Liu, R. W. (2024). Application of
artificial intelligence in maritime transportation. Journal
of Marine Science and Engineering, 12(3), 439.
https://doi.org/10.3390/jmse12030439
[3] Ichimura, Y., Dalaklis, D., Kitada, M. & Christodoulou, A.,
(2022). Shipping in the era of digitalization: Mapping the
future strategic plans of major maritime commercial
actors, Digital Business, 2(1), 100022.
https://doi.org/10.1016/j.digbus.2022.100022
[4] MARITIME EXECUTIVE. (2021). Autonomous Vessels
are Becoming a Commercial Reality. https://maritime-
executive.com/editorials/autonomous-vessels-are-
becoming-a-commercial-reality, accessed 17 June 2024.
[5] Encyclopedia Britannica: Technology,
https://www.britannica.com/technology/technology,
retrieved 09/10/2024
[6] Dalaklis, D., Fonseca, T. & Schröder-Hinrichs, J.U. (2019).
How will automation and digitalization impact the future
of work in cargo transport and handling? The ITF/WMU
Transport 2040 Report, The International Cargo Handling
Coordination Association (ICHCA) International 2019
Conference: NEW RULES, NEW TECH, NEW WORLD,
Valetta-Malta, 12 November 2019.
https://doi.org/10.13140/RG.2.2.18537.44644
[7] Sanchez-Gonzalez, P.L., Díaz-Gutiérrez, D., Leo, T.J. &
Núñez-Rivas, L.R. Toward Digitalization of Maritime
Transport? Sensors. 2019, 19(4), 926.
https://doi.org/10.3390/s19040926
[8] WORLD MARITIME UNIVERSITY (WMU). (2023).
Transport 2040: Impact of Technology on Seafarers - The
Future of Work.
https://doi.org/http://dx.doi.org/10.21677/230613
[9] Yara Birkeland (Press kit, 2024),
https://www.yara.com/news-and-media/media-
library/press- kits/yara-birkeland-press-kit/, retrieved
09/03/2025.
[10] Report of the MSC-LEG-FAL Joint Working Group on
MARITIME AUTONOMOUS SURFACE SHIPS (MASS)
on its second session, MSC 107/5/1 C.F.R. (2023).
[11] DEVELOPMENT OF A GOAL-BASED INSTRUMENT
FOR MARITIME AUTONOMOUS SURFACE SHIPS
(MASS), MSC 108/WP.7 C.F.R. (2024).
[12] FORTUNE BUSINESS INSIGHTS. (2024). Autonomous
Ships Market Size, Share, Industry Analysis,
https://www.fortunebusinessinsights.com/industry-
reports/autonomous-ship-market-101797, accessed 02
July 2024.
[13] O’Regan, G., 2012. “Artificial Intelligence. In: A Brief
History of Computing” Springer, London.
https://doi.org/10.1007/978-1-4471-2359
[14] Ilkou E. and Koutraki M., (2020), “Symbolic Vs Sub-
symbolic AI Methods: Friends or Enemies?”. Paper
presented at the CSSA'20: Workshop on Combining
Symbolic and Sub-Symbolic Methods and their
Applications co-located with CIKM2020, https://ceur-
ws.org/Vol-2699/paper06.pdf, accessed 02 August 2024.
[15] Kovalishin, P., Nikitakos, N., Svilicic, B., Zhang, J.,
Nikishin, A., Dalaklis, D., Kharitonov, M. & A.A.
Stefanakou, (2023), Using Artificial Intelligence (AI)
methods for effectively responding to climate change at
marine ports, Journal of International Maritime Safety,
Environmental Affairs Shipping, 7, 1, 2186589. DOI:
10.1080/25725084.2023.2186589
[16] Tsaganos, G., Nikitakos, N., Dalaklis, D., Ölcer, A.I. &
Papachristos, D. (2020). Machine Learning Algorithms in
Shipping: Improving Engine Fault Detection and
Diagnosis via Ensemble Methods, Journal of Maritime
Affairs, 19(1), 51-72.
https://link.springer.com/article/10.1007/s13437-019-
00192-w
[17] UNITED NATIONS CONFERENCE ON TRADE AND
DEVELOPMENT (UNCTAD. (2023). Review of Maritime
Transport 2023. https://unctad.org/system/files/official-
document/rmt2021_en_0.pdf, accessed 03 April 2025.
[18] Munim, Z. H., Dushenko, M., Jimenez, V. J., Shakil, M.
H., & Imset, M. (2020). Big data and artificial intelligence
in the maritime industry: a bibliometric review and future
research directions. Maritime Policy & Management,
47(5), 577-597. DOI: 10.1080/03088839.2020.1788731
[19] Xiao, G., Yang, D., Xu, L., Li, J., & Jiang, Z. (2024). The
application of artificial intelligence technology in
shipping: A bibliometric review. Journal of Marine
Science and Engineering, 12(4), 624.
https://doi.org/10.3390/jmse12040624