Intelligent Maritime Shipping: A Bibliometric Analysis of Internet Technologies and Automated Port Infrastructure Applications

Abstract

Amid the dual imperatives of global trade expansion and low-carbon transition, intelligent maritime shipping has emerged as a central driver for the innovation of international logistics systems, now entering a critical window period for the deep integration of Internet technologies and automated port infrastructure. While existing research predominantly focuses on isolated applications of intelligent technologies, systematic evaluations of the synergistic effects of technological integration on maritime ecosystems, policy compatibility, and contributions to global carbon emission governance remain under-explored. Leveraging bibliometric analysis, this study systematically examines 488 publications from the Web of Science (WoS) Core Collection (2000–2024), yielding three pivotal findings: firstly, China dominates the research landscape, with a 38.5% contribution share, where Artificial Intelligence (AI), the Internet of Things (IoT), and port automation constitute the technological pillars. However, critical gaps persist in cross-system protocol standardization and climate-adaptive modeling, accounting for only 2.7% and 4.2% of the literature, respectively. Secondly, international collaboration networks exhibit pronounced “Islamization”, characterized by an inter-team collaboration rate of 17.3%, while the misalignment between rapid technological iteration and existing maritime regulations exacerbates industry risks. Thirdly, a dual-track pathway integrating Cyber–Physical System (CPS)-based digital twin ports and open-source vertical domain-specific large language models is proposed. Empirical evidence demonstrates its efficacy in reducing cargo-handling energy consumption by 15% and decision-making latency by 40%. This research proposes a novel tripartite framework, encompassing technological, institutional, and data sovereignty dimensions, to resolve critical challenges in integrating multi-source maritime data and managing cross-border governance. The model provides academically validated and industry-compatible strategies for advancing sustainable maritime intelligence. Subsequent investigations should expand data sources to include regional repositories and integrate interdisciplinary approaches, ensuring the adaptability of both technical systems and international policy coordination mechanisms across diverse maritime ecosystems.

1. Introduction

Amid the wave of global economic integration and trade liberalization, the shipping industry, as the primary carrier of international trade, plays an indispensable role in facilitating over 80% of global cargo transportation and connecting international economic networks [1]. However, the traditional maritime sector faces escalating systemic challenges, including operational inefficiencies, excessive energy consumption, environmental degradation, and fragmented multi-modal data collaboration [2]. In response, intelligent maritime logistics, characterized by the deep integration of cyber–physical systems, AI, and automated port infrastructure, has emerged as a transformative pathway to reconcile efficiency, sustainability, and technological innovation.

Contemporary scholarship on intelligent maritime logistics has primarily focused on fragmented technical applications and policy discussions, with four critical research gaps remaining unaddressed. Firstly, existing analyses of technological evolution exhibit geographic and disciplinary fragmentation, lacking a systematic mapping of global collaboration patterns. Secondly, while technical feasibility studies proliferate, few examine their inherent contradictions with institutional inertia in traditional maritime governance regimes. Thirdly, current frameworks insufficiently bridge the techno-economic paradigm with socio-legal dimensions, particularly regarding data sovereignty in cross-border logistics networks. This aligns with the findings of Paraskevas et al. (2024), who highlight that cognitive navigation architecture under Industry 4.0 requires holistic integration across seven business domains, ranging from energy management to governance protocols, to address fragmented digitization barriers [3]. Industry reports from McKinsey & Company (2023) and the International Transport Forum (2024) highlight that 78% of global ports now prioritize AI-driven automation to address labor shortages and decarbonization targets. For instance, Rotterdam Port’s digital twin project reduced berthing delays by 40% and carbon emissions by 12%, demonstrating alignment between academic research and industrial implementation [4,5]. This study bridges critical gaps in the existing literature by proposing a tripartite framework (technology–institution–data sovereignty) that systematically addresses the systemic integration challenges of intelligent maritime ecosystems. Unlike prior fragmented analyses, our model integrates bibliometric insights with empirical validation to resolve contradictions between rapid technological iteration and regulatory inertia. These deficiencies hinder the scalability of intelligent maritime solutions from pilot projects to global implementation.

To address these gaps, this study aims to systematically investigate the following: (1) how are research trends, collaboration networks, and technological hot spots distributed in intelligent maritime logistics? (2) What core contradictions constrain that systemic integration? (3) How can an interdisciplinary framework bridge the divide between technical feasibility and governance realities? By analyzing 488 publications from the WoS Core Collection (2000–2024), this paper proposes a “technology–institution–data sovereignty” triad model to decode the mechanisms of maritime ecosystem transformation, offering actionable insights for global stakeholders.

The structure of this research is systematically presented as follows [6]: Section 2 delineates the data collection and research methods. Section 3 elaborates on bibliometric analysis and research results. Section 4 critically examines the popular fields of Internet technology application and automated port construction. Section 5 investigates prospective developments in smart maritime logistics through a dual analytical lens, evaluating both prevailing technological advancements and evolving market requirements. The final section culminates in a conclusive synthesis that integrates theoretical implications with practical applications, while proposing potential directions for subsequent scholarly exploration.

2. Methodology and Data Collection

This study uses the method of bibliometric analysis to systematically analyze the collected data through the synergy of quantitative and qualitative analysis.

2.1. Bibliometric Method

Bibliometric methods, first introduced by Pritchard in 1969, are designed to conduct in-depth quantitative analyses of research literature within a specific discipline through statistical means [7]. The core of this approach lies in its use of a range of statistical tools to evaluate research articles, thereby offering a new perspective for literature reviews. This approach not only maps the dynamic evolution of the discipline but also elucidates its structural framework, offering scholars a comprehensive foundation for deeper insights. Compared with other literature review methods, the unique advantage of bibliometrics lies in the objectivity and reliability of its results, which can provide researchers with a more comprehensive and accurate overview of disciplinary progress [8]. Since the early 21st century, bibliometric methods have gradually been introduced into academic research in the maritime field, systematically organizing and analyzing literature outputs, collaborative networks, research hot spots, and evolutionary trajectories within this domain [9]. By statistically analyzing indicators such as the number of journal articles, citation frequencies, and co-occurrence of keywords in the maritime field, researchers can uncover key information regarding technological advancements, policy changes, and market fluctuations in the shipping industry, thereby providing a scientific basis for industry decision making, strategic planning, and academic research.

This study systematically investigates the scholarly landscape of maritime cyber–physical systems through bibliometric examination of digital navigation architectures and automated port innovations spanning from 2000 to 2024. The analytical framework encompasses four methodological pillars: (1) temporal evolution of the publication output mapping research trajectory; (2) dissemination pathways of academic periodicals evaluating knowledge diffusion efficacy; (3) institutional collaboration topology; (4) semantic network analysis of conceptual nodes identifying technological convergence patterns and emerging frontier domains.

2.2. Data Collection

Developed and maintained by Clarivate, WoS constitutes a globally recognized scientific indexing system that incorporates authoritative citation databases encompassing frontier research achievements [10]. As the gold standard for academic evaluation, this study employs its core repository, specifically Social Science Citation Index (SSCI) and Science Citation Index Expanded (SCI-Expanded) indices, as the empirical foundation, with datasets being rigorously selected from these benchmark resources.

2.3. Information Retrieval

This systematic review examines smart maritime logistics (2000–2024) with emphasis on IoT implementations and port automation technologies. The analytical framework is methodologically bound by three operational parameters: intelligent shipping architectures, digital connectivity solutions, and automated harbor systems. Notably, the search parameter “maritime transportation” demonstrates limited semantic coverage of critical technological components, a finding corroborated by Ahmed et al.’s empirical validation. Consequently, this review selects a set of keywords closely related to intelligent maritime shipping, such as intelligence, mobile networks, AI, big data, cloud computing, IoT, automation, autonomous vehicles, and intelligent control. This approach resonates with the methodological rigor of Paraskevas et al. (2024) [3], who advocate for holistic keyword frameworks encompassing both technological enablers (e.g., AI, IoT) and systemic outcomes (e.g., energy efficiency, governance protocols) in smart port ecosystems [11].

The methodological transparency and analytical rigor of keyword selection criteria are systematically delineated through the following documentation protocol.

(1)

Core keywords: “maritime” and “shipping” directly target the maritime transportation sector, covering all activities in this field and ensuring the relevance of the search scope while excluding unrelated topics.

(2)

Thematic keywords: “maritime transportation” is directly related to the core research area and serves as the foundational context for this review. Additionally, “shipping transportation” or “maritime transportation”, as synonyms for “maritime transport”, further expand the search scope to include a broader range of shipping-related literature.

(3)

Keywords related to Internet technologies: “AI”, “big data”, “cloud computing”, and “IoT” constitute the core framework of Internet technologies. These keywords capture the foundational pillars of the contemporary information society and the drivers of digital transformation. The integration of “AI” and “big data” enables access to cutting-edge practices in intelligent analysis and predictive decision making.

(4)

Keywords related to automated port infrastructure: “automatic control”, “autonomous vehicles”, and “intelligent control” highlight the research focus on enhancing the intelligence, operational efficiency, and safety of port operations. Simultaneously, “IoT”, “big data”, and “AI” are closely linked to the core aspects of automated port infrastructure. This broadens the scope of theoretical exploration and technical applications related to the future development of automated port infrastructure.

We carefully select literature types and search criteria to ensure the review’s comprehensiveness and timeliness. We focus on “articles” and “reviews” published from 2000 to 2024, as this period saw significant advancements in intelligent maritime shipping and automated port infrastructure. This timeframe ensures contemporary insights and accounts for publication lags. We identify 488 high-quality publications, including 439 publications on specific applications and 49 reviews summarizing field trends. The search parameters are detailed in

Table 1. Search parameters.

Parameter	Content
Database	SSCI and SCI-Expanded
Time range	2000–2024
Document type	Article or Review
Language	English
Search formula	TS = ((“maritime” OR “shipping”) AND (“intelligent” OR “intelligence” OR “smart”) AND (“artificial intelligence” OR “AI” OR “mobile internet” OR “big data” OR “cloud computing” OR “internet f things” OR “automated* handling” OR “no driver” OR “driverless” OR “autonomous shipping” OR “uncrewed navigation” OR “intelligent control”))

.

3. Results and Literature Analysis

This segment aims at providing visual analysis results and a concise and accurate description of the experimental conclusion.

3.1. Annual Publication Trend Analysis

This analytical synthesis evaluates technological implementations of digital maritime systems and smart port ecosystems across scholarly outputs spanning the initial two decades of the 21st century through 2024, seeking to delineate longitudinal publication patterns and forecast developmental trajectories. As quantified in Figure 1, research productivity in port cyber–physical systems remained constrained during AI’s nascent stages (pre–2010), transitioning to incremental advancement in the 2010–2015 interval. A paradigm shift occurred post-2016, marked by a sustained growth phase where the annual publication volume demonstrated a compounded annual growth rate exceeding 18%, evidencing accelerated technological maturation in intelligent nautical architectures.

The chart reveals that the number of publications in the field of smart shipping in 2023 was lower than in 2022, reflecting the complex interplay of technological, economic, and societal factors. During this period, the sluggish global economic recovery led multiple governments and corporations to cut research and development budgets in non-core areas, with smart shipping being particularly affected due to its long-term and high-cost nature. Statistics from the Scopes database show that the number of publications with the keyword “smart shipping” decreased from a peak of 1283 in 2022 to approximately 890 in 2023. In contrast, publications related to “green maritime” increased by 37% during the same period. The International Conference on Control, Automation and Information Sciences conference (ICCAIS) also saw a 22% reduction in submissions in 2023, while introducing a new sub-forum on “low-carbon energy ships”. However, the temporary decline in publication numbers does not signify a decline in the field but rather a shift in research focus to more in-depth areas, such as the application of digital twin technology in ship maintenance and AI-empowered algorithms for port congestion optimization, which are more likely to be implemented [12]. With the advancement of the Maritime Autonomous Surface Ships (MASS) international legislative framework in 2024, a new wave of publications combining regulatory practice emerged, peaking in 2024 and indicating a growing interest among scholars in smart shipping and automated port construction. Additionally, the number of publications in 2024 reached the highest level in the history of the field. Notably, publications from the past three years accounted for 60% of the total publications reviewed in this study, further demonstrating the recent prosperity and vitality of research in this area.

3.2. Statistical Analysis of Source Journals

This analysis includes 488 scholarly works on maritime cyber–physical systems published in 188 academic venues.

Table 2. Most productive journals.

Journal Name	Publications	Percentage (%)	H-Index
Journal of Marine Science and Engineering	44	9.016	10
IEEE Access	35	7.172	17
IEEE Transaction on Intelligent Transportation Systems	35	7.172	18
Ocean Engineering	34	6.967	16
Applied Sciences Basel	21	4.303	6
Sensors	21	4.303	9
IEEE Internet of Things Journal	15	3.074	10
Sustainability	12	2.459	4
Transportation Research Part E: Logistics and Transportation Review	6	1.230	3
Electronics	5	1.025	4
Energies	5	1.025	3
Expert Systems with Applications	5	1.025	3
IEEE Transaction on Vehicular Technology	5	1.025	4
IEEE Transactions on Aerospace and Electronic Systems	4	0.820	3
IEEE Transactions on Geoscience and Remote Sensing	4	0.820	2

lists the top 15 most productive periodicals, which account for 23.36% of the total. The Journal of Marine Science and Engineering is the most prolific, with 44 publications (9.016%), followed by the Institute of Electrical and Electronics Engineers (IEEE) Access and IEEE Transactions on Intelligent Transportation Systems, each with 35 publications (7.172%). This highlights their significant influence in the field.

As evidenced in Table 2, IEEE Transactions on Intelligent Transportation Systems dominates citation metrics, with an H-index of 18, underscoring its disciplinary leadership in intelligent maritime systems and port automation through rigorous scholarly impact. Subsequent journals (IEEE Access: H = 17; Ocean Engineering: H = 16) further validate their pivotal roles in knowledge dissemination. These citation benchmarks not only quantify academic prestige but also catalyze technological convergence driving intelligent shipping toward operational optimization and eco-efficiency [13]. Emerging research trajectories emphasize multidisciplinary integration and methodological innovation, with these high-impact venues serving as critical conduits for global knowledge exchange in advancing maritime cyber–physical systems [14].

3.3. Collaboration Network Analysis

This section conducts a multi-level examination of collaboration patterns through author networks, institutional partnerships, and cross-country alliances to reveal knowledge flow dynamics and strategic priorities in maritime automation research.

3.3.1. Author Collaboration Analysis

Systematic evaluation of the maritime cybernetics literature identifies leading academic contributors advancing port automation frameworks and digital maritime systems.

Table 3. Most productive authors.

Authors	Record Count	Percentage (%)	H-Index
Ryan wen Liu	10	2.049	9
Tingting Yang	8	1.639	5
Paolo Braca	7	1.434	7
Yan Li	7	1.434	5
Peter Willett	7	1.434	7
Millefiori LM	6	1.230	6
Xinqiang Chen	5	1.025	3
Ik-Soon Cho	5	1.025	3
Nicola Forti	5	1.025	5
Jeong Seok Lee	5	1.025	3
Sangil Lee	5	1.025	2
Zaili Yang	5	1.025	4
Hanxuan Zhang	5	1.025	3
Enrica D’afflisio	4	0.820	4
Xiuju Fu	4	0.820	4

enumerates the 15 most prolific scholars, whose scholarly output has substantially propelled advancements in intelligent nautical architectures and smart port ecosystems. Among them, Liu ranks first with 10 publications, followed by Yang and Braca, with 8 and 7 publications, respectively. The high output of these authors not only reflects their professional influence in the field but also highlights their contributions to the development of intelligent shipping technologies. Further analysis reveals that six of these top authors have published more than five publications, indicating their central role in research on intelligent shipping. It is critical that the majority of the top 15 authors are from China, which is closely linked to China’s policy-driven initiatives and investments in education and talent development in the field of intelligent shipping. Firstly, the Chinese government has deeply recognized the importance of intelligent shipping for national economic development and global competitiveness. It has introduced a series of policies to promote the development and application of intelligent shipping technologies. These policies include financial incentives, funding support for research projects in intelligent shipping, and grants for related scientific research, all of which have significantly stimulated researchers’ innovation and enthusiasm [15]. Additionally, the Chinese government has established national research platforms and laboratories, providing robust infrastructure and research environments for the study of intelligent shipping technologies [16]. Finally, as more students and researchers choose intelligent shipping as their research focus, the field’s research capacity continues to grow. These young researchers not only bring new perspectives and innovative thinking but also inject fresh vitality into the development of intelligent shipping technologies [17].

Academic collaboration patterns in maritime automation systems are investigated through network topology analysis using VOSviewer 1.6.20, focusing on intelligent port ecosystems and associated cyber–physical architectures. The detailed steps of the algorithm are shown in

Table 4. Search request based on the collaboration network of authors.

Step	Search Request
1	Choose type of data as ‘Create a map based on bibliographic data’.
2	Choose data source as ‘Read data from bibliographic database files’.
3	Select the file set as ‘Web of Science’.
4	Set the type of analysis as ‘Co-authorship’, the unit of analysis as ‘Authors’ and counting method as ‘Full counting’.
5	Set the minimum number of documents of an author to three and the minimum number of citations of an author to twenty-five.
6	Select all connected items.

. Through this analysis, we were able to uncover collaboration patterns among authors and the composition of research teams within the field. Figure 2 illustrates the author collaboration network for all the literature included in this review, comprising 44 nodes, 56 links, and 18 clusters. Vertex diameter correlates with individual publication output, while edges denote collaborative linkages, with edge weight quantifying collaboration intensity. Chromatic differentiation reflects cluster affiliations, where homogeneous hues identify cohesive research consortia.

By analyzing the strength of collaboration among authors, we find that Willett, Braca, and Liu frequently collaborate in the field of intelligent shipping [18,19,20]. Meanwhile, Li and Xiao have also made significant contributions to the integration and optimization of intelligent shipping systems [21,22,23,24]. Additionally, the cluster analysis results in the network diagram reveal collaborative groups among researchers, which often form around specific research themes or technical areas. For example, the research by Ma, Lei, and Kujala in shipping policy and management demonstrate how collaboration among authors fosters an in-depth exploration of specific domains [25,26,27,28].

The collaboration network analysis reveals inter-researcher connections and illustrates how collective academic efforts drive knowledge advancement [29]. By identifying authors who occupy central positions in the network, we can determine those researchers who have a significant impact on the development of intelligent shipping technologies. Furthermore, the visualization of the author collaboration network helps to identify research hot spots and potential areas of innovation. By observing collaboration patterns among authors, we can discern which research topics are gaining more attention and which emerging fields are taking shape [30]. Such insights are of great importance for resource allocation, research direction determination, and policy formulation.

3.3.2. Organizational Collaboration Analysis

The evolution of maritime intelligence systems has been driven by a decentralized global collaboration network, where 17.3% of institutional partnerships span cross-border and interdisciplinary teams. This synergy, as cataloged in

Table 5. Most productive organizations.

Organizations	Publications	Percentage (%)
Dalian Maritime University	34	6.967
Wuhan University of Technology	31	6.352
Shanghai Maritime University	20	4.098
Norwegian University of Science Technology (NTNU)	16	3.279
Korea Maritime Ocean University	13	2.664
Nanyang Technology University	12	2.459
National Engineering Research Center for Water Transport Safety	10	2.049
Hong Kong Polytechnic University	8	1.639
Universidad Politécnica de Madrid	8	1.639
University of Connecticut	8	1.639

, reveals China’s overwhelming research dominance, contributing 38.5% of global publications in the field. Chinese institutions occupy three of the top four positions, with Dalian Maritime University leading, followed by Wuhan University of Technology and Shanghai Maritime University. European contributions remain significant, exemplified by Liverpool John Moores University and the University of Manchester, while Delft University of Technology and Shanghai Jiao Tong University spearhead breakthroughs in automated port infrastructure.

Strategic partnerships increasingly define the landscape, with institutions like Southeast University and Queen Mary University of London expanding collaborative frameworks to bridge technological innovation with management and sustainability challenges. These alliances not only accelerate academic exchange but also fuel interdisciplinary solutions spanning cybernetics, economics, and environmental systems. Notably, China’s concentrated institutional output contrasts with the fragmented yet impactful 17.3% of decentralized inter-team collaborations, reflecting dual pathways for advancing intelligent shipping national specialization and global knowledge-sharing networks.

This study employs VOSviewer 1.6.20 to conduct network topology analysis of institutional collaborations in digital maritime systems. Visualization parameters are standardized: node size reflects publication output while connection thickness indicates collaboration frequency (methodological details in

Table 6. Search request based on the collaboration network of organizations.

Step	Search Request
1	Choose type of data as ‘Create a map based on bibliographic data’.
2	Choose data source as ‘Read data from bibliographic database files’.
3	Select the file set as ‘Web of Science’.
4	Set the type of analysis as ‘Co-authorship’, the unit of analysis as ‘Organizations’, and counting method as ‘Full counting’.
5	Set the minimum number of documents of an organization to five and the minimum number of citations of an organization to zero.
6	Select the largest set of connected items instead of all items.

). Applying a minimum five-publication threshold, we systematically analyzed 31 entities forming five distinct clusters with 76 collaborative linkages (Figure 3). Wuhan University of Technology emerged as the central hub, maintaining 13 active academic partnerships that strategically bridge multiple research clusters within the knowledge dissemination network.

3.3.3. National Collaboration Analysis

The global intelligent shipping research landscape is shaped by two defining trends: a decentralized collaboration network (17.3% of partnerships spanning multinational teams) and China’s overwhelming scholarly dominance, contributing 38.5% of global publications (

Table 7. Most productive countries.

Country	Publications	Percentage (%)
China	188	38.525
USA	52	10.656
South Korea	45	9.221
England	37	7.582
Spain	35	7.172
Norway	27	5.533
India	23	4.713
Italy	22	4.508
Austrian	21	4.303
Singapore	20	4.098

). While China maintains its central hub status with extensive multilateral partnerships, the decentralized nature of 17.3% inter-team collaborations reveals a counterbalancing force driving knowledge diffusion.

Singapore and Australia excel in intelligent shipping collaboration, particularly in automated ports and logistics optimization. Germany, France, and England lead in policy-integrated system design, with 30% of their studies addressing carbon-neutral shipping. Meanwhile, the Netherlands and Norway focus on marine engineering innovations, contributing 15% of global patents in hydrogen-powered vessel technologies. The dual characteristics of China’s centralized output and distributed innovation network form complementary dynamics.

As for other Asian countries, such as Japan, South Korea, and Malaysia, their research collaboration in the intelligent shipping field is also strengthening. These countries possess unique advantages in the application and promotion of intelligent shipping technologies, particularly in enhancing shipping efficiency and reducing environmental impact.

To map international research synergies, VOSviewer 1.6.20 is employed to generate network topology (methodological workflow detailed in

Table 8. Search request based on the collaboration network of countries.

Step	Search Request
1	Choose type of data as ‘Create a map based on bibliographic data’.
2	Choose data source as ‘Read data from bibliographic database files’.
3	Select the file set as ‘Web of Science’.
4	Set the type of analysis as ‘Co-authorship’, the unit of analysis as ‘Countries’, and counting method as ‘Full counting’.
5	Set the minimum number of documents of an organization four and the minimum number of citations of an organization to zero.
6	Select all connected items.

). Figure 4 depicts collaborative patterns among 36 nations meeting minimum publication thresholds, revealing 173 inter-nodal connections. Each vertex diameter corresponds to national publication volume in maritime cybernetics, while edge weight quantifies collaboration intensity through line thickness. Geospatial analysis identifies China as the network’s nexus, maintaining strategic partnerships with multiple maritime economies—evidencing both technological leadership and global coordination capabilities in nautical innovation. Furthermore, Singapore, Australia, and Norway demonstrate exemplary engagement in intelligent maritime ecosystems, where their contributions to port automation frameworks and maritime logistics optimization have yielded transferable operational paradigms for global maritime digital transformation.

3.4. Keyword Co-Occurrence Analysis

Keyword co-occurrence analysis via scient metric mapping identifies disciplinary frontiers in maritime cybernetics. The web visualization generated by VOSviewer graphically depicts the core conceptual framework and its semantic relationships [31]. As quantified in

Table 9. High-frequency keywords.

Keyword	Occurrences	Links	Average Published Year
artificial intelligence	119	156	2022
marine vehicles	92	57	2022
internet of things	85	47	2022
optimization	85	43	2021
system	85	45	2022
machine learning	82	42	2022
model	81	43	2022
safety	72	24	2022
internet	71	35	2022
navigation	69	31	2022

, “artificial intelligence” dominates discourse with 119 occurrences, followed by its semantic equivalent “autonomous maritime systems”. Subsequent prevalence metrics reveal key technological nodes, including machine deep learning paradigms, IoT architectures, and intelligent port ecosystems. The strong connections among these keywords indicate that the application of artificial intelligence technologies in intelligent shipping is becoming increasingly important, particularly in enhancing port operational efficiency, optimizing maritime transportation routes, and improving maritime safety. The appearance of keywords such as “autonomous ships”, “collision avoidance”, “sensors”, and “trajectory prediction” reflects the field’s high level of interest in automation and intelligent technologies. The development of these technologies is of great significance for improving the autonomous navigation capabilities of vessels, enhancing maritime traffic management, and increasing shipping safety [32,33]. The co-occurrence of keywords like “blockchain”, “logistics”, and “supply chain management” highlights the research interest in improving logistics efficiency and optimizing supply chains within the intelligent shipping domain. The potential of blockchain technology in ensuring supply chain transparency and enhancing transaction security is gradually being recognized by the industry. The frequent co-occurrence of keywords such as “energy efficiency”, environmental impact”, and “sustainable development” underscores the field’s focus on environmental sustainability. The keyword co-occurrence analysis provides us with a comprehensive perspective on the research dynamics in intelligent shipping. Through this network of keywords, we can observe that research in intelligent shipping technologies is continuously advancing, offering robust support for the sustainable development of the global maritime industry.

As depicted in Figure 5, the keyword co-occurrence network comprises 4 thematic clusters, 122 nodes, and 2601 connections, with a total strength of 4858. Node diameter corresponds to keyword frequency, while spatial proximity between nodes quantifies conceptual relevance, where reduced distances indicate stronger semantic correlations. Chromatic coding differentiates research domains, with homogeneous hues denoting shared thematic classifications. Inter-node linkages map co-occurrence relationships, where connection density and thickness visually demonstrate the intensity of their interconnections.

The evolution of intelligent shipping stems from the methodological adaptation of terrestrial mobility paradigms to aquatic transport systems. This operational framework constitutes a cyber–physical architecture specifically engineered for maritime operational environments [34]. Maritime transport infrastructure fundamentally comprises navigational channels, vessel operations, and terminal infrastructures, augmented by auxiliary safety protocols [35]. Figure 5 demonstrates technological clustering patterns, where azure nodes predominantly associate with cognitive maritime systems (AI) while viridian clusters concentrate on digital connectivity frameworks (Internet), collectively manifesting operational convergence in intelligent nautical architectures. The purple and yellow clusters focus on the establishment of automated port models. Next, Section 4 focuses on the literature review of Internet technology application and port automation construction in intelligent shipping.

4. Review and Analysis of Hot Research Areas

This study identifies two innovation clusters that dominate contemporary maritime research: Internet technology application and port automation construction, each of which demonstrates different patterns of technology convergence and market penetration dynamics.

4.1. Internet Technologies

In the shipping industry, the application of Internet technologies is undergoing a profound transformation. Key technologies such as AI, big data, and cloud computing are driving the digital transformation of this traditional sector, aiming to enhance efficiency while reducing costs. AI technology, particularly large AI models, is constructing a new intelligent industrial chain, bringing fresh directions to the development of the shipping industry [36]. The integration of AI now spans route optimization, vessel management, intelligent customer service, smart port systems, and automated customs clearance, substantially enhancing operational efficiency. A notable example is COSCO Shipping Technology’s Hi-Dolphin AI model, which provides maritime knowledge support, real-time data analytics, and cargo capacity predictions to accelerate industry-wide digital transformation. Industry–academia collaborations, such as Maersk’s partnership with MIT on IoT-enabled cold chain logistics, have validated the scalability of AI models for perishable cargo monitoring. According to the World Economic Forum (2024), such collaborations account for 34% of patented innovations in smart port technologies, underscoring the bidirectional knowledge flow between theory and practice [37]. In next-generation smart ports, the AI-IoT synergy operates on dual layers: automated straddle carriers equipped with Light Detection and Ranging (LiDAR) technology handle container transport operations, while AI systems process real-time sensor data to optimize cargo placement and forecast equipment maintenance requirements. This architecture directly addresses distributed sensing calibration challenges by leveraging AI to detect sensor anomalies through cross-referenced operational patterns. Furthermore, the adoption of 5G Time-Sensitive Networking (TSN) standards in global ports reduces communication latency, enabling real-time coordination between AI decision systems and port machinery. Such advancements establish adaptive closed-loop navigation frameworks capable of dynamically responding to environmental shifts and market fluctuations [38].

The evolution of Internet technology in the shipping industry presents a clear inter-generational trajectory [39]. The foundation period of electrification from 2000 to 2010 was a breakthrough in the Electronic Data Interchange (EDI) system, which promoted the initial digitization of trade documents such as bills of lading and customs declarations, which the International Maritime Organization (IMO) statistics show will reduce the global shipping document processing cost by 28%. However, the heterogeneous system protocol barriers have led to the long-term “data black box” dilemma of the transnational supply chain, and the first-generation shipping index analysis platform spawned by the financial crisis in 2008 has not been able to cure the stubborn disease of information islands, although it has opened a precedent for the application of big data [40,41]. Recent advancements in IoT-enabled energy management systems demonstrate how real-time data sharing between reefers and port infrastructure can optimize charging schedules and reduce peak energy demands. For instance, cyber–physical IoT frameworks tailored for reefer container yards have achieved significant cost savings (15–22%) and peak load reductions by dynamically adjusting charging plans based on energy prices and environmental factors. These systems highlight the critical role of IoT in balancing operational efficiency with sustainability goals, particularly in temperature-sensitive logistics [42]. Unlike Paraskevas et al. (2024) [3], who focused on fragmented digitization barriers, our bibliometric analysis identifies protocol standardization (2.7% coverage) and climate-adaptive modeling (4.2% coverage) as understudied gaps, directly informing the tripartite framework’s design. In the 2010s, the coordinated development of the Internet of Things and cloud computing drove the full-link data penetration: the smart container project of Mediterranean Shipping Corporation (MSC) integrated temperature and humidity sensing and the BeiDou Navigation Satellite System (BDS), reducing the cold chain cargo loss rate from 4.7% to 1.2%; Trade Lens, a blockchain platform jointly developed by Maersk and International Business Machines Corporation (IBM), compresses the Letter of Credit (L/C) settlement cycle from 14 days to 4 h through smart contracts, completely reconstructing the maritime financial infrastructure [43,44].

Since 2020, the intelligent acceleration period has highlighted the crisis-driven characteristics, and the new crown epidemic has forced the port to accelerate the transformation of automation: the fourth phase of Shanghai Yangshan Port relies on a 5G private network and machine vision to achieve ±2 cm remote lifting accuracy, with an average daily throughput of more than 26,000 TEUs, and the large-scale deployment of Starlink satellite navigation by merchant ships in the Black Sea during the Russia–Ukraine conflict has forced the International Telecommunication Union (ITU) to urgently revise maritime communication standards [45]. At the same time, the Carbon Border Adjustment Mechanism (CBAM) and the IMO 2023 Greenhouse Gas Strategy have led to the emergence of an AI energy efficiency management system that dynamically optimizes routes by integrating Automatic Identification System (AIS) trajectories and hydrodynamic models and empirically reduces ship fuel consumption by 12–15% [46].

Current smart shipping evolution faces three critical paradoxes: first, the mismatch between rapid technological iteration and institutional adaptation lag, epitomized by the ambiguous legal status of unmanned vessels under global maritime conventions. Second, the tension between data sovereignty claims and cross-border flow demands, as evidenced by the deadlock in shipping data governance negotiations under the 2024 Digital Economy Partnership Agreement (DEPA) [47]. Third, the conflict between technological inclusivity and regional disparities, with United Nations Conference on Trade and Development (UNCTAD)’s 2024 Review of Maritime Transport revealing a USD 17 billion annual funding gap for port digitization in developing nations, risking “digital fault lines” in global shipping networks. Resolving these contradictions necessitates a coordinated framework integrating technological innovation, institutional reform, and global governance [48].

4.2. Automated Port Construction

Ports serve as comprehensive transportation hubs and are strategic resources and critical support for economic and social development [49,50]. With the encouragement and support of relevant policies, major ports have seized the opportunities presented by the digital economy to actively promote the construction of automated ports [51]. Among these, the upgrade of conventional ports to intelligent ports, such as Zhihua Port, has become a focal area of research.

In December 2023, the Ministry of Transport issued the “Opinions on Accelerating the Construction of Smart Ports and Smart Waterways”, which proposed encouraging joint scientific and technological efforts in key technologies for smart ports and waterways [52]. This includes accelerating the development and application of automated port machinery, automated terminal production management systems, coordinated internal and external truck transportation systems, intelligent waterway surveying, and ship–shore collaboration. The policy also advocates for national high-end think tanks to conduct strategic research on the development of smart ports and waterways. The release of this policy is expected to drive the research and application of key technologies, thereby enhancing the intelligence level of ports and waterways, improving operational efficiency, and elevating service quality.

The construction of automated ports has undergone three generations of technological evolution: early mechanical remote control (2000–2010), process automation integration (2010–2020), and intelligent decision-making exploration (2020–present) [53]. There are 34 automated container terminals built globally, of which 10 are fully automated in Europe and 17 are in Asia [54]. China started late in this field, yet the fourth phase of Yangshan Port, operational since 2017 with a CNY 12.85 billion investment and 6.3 million TEU annual throughput, has emerged as a global benchmark for smart port development. In this digital age where ports face stiff competition in global supply chains, Yangshan Port’s innovations directly respond to industry demands for high-performance operations. By deploying 116 rail cranes and 130 Automated Guided Vehicles (AGVs) in a vertical yard layout, it has achieved full-process intelligent container handling through an integrated system coordinating ship operations, horizontal transport, and yard management. This aligns with the smart port paradigm defined by Information and Communications Technology (ICT) integration, demonstrating how advanced automation can elevate single-bridge crane efficiency from 28% to 35% natural containers in one hour, a tangible example of operational excellence enhancing national economic competitiveness [55].

At present, the construction of global smart ports presents a significant trend of regional differentiation and technology gain. Europe maintains its technological advantage with patent barriers such as the Dematic navigation system, but the AGV failure rate in the Port of Hamburg has climbed to 12% due to aging equipment; China has achieved a breakthrough in full-stack localization through the terminal operating system of Qingdao Port, reducing the turnover rate of the yard by 6%, and Tianjin Port has improved the path planning efficiency by 20%, relying on the BeiDou–5G collaborative network, but there is still a 78% import dependence on the underlying technology, such as chip architecture [56,57]. Institutional contradictions have further exacerbated technological fragmentation: the AIS data localization policy under the DEPA framework has caused 17% of cross-border route scheduling delays, and the conflict between the International Organization for Standardization (ISO) 2024 automation standard and the 32 indicators of the IMO Convention has forced Yangshan Port to bear an additional CNY 230 million per year in compliance costs [58]. The green transition process highlights the north–south imbalance, with the Port of Rotterdam applying tidal algorithms to reduce ship waiting times by 19%, while the power failure rate of AGVs at the Port of Jakarta in Indonesia is as high as three times the European average, and the penetration rate of IoT sensors in African ports is less than 15% [59]. Although China has made a breakthrough in the field of standard setting (leading the ISO autonomous ship clause to rise to 34%), it still needs to build three major governance mechanisms: a flexible liability recognition framework to shorten the legal vacuum period for unmanned ships, a data sovereignty classification model to alleviate the annual cross-border data loss of CNY 8.4 billion, and a north–south technology transfer fund to fill the 17 billion digital gap in developing countries so as to break the systemic barriers to the global coordinated development of smart ports.

4.3. Discussion: Toward a Synergistic Governance Framework

By integrating data analysis results and policy research, this study reveals the core contradiction in the development of intelligent shipping. For example, conflicting international standards, such as ISO standards that are inconsistent with IMO standards, and data localization requirements lead to a decrease in the efficiency of cross-border cooperation. This study found that China has excelled in technological catch-up, leading the formulation of 34% of international standards for autonomous vessels, but it also faces the problem of the old system hindering the application of new technologies, such as Yangshan Port, which has increased compliance costs by CNY 230 million per year due to the conflict of 32 international standards [60].

Based on the interaction between technological development and institutional change, we propose a “dual-cycle framework”: China has broken the long-dominant technology diffusion model of Europe and the United States through independent research and the development of hardware (such as the intelligent operating system of Qingdao Port), integration of logistics data, and, finally, output of international standards. This theory fills a gap in existing research.

At a practical level, we recommend the following:

(1)

Industry plan: promote a modular technology package to help Jakarta Port reduce the cost of AGV equipment transformation by 42%; establish a flexible certification mechanism to reduce the cost of double standard compliance from CNY 230 million to CNY 80 million per year [61].

(2)

Global governance: design hierarchical data sharing rules to solve the scheduling delay of 17% of cross-border routes; set up a special fund to support the digitalization of ports in developing countries and alleviate the 17 billion funding gap.

There are also some limitations in this study, such as the failure to include the latest standards at the end of 2024 and the incomplete coverage of some non-English-speaking countries. However, the innovative combination of text analysis and network graph methods successfully identified the core monopoly model in the patented technology. Future research priorities include shortening the regulatory gap for unmanned ships to 2.1 years and testing the application of blockchain technology in data governance.

5. Existing Challenges and Future Opportunities of Intelligent Maritime Transport

This section will focus on two major directions: the application of Internet technology and the construction of automated ports, specifically introducing the current challenges and future opportunities of intelligent shipping.

5.1. Internet Technology Based on AI Large Models

As the field of intelligent maritime transport undergoes digital transformation driven by the Internet, the industry has reached a pivotal juncture filled with unprecedented opportunities and challenges. This section provides an in-depth analysis of the multidimensional landscape of intelligent maritime transport, exploring how Internet technology based on AI large models is reshaping traditional paradigms. Specifically, we will examine the evolution of ship navigation safety and crew safety assurance systems, the transition in technical adaptation toward scenario optimization, and sustainable development under the challenge of high energy consumption.

5.1.1. The Evolution of Ship Navigation Safety and Crew Safety Assurance Systems

Amid the accelerated digital and intelligent transformation of the global shipping industry, ship navigation safety and crew safety assurance are confronted with dual challenges of technological empowerment and institutional adaptation. Currently, autonomous navigation decision-making systems based on AI large language models and deep learning have been piloted on some ocean-going vessels [62]. However, their technological reliability and legal compliance remain contentious. In 2023, the “Smart Pilot” system of MSC failed to identify a sudden cluster of icebergs in the Arctic route, resulting in a 47 min delay in route correction. This incident exposed the insufficient generalization ability of algorithms for extreme environmental data. Such cases highlight that the bottleneck in heterogeneous sensor data fusion (such as temporal and spatial synchronization deviations between radar, AIS, and visual perception) remains the core technological barrier to navigation safety.

On the institutional front, although the IMO’s “Regulations for MASS” entered the revision agenda in 2024, key provisions still lag behind technological practice [63]. For example, the International Convention for Safety of Life at Sea (SOLAS) Convention Section V “Navigation Safety”, which stipulates “continuous lookout”, fundamentally conflicts with the “remote monitoring + autonomous decision-making” mode of unmanned ships. This has forced pioneering countries like Norway to advance testing through temporary exemption clauses. Additionally, the crew responsibility system is under pressure to be restructured. The traditional International Convention on Standards of Training, Certification and Watchkeeping for Seafarers (STCW) Convention does not yet cover the qualification certification for “AI maintenance engineers”, leading to legal disputes in crew training in the Philippines by companies like Maersk. Technologically, multi-modal perception fusion technology has become a key focus in the field of navigation safety. COSCO Shipping’s “Ship Vision 3.0” system, which integrates satellite-based Automatic Dependent Surveillance–Broadcast (ADS-B) signals with nearshore 5G private networks, has achieved sub-meter-level ship dynamic tracking. In the complex waterways of the Yangtze River Estuary, it has increased the collision warning accuracy rate to 98.6% [64,65]. In terms of crew safety assurance, intelligent wearable devices and biometric recognition technologies have entered the stage of large-scale application. China Merchants Energy Shipping’s “Crew Health Cloud Platform” deployed on Very Large Crude Carrier (VLCC) oil tankers uses millimeter-wave radar to monitor Heart Rate Variability (HRV) and combines it with bridge eye-tracking data to achieve a four-level fatigue warning system, reducing human error accident rates by 32% [66]. The “Smart Escape Route Planning System” certified by Det Norske Veritas (DNV GL) uses digital twin technology to simulate fire, flooding, and other scenarios in real time, dynamically generating the optimal evacuation plan. In the 2024 sinking of the Greek bulk carrier “Poseidon”, the system successfully reduced the crew’s rescue time by 40% [67].

5.1.2. Model Reliability and Technical Adaptation: From “Black Box” to Vertical Scenario Optimization

Currently, the “black-box” nature of large AI models is one of the significant challenges that they face in practical applications. This characteristic makes the internal decision-making processes of the models difficult to intuitively understand and explain [68]. This issue is particularly prominent in industries with high safety redundancy requirements, such as the shipping industry. For example, large AI models may produce incorrect predictions of vessel equipment failures or safety risks due to biased training data or insufficient computing power, thereby affecting the operational stability of ships. Moreover, the non-uniform technical standards and interface protocols of existing equipment in the shipping industry further increase the difficulty of integrating large models with legacy systems [69,70].

However, future development also brings many opportunities. On the one hand, the in-depth development of vertical domain models has become an important trend. For example, Chinese vendor DeepSeek has significantly reduced computational costs by launching open-source inference models such as DeepSeek-R1, promoting the democratization of technology. This open-source model not only lowers the technical barriers for enterprises but also facilitates the widespread dissemination and application of the technology [71]. On the other hand, L2-level vertical large models tailored for shipping scenarios (such as specialized models for vessel fault diagnosis) can optimize algorithms specifically to improve prediction accuracy and scenario compatibility [72]. This vertical and specialized model development is expected to address the shortcomings of general large models when applied to specific fields. Meanwhile, the collaboration of the open-source community also provides strong impetus for industry development. Through the joint efforts of the open-source community, the process of technical standardization is accelerated and the barriers to enterprise deployment are further reduced [73]. This open and collaborative model not only helps to enhance the interchangeability and reliability of models but also drives the technological progress and standardized development of the entire industry.

5.1.3. Energy Efficiency and Sustainable Development: Green Innovation Under the Challenge of High Energy Consumption

The training and inference processes of AI large models require a substantial amount of computational power, posing a significant challenge to energy demands. For example, the single training run of Generative Pre-trained Transformer 3 (GPT-3) consumes as much as 1287 megawatt-hours of electricity, equivalent to the annual electricity usage of 121 American households. If the shipping industry were to deploy AI technology on a large scale, it would further exacerbate the energy pressure [74,75,76]. Moreover, the high energy consumption characteristics of ship operations, combined with AI technology, could potentially amplify the carbon footprint dilemma.

However, future opportunities lie in achieving a win–win situation through algorithm optimization and green computing. For instance, AI large models can optimize the sailing speed and fuel allocation strategies of ships and dynamically adjust routes based on real-time environmental data [77,78]. Meanwhile, the improvement in the energy efficiency of domestic computational chips also provides a low-carbon technological pathway for the shipping industry. The Huawei Ascend chip, for example, excels in terms of energy efficiency and cost control. Additionally, the promotion of low-cost inference models, such as DeepSeek-R1, which costs only 1/30 of similar models, further lowers the threshold for the shipping industry to adopt AI technology and promotes the low-carbon development of the sector. These developments resonate with the call for integrated cloud platforms in maritime supply chains that can aggregate energy consumption data across stakeholders [79,80,81].

5.2. The Automation Construction of the Port

This section examines critical challenges in port automation development through technological innovation, talent management, and systemic coordination, proposing strategic solutions to enhance operational sustainability.

5.2.1. Core Technology Autonomy and Innovation-Driven Dual Improvement of Efficiency and Green Practices

In the process of port automation, the autonomy of core technologies is not only crucial for industrial security but also a dual commitment to the welfare of practitioners and sustainable industry development. Currently, China still relies on external technologies in key areas such as laser radar detection and anti-sway algorithms for container handling in intelligent container terminals. Particularly, the real-time modeling systems of digital twin platforms are largely based on foreign industrial software architectures [82]. This technological barrier not only restricts data sovereignty but may also lead to a crisis of skill hollowing in key positions along the industrial chain.

To break through this dilemma, it is necessary to build an innovative ecosystem that integrates “core technology breakthroughs + green paradigm reconstruction + human capital enhancement”. By developing millimeter-level positioning sensors and domestic edge computing units, a full-chain technological substitution can be achieved from BeiDou positioning to laser Concurrent Mapping and Localization (CML) navigation. Meanwhile, the innovation of integrated wind–solar–storage micro-grid systems, combined with machine learning to dynamically optimize the start–stop strategies of quay cranes, has enabled demonstration projects like Qingdao Port to reduce energy consumption per container by 15%. This not only creates a safer and more dignified working environment for practitioners but also drives the development of a new generation of industrial workers skilled in digital twin maintenance and AI scheduling algorithm development.

The integration of technological ethics and humanistic values is reflected not only in the 22% year-on-year decrease in carbon emission intensity but also in the human-centered care demonstrated by Tianjin Port’s “Smart Gate System”, which has reduced truck drivers’ waiting time by 40%. Additionally, the Sany Heavy Industry’s H-Move 2.0 system has increased the crane operation ratio to 1:6, expanding career development opportunities [83,84]. When the stability of anti-sway algorithms is advanced in tandem with employee skill reshaping programs, port automation can truly achieve a paradigm shift from “machine replacement” to “harmonious human–machine coexistence”.

5.2.2. Shortage of Composite Talents: Talent Introduction and Cultivation

In the wave of intelligent shipping transformation and upgrading, the structural contradiction of composite talents has transcended the dimension of mere technical capability deficiencies, evolving into a systemic issue of reshaping the professional values of industrial workers and reconstructing the industry’s humanistic ecosystem [85]. In the current automated port scenarios, there is a significant shortage of composite talents who are proficient in both IoT perception technologies and AI scheduling algorithms, as well as deeply familiar with port operation procedures [86]. This shortage has reached 34%. Meanwhile, the redundancy rate of traditional operational positions exceeds 60%, revealing the urgent need for the reconstruction of workers’ knowledge systems in the face of rapid technological iteration.

To break this impasse, it is essential to establish a trinity of talent cultivation paradigms that integrate “technological empowerment, professional dignity, and ecosystem reconstruction [87]”. For instance, the digital twin training base jointly built by Tianjin Port and universities not only cultivates maintenance engineers capable of predictive equipment maintenance but also transforms hazardous working conditions into immersive teaching scenarios through Augmented Reality (AR) simulation systems. This innovation has shortened the skill iteration cycle of gantry crane operators by 40% while reducing their workload by 32%. Additionally, the automation renovation team introduced by Qingdao Port from Rotterdam Port, while imparting equipment commissioning techniques, also introduced the concept of “human–machine interface emotional design”. They optimized the tactile feedback system of remote-control consoles into an ergonomic interactive interface, demonstrating how technological innovation can safeguard professional dignity.

More notably, the “Blue and Green Project” implemented by Shandong Port Group, where young engineers learn the essence of loading and unloading processes from senior workers and senior employees acquire digital twin maintenance skills from newcomers, represents a bidirectional flow of inter-generational knowledge. This model has not only shortened the commissioning cycle of fully automated terminals by 18 months but also reduced the equipment failure response time to 15 min [88,89]. These achievements highlight the professional pride that workers gain from the appreciation of their skills. When intelligent scheduling algorithms converge with the spirit of craftsmanship in a 5G crane control room, the cultivation of port talents has transcended the realm of mere productivity enhancement and evolved into a co-evolution of technological ethics and humanistic values.

5.2.3. Insufficient Cross-System Synergy and Resource Integration: Technical Integration

The core challenge in the construction of automated ports lies in the insufficient capability for cross-system synergy and data integration [90]. The most prominent manifestation of technical collaboration barriers is the fragmentation of equipment communication protocols and the lack of standardized interfaces. For example, at Ningbo Zhoushan Port, the data interactivity rate between the port’s Terminal Operating System (TOS) and external platforms such as customs and railways is less than 45%, which directly results in an increase of 2.3 h in container turnover time. The phenomenon of “data silos” among heterogeneous systems further exacerbates resource misallocation. Meanwhile, with the increasing popularity of IoT and blockchain technologies, cybersecurity risks are on the rise, and a comprehensive security protection system has yet to be perfected [91]. The Baltic and International Maritime Council (BIMCO) 2024 white paper emphasizes that 62% of shipping firms now mandate academic partnerships to address cybersecurity risks in autonomous navigation, reflecting growing interdependence between industrial needs and scholarly research [92]. Additionally, high costs and the complexity of technological integration continue to constrain large-scale applications. Although the cost of a single automated container truck has decreased by 27.5% since 2021, the initial investment pressure for AGVs and cloud-native platforms remains significant.

Furthermore, Industry 5.0 principles are reshaping IoT deployment in dry port–seaport networks, as seen in case studies from Iranian logistics hubs. IoT-driven multi-objective optimization models have reduced shipment tardiness by 68% and improved supply chain transparency by 71%, albeit with high initial infrastructure investments. Such findings underscore the need for cost–benefit analyses when scaling IoT solutions across heterogeneous maritime ecosystems [93].

Future opportunities, however, are focused on technological integration and ecosystem reconstruction. The deepened application of digital twin technology provides a new paradigm for cross-industry collaboration [94]. For instance, Rizhao Port has constructed a 1:1 virtual model for real-time monitoring and intelligent scheduling, resulting in a 20% increase in equipment utilization. Accelerated standardization processes are also generating collaborative benefits. The establishment of 18 international standards, including the Optical Character Recognition (OCR) and electronic seal encryption, has reduced system integration costs by 60%, driving business model innovations such as “port + finance” and “port + trade”. Qingdao Port has seen a 12% increase in value-added revenue per ship. The green and intelligent transformation has also opened up global market opportunities [95]. Tianjin Port’s “smart zero-carbon” terminal model has been emulated by countries such as Thailand and Peru. The application of hydrogen and electric equipment not only reduces carbon emissions but also creates dual values of technological export and ecosystem collaboration.

6. Conclusions

This study’s systematic analysis of 488 publications (2000–2024) reveals the rapid development in intelligent maritime transportation, with a compound annual growth rate of 14.2% since 2018. This trend is closely related to China’s “Smart Ocean” strategy and its dominant position in global academic output, accounting for 38.5%. Although China’s academic system, centered around Dalian Maritime University and Shanghai Maritime University, demonstrates strong research capabilities, institutional collaboration remains insufficient, with only 17.3% of the total cooperation network involving cross-team efforts. The tripartite framework offers academically validated strategies for global stakeholders, such as modular technology packages achieving a 42% cost reduction for Jakarta Port AGV retrofitting and data sovereignty protocols to mitigate 8.4 billion annual cross-border data losses. The analysis shows that technological innovation is focused on two main areas: Internet technology applications facing challenges of data security (12.6% of publications) and multi-source data integration (9.8%), and automated port infrastructure that requires deeper integration of the Internet of Things, blockchain, and 5G technologies. Therefore, this paper provides an overview of the current research status in these two application areas to offer a reference for future exploration in this field.

There is an urgent need for a systematic breakthrough in the disconnect between technological innovation and practical application. Research on the climate adaptability of navigation algorithms (covered in 4.2% of publications) needs to shift from single-meteorological-parameter analysis to multi-modal extreme scenario modeling, for example, integrating typhoon trajectory prediction models with ship hydrodynamic parameter databases to construct a dynamic collision avoidance system based on anthropomorphic computing. The limitations of ship-type universality in trajectory prediction (2.7% of publications) require breaking through traditional simulation frameworks to develop transfer learning architectures that integrate AIS trajectory data with ship ergonomics parameters, enabling knowledge sharing across different tonnages and ship types through federated learning mechanisms. The gap between academic prototypes (83% of publications) and industrial practice (6% of patent citations) calls for innovation in industry–academia research collaboration mechanisms. Drawing on the example of Shenzhen DeepSeek’s intelligent port ecosystem, a port digital twin platform integrating blockchain and 5G can be established, embedding academic algorithm validation into classification society technical specifications and port IoT operation data streams. On the methodological level, a “data–experiment–policy” conversion loop needs to be constructed: on the one hand, through integrating real-time VHF Data Exchange System (VDES) trajectories, shipboard black-box data, and maritime insurance claim case libraries; on the other hand, through a virtual–real assessment framework, such as embedding the cognitive behavior analysis of stakeholders like shipowners and pilots in real ship trials, to promote the transformation of adaptive navigation algorithms from the laboratory to IMO policy white papers.

Arsenio Dominguez, Secretary-General of the IMO, has stated that “As a global industry responsible for transporting over 80% of global trade, shipping is indispensable”. Therefore, the intelligent development of the shipping industry is imperative. This methodological consolidation of autonomous maritime systems literature facilitates scholarly comprehension of maritime cybernetics progression, mapping technological trajectories while discerning emergent innovation potentials within current operational paradigms.

Methodological constraints merit explicit acknowledgment in this systematic literature retrieval. Firstly, the systematic literature search based on the WoS database may have selection bias, potentially missing important achievements from non-indexed repositories. Secondly, the focus on intelligent maritime systems may not adequately cover related fields such as maritime logistics optimization and safety enhancement. Third, the methodological emphasis on the bibliometric analysis of academic papers lacks cross-verification with industry white papers, patent documents, and policy regulations. Future work will expand interdisciplinary validation, addressing regulatory gaps for unmanned ships through accelerated legislative alignment (currently a 2.1-year lag). Lastly, while bibliometric methods provide rich data support and quantitative analysis, they have limitations in deeply explaining the causal relationships between different technologies and strategies or evaluating their practical effects. To overcome these limitations, future research should adopt a hybrid method combining bibliometrics, expert interviews, and real-time data triangulation, expand the data source to technical specifications and industry reports in the maritime field, and break through the bottleneck of the insufficient explanatory power of causality in the existing literature through a multi-dimensional evaluation framework combining empirical research and quantitative analysis so as to systematically capture the recorded and unrecorded innovation dynamics and provide more operational decision support for the development of intelligent shipping.