International Journal of Economics.
Business and Accounting Research (IJEBAR) Peer Reviewed Ae International Journal Vol-9.
Issue-3, 2025 (IJEBAR) E-ISSN: 2614-1280 P-ISSN 2622-4771 https://jurnal.
stie-aas.
id/index.
php/IJEBAR AI-BASED SCHEDULING FOR COST-EFFECTIVE
MARITIME ENERGY MANAGEMENT
Agus Aji Samekto Maritime University AMNI.
Semarang.
Indonesia Email: agus.
amni@google.
Abstract:
This study investigates the role of artificial intelligence-based scheduling in optimizing maritime energy management systems to enhance sustainability and cost efficiency.
Using quantitative analysis derived from real-time operational data of 15 international shipping routes between 2015 and 2024, the research applies multi-vector energy optimization models that integrate machine learningbased predictive scheduling, port turnaround time analysis, and adaptive fuel management algorithms.
The empirical findings indicate that artificial intelligence scheduling reduces overall operational energy consumption by 14Ae18% and improves system reliability by approximately 25%, while achieving significant reductions in carbon intensity compared with conventional scheduling practices.
Regression and sensitivity analyses confirm that adaptive optimization in voyage planning contributes directly to both financial and environmental performance The study concludes that AI-driven scheduling frameworks provide a measurable pathway toward achieving International Maritime Organization decarbonization targets and ensuring economically viable shipping operations in a competitive global environment.
Keywords:
AI scheduling.
maritime energy management.
cost optimization.
digital maritime operations Introduction The decarbonization and digital transformation of maritime transport have become central to the global sustainability agenda, driving the adoption of intelligent management systems that integrate artificial intelligence into operational planning.
Maritime transport contributes nearly 3% of total global greenhouse gas emissions, emphasizing the urgent need for technological frameworks capable of optimizing energy consumption while maintaining operational reliability (Deja and Ulewicz, 2021.
Zanobetti et al.
, 2.
Recent developments in digitalization and automation (Ichimura et al.
, 2022.
Lind et al.
, 2.
have shown that advanced data-driven tools such as artificial intelligence and machine learning can substantially improve the efficiency of scheduling, routing, and energy utilization in maritime operations.
However, despite these advances, the practical implementation of AI-based energy management remains limited, particularly in long-distance freight and container transport, where multi-vector energy sources and complex operational constraints dominate decision-making.
Studies by Taghavi et al.
and Perna et al.
demonstrate that integrating AI optimization with hybrid energy systems.
International Journal of Economics.
Business and Accounting Research (IJEBAR) Page 302 International Journal of Economics.
Business and Accounting Research (IJEBAR) Peer Reviewed Ae International Journal Vol-9.
Issue-3, 2025 (IJEBAR) E-ISSN: 2614-1280 P-ISSN 2622-4771 https://jurnal.
stie-aas.
id/index.
php/IJEBAR including hydrogen and methanol, can provide significant improvements in economic feasibility and emissions reduction.
Similarly.
Elgharbi et al.
emphasize the necessity of resilient data networks such as LoRaWAN to ensure reliable, low-latency communication among distributed maritime assets, thereby enabling real-time energy scheduling.
In the broader economic context.
SzczepaEska-Przekota and Przekota .
identify macroeconomic fluctuations as critical determinants of maritime transport performance, suggesting that costeffective scheduling models can serve as stabilizing mechanisms against volatile fuel markets.
the same time, the bibliometric synthesis by DragoviN et al.
highlights an emerging research trend toward integrated digital and energy management systems that combine operational efficiency with environmental responsibility.
This research builds upon these insights by developing a quantitative framework for AI-based scheduling in maritime energy management, emphasizing empirical validation through panel data from international routes.
employing optimization algorithms trained on voyage duration, load capacity, and meteorological variables, the study quantifies the potential of AI-driven scheduling to reduce both operational costs and carbon emissions while improving fleet energy efficiency.
The findings aim to contribute to the evolving paradigm of sustainable maritime management, in line with International Maritime Organization 2050 targets and the transition toward carbon-neutral maritime logistics.
Table 1.
Global Context of AI-Based Maritime Energy Management No.
Indicator / Variable Quantitative Data Source (Yea.
5Ae2.
Global maritime transport 280 million tonnes IMO .
energy consumption of fuel oil equivalent annually Research Implication Highlights the large-scale energy demand requiring AIbased optimization models Share of global GHG emissions from shipping 9Ae3.
1% of total global emissions Deja & Ulewicz .
Zanobetti et al.
Establishes the environmental significance of improving energy management efficiency Average vessel energy cost 45Ae55% of total component in total voyage cost operating costs SzczepaEskaPrzekota & Przekota .
Justifies the economic necessity of cost-efficient AI
scheduling models Projected annual increase in digitalization adoption in maritime sector
6% CAGR
9Ae2.
Ichimura et al.
Supports the integration of AI .
DragoviN et and digital twins in scheduling .
Reduction in energy use through AI-based predictive maintenance 14Ae18% average Taghavi et al.
Empirical evidence of AI potential in energy cost Improvement in voyage 22Ae28% decrease in Elgharbi et al.
scheduling accuracy using delay variance .
AI algorithms Demonstrates operational benefits of AI-enhanced scheduling in variable Proportion of maritime 34% of global ports FrkoviN et al.
companies adopting smart by 2024 Indicates growing readiness for AI-based energy coordination International Journal of Economics.
Business and Accounting Research (IJEBAR) Page 303 International Journal of Economics.
Business and Accounting Research (IJEBAR) Peer Reviewed Ae International Journal Vol-9.
Issue-3, 2025 (IJEBAR) E-ISSN: 2614-1280 P-ISSN 2622-4771 https://jurnal.
stie-aas.
id/index.
php/IJEBAR No.
Indicator / Variable Quantitative Data .
5Ae2.
Source (Yea.
port energy systems Research Implication between ships and ports Average cost savings from 12Ae20% reduction hybrid AI-energy in operating Reduction in COCC intensity .
COCC /tonmil.
via AI optimization Perna et al.
Validates cost-effectiveness as a measurable performance 10Ae15% decline vs Taghavi et al.
baseline scheduling .
10 Estimated contribution of 20Ae25% of total AI scheduling to meeting required reduction IMO 2050 emission targets IMO .
Lind et al.
Quantifies direct environmental outcomes of AI scheduling adoption Confirms strategic alignment of AI-driven energy management with global decarbonization pathways The quantitative data presented in Table 1 demonstrate that global maritime transport is undergoing an accelerated transition toward digital and sustainable energy management.
The integration of artificial intelligence into voyage scheduling and energy optimization has shown 14Ae18% efficiency improvement, 12Ae20% cost reduction, and up to 25% progress toward IMO 2050 decarbonization goals.
These statistics provide a strong empirical basis for the argument that AI-based scheduling is both an economic and environmental imperative in the maritime Literature Review Maritime energy management has become a critical research domain due to the increasing pressure on shipping industries to reduce greenhouse gas (GHG) emissions and operational costs (Deja & Ulewicz, 2.
The International Maritime Organization (IMO) has mandated a reduction of 50% in total GHG emissions by 2050 compared to 2008 levels.
This ambitious goal has accelerated the integration of artificial intelligence (AI), digitalization, and predictive systems in maritime operations (Zanobetti et al.
, 2023.
Taghavi et al.
, 2.
Energy management within maritime operations encompasses voyage optimization, propulsion efficiency, cargo handling energy use, and port interface processes (Filina-Dawidowicz et al.
Traditional optimization methods, while effective for static systems, struggle to address dynamic maritime environments characterized by unpredictable weather, variable cargo weights, and changing port conditions.
Hence.
AI-based scheduling offers adaptive, data-driven, and realtime decision support for achieving cost-effective energy performance (Taghavi et al.
, 2.
Digital transformation has significantly reshaped maritime logistics and vessel operations.
Ichimura et al.
mapped digitalization trends and found a 12.
6% compound annual growth rate (CAGR) in technological adoption between 2019Ae2024, with AI and data analytics leading Similarly.
Kaklis et al.
emphasized that digital twins integrated with AI provide predictive insights into vessel performance and fuel consumption, enabling proactive energy Surucu-Balci et al.
identified blockchain and cloud integration as enablers for transparent data exchange across maritime supply chains, while Hamidi et al.
proposed a International Journal of Economics.
Business and Accounting Research (IJEBAR) Page 304 International Journal of Economics.
Business and Accounting Research (IJEBAR) Peer Reviewed Ae International Journal Vol-9.
Issue-3, 2025 (IJEBAR) E-ISSN: 2614-1280 P-ISSN 2622-4771 https://jurnal.
stie-aas.
id/index.
php/IJEBAR three-stage digital maturity model for assessing AI readiness in maritime logistics.
In Asia, the study by Janmethakulwat and Thanasopon .
revealed that institutional and regulatory support significantly influence digital technology adoption rates among Thai shipowners.
Despite progress, barriers remain.
Zhao et al.
identified data fragmentation, limited interconnectivity, and insufficient analytics capability as the main obstacles preventing maritime enterprises from leveraging AI for energy efficiency.
Recent advancements show that AI can drastically improve maritime scheduling accuracy and fuel efficiency.
Bourzak et al.
developed a machine learning model predicting vessel speed over ground, improving route efficiency by up to 15%.
Similarly.
Cuong et al.
introduced robust adversarial reinforcement learning to optimize maritime supply chains, demonstrating 12Ae18% cost reduction through adaptive scheduling algorithms.
Taghavi et al.
performed a techno-economic assessment on AI-based coordinated energy systems, indicating potential cost savings of 12Ae20% in multi-vector energy networks.
Furthermore.
Buonomano et al.
integrated AI-driven dynamic simulations into ship design, achieving 14Ae18% improvement in energy performance.
The integration of AI scheduling and renewable energy sources is increasingly emphasized.
FrkoviN et al.
proposed a shore-to-ship electrification system supported by renewable energy, while Perna et al.
demonstrated the feasibility of green hydrogen supply chains optimized by AI-based predictive systems for energy The decarbonization of maritime transport is an imperative aligned with the IMO 2050 vision (Lind et al.
, 2.
Akac et al.
evaluated methanol as a sustainable fuel, while Mohd Tamam et al.
reviewed pathways to carbon neutrality, including hybrid propulsion and battery-based energy storage.
Zanobetti et al.
and Fadiga et al.
confirmed that decarbonizing maritime ports and vessels requires AI-assisted coordination of energy systems to balance cost and emission goals.
Moreover, the circular economy approach explored by Okumus et al.
stresses the role of data analytics and machine intelligence in resource reuse and operational sustainability.
Digital innovation in ports, such as smart gate models and IoT-based container monitoring (Ferreira et al.
, 2025.
Ledesma & Lamo, 2.
, further support integrated energy management along the maritime logistics chain.
Energy expenditure accounts for 45Ae55% of vessel operational costs (SzczepaEska-Przekota & Przekota, 2.
AI-driven scheduling systems directly impact cost-efficiency through dynamic route optimization, predictive maintenance, and fuel consumption control.
Maternovy and Materna .
noted that human error remains a major factor in operational inefficiency, suggesting AI-based systems as essential tools for minimizing risk and cost variability.
From a macroeconomic perspective.
Alves da Costa et al.
applied multimodal super network analysis to assess cabotage competitiveness, revealing that digital scheduling significantly enhances maritime economic efficiency.
In a similar vein.
Dimakis et al.
employed textual analytics to highlight that sustainability-driven supply chain innovations in European maritime logistics correlate with AI adoption intensity.
International Journal of Economics.
Business and Accounting Research (IJEBAR) Page 305 International Journal of Economics.
Business and Accounting Research (IJEBAR) Peer Reviewed Ae International Journal Vol-9.
Issue-3, 2025 (IJEBAR) E-ISSN: 2614-1280 P-ISSN 2622-4771 https://jurnal.
stie-aas.
id/index.
php/IJEBAR Human-AI collaboration is crucial for successful implementation.
Wu et al.
discussed the sociotechnical implications of AI socialisation, emphasizing skill transformation and digital readiness of seafarers.
Shahbakhsh et al.
argued that the autonomous era of shipping necessitates redefining seafarersAo roles from operational to supervisory.
Dewan and Godina .
highlighted the cultural transformation among seafarers embracing energy-efficient practices aided by AI systems.
Educational institutions are now adapting curricula to support digital sustainability.
Karahalios .
emphasized the need for AI literacy and digital sustainability education in maritime academies to strengthen human readiness in maritime 5.
Despite rapid progress, significant research gaps persist:
Integration Challenges: Limited interoperability between AI scheduling systems and traditional maritime energy infrastructures (Zhao et al.
, 2.
Data Reliability: Inconsistent and incomplete energy consumption data across fleets limit model accuracy (Mojica Herazo et al.
, 2.
Cybersecurity Risks: As AI and IoT integration increases, maritime cyber supply chains become vulnerable (Diaz et al.
, 2.
Socioeconomic Equity: Few studies have explored how AI-based energy systems affect developing maritime economies, such as Southeast Asia (Tahsin et al.
, 2.
Emerging directions include multi-agent reinforcement learning for coordinated scheduling (Cuong et al.
, 2.
AIAeblockchain convergence for transparent energy trading (Surucu-Balci et , 2.
, and AI-assisted hydrogen logistics networks for zero-emission shipping (Perna et al.
Based on the reviewed literature, this study conceptualizes AI-based maritime energy scheduling as a system integrating four key pillars:
Predictive Analytics: Machine learning algorithms for real-time forecasting of vessel energy .
Optimization Algorithms: Reinforcement learning models minimizing operational costs while maintaining route efficiency.
Renewable Energy Integration: AI-enabled scheduling for hybrid propulsion and green port .
Human-AI Synergy: Digital literacy and adaptive maritime education for sustainable operational transition.
This framework aligns with the Maritime 5.
0 paradigm, emphasizing digital, sustainable, and human-centric maritime operations.
Table 2.
Systematic Literature Review (SLR) on AI-Based Scheduling for Cost-Effective Maritime Energy Management No Author.
& Year Methods 1 Taghavi et al.
Techno-economic Energy Conversion multiand Management vector AI Key Findings Demonstrated cost-effective integration of underground storage using AI-based scheduling to optimize energy flows in maritime Identified Research Gaps Lack of integration between AI-based scheduling and real-time ship operations data.
limited validation in real International Journal of Economics.
Business and Accounting Research (IJEBAR) Page 306 International Journal of Economics.
Business and Accounting Research (IJEBAR) Peer Reviewed Ae International Journal Vol-9.
Issue-3, 2025 (IJEBAR) E-ISSN: 2614-1280 P-ISSN 2622-4771 https://jurnal.
stie-aas.
id/index.
php/IJEBAR No Author.
& Year Methods Key Findings Identified Research Gaps maritime ports.
Buonomano et al.
Dynamic Developed energy .
Applied energy optimization models for Thermal Engineering design optimization large ships via information achieved 14% energy savings.
Did not employ adaptive AI learning.
still rule-based and nonautonomous.
Lind et al.
Transportation Research Procedia Absence of AI integration for predictive scheduling and energy balancing.
FrkoviN et al.
System-level Ocean Engineering modeling of Proposed shore-to-ship electrification system with renewable energy nexus.
improved port sustainability by 20%.
Ravi et al.
Procedia CIRP Life cycle assessment (LCA) Conducted LCA of maritime Lacked predictive AI battery systems, highlighting models for optimizing future cost reduction battery use under dynamic energy demand.
Perna et al.
Energy Conversion and Management Simulation and feasibility analysis Developed green hydrogen supply chain model for maritime transport.
lifecycle emissions by 35%.
Zanobetti et al.
Journal of Cleaner Production Sustainability Assessed alternative No decision-support AI propulsion systems.
electric tool for dynamic costand hybrid systems showed energy optimization.
best sustainability trade-offs.
DragoviN et al.
Ocean Engineering Bibliometric Reviewed maritime bibliometric trends .
4Ae identified AI and digitalization as dominant Deja & Ulewicz Transportation Research Procedia Environmental threat assessment Highlighted major No energy scheduling environmental risks in framework for real-time maritime transport including emission reduction.
fuel inefficiency.
Strategic foresight and mapping Identified digitalization trends shaping maritime strategies through AI.
IoT, and data analytics.
Lack of cost-energy integration framework for AI-driven scheduling 11 Eddine Elgharbi et al.
Network simulation Proposed resilient maritime Limited exploration of how 10 Ichimura et al.
Digital Business Case analysis.
Introduced Sea Traffic system architecture Management concept benefiting all stakeholders through synchronized Did not evaluate cost optimization under variable energy pricing.
no AI scheduling applied.
Need for AI scheduling to coordinate hydrogen logistics with vessel Lack of empirical studies quantifying energy-saving impacts of AI scheduling.
International Journal of Economics.
Business and Accounting Research (IJEBAR) Page 307 International Journal of Economics.
Business and Accounting Research (IJEBAR) Peer Reviewed Ae International Journal Vol-9.
Issue-3, 2025 (IJEBAR) E-ISSN: 2614-1280 P-ISSN 2622-4771 https://jurnal.
stie-aas.
id/index.
php/IJEBAR No Author.
& Year .
Computer Communications Methods Key Findings (LoRaWAN mes.
data transmission architecture for real-time Identified Research Gaps these networks support AIbased energy management.
12 Wu et al.
Transportation Research Part E Quantitative survey.
Demonstrated human-AI Focused on service structural modeling collaboration improves innovation, not AI maritime service efficiency.
scheduling for energy 13 Zhang et al.
Science of The Total Environment Literature review and thematic Green port digitalization significantly enhances sustainability metrics.
14 Fadiga et al.
Journal of Cleaner Production Systematic review Summarized decarbonization Absent model integrating strategies for maritime ports.
AI scheduling to align port emphasized renewable energy and vessel demand.
energy transitions.
AI scheduling for portAe ship energy synchronization remains 15 Hamidi et al.
Digital maturity Journal of Industrial assessment Information Integration Developed 3-stage digital readiness model for blockchain in maritime AI-based scheduling maturity not yet evaluated in operational contexts.
16 Cuong et al.
Engineering Applications of AI Proposed adversarial RL for maritime supply chain resilience by 15%.
No specific focus on energy cost optimization.
potential extension to AI 17 Bourzak et al.
Machine learning IFAC-PapersOnLine regression Predicted vessel speed with 93% accuracy to improve transport efficiency.
Missing link to energy consumption forecasting and AI cost scheduling.
18 Al-Okaily et al.
Heliyon Systematic review Found that Industry 4.
Lack of AI-based costAe technologies improve supply energy integration chain sustainability.
frameworks for maritime 19 Maternovy & Materna .
Transportation Research Procedia HFACS analysis Studied human factors in maritime accidents.
Reinforcement 20 Samekto et al.
Quantitative Int.
Mechanical analysis during Engineering COVID-19 Future research could combine AI scheduling with human reliability data.
Identified green supply chain AI-based scheduling for models sustaining maritime cost-energy optimization logistics during disruptions.
under crisis scenarios Table 2 demonstrate that Dominant Methods: .
Simulation .
%).
Optimization modeling .
%).
Bibliometric/systematic review .
%).
Quantitative analysis .
%).
Main Themes Identified: .
AI-driven decision support and scheduling (Taghavi et al.
, 2025.
Cuong et al.
International Journal of Economics.
Business and Accounting Research (IJEBAR) Page 308 International Journal of Economics.
Business and Accounting Research (IJEBAR) Peer Reviewed Ae International Journal Vol-9.
Issue-3, 2025 (IJEBAR) E-ISSN: 2614-1280 P-ISSN 2622-4771 https://jurnal.
stie-aas.
id/index.
php/IJEBAR .
Digital transformation and Industry 4.
0 (Ichimura et al.
, 2022.
Al-Okaily et al.
, 2.
Energy transition technologies .
ydrogen, electrification, battery, renewable.
Primary Gaps for Future Research: .
Lack of real-time AI scheduling integration for dynamic vessel-port energy optimization.
Limited quantitative validation using actual maritime operational data.
Need for multi-objective costAeenergyAeemission trade-off models driven by AI.
Absence of interoperable frameworks connecting port infrastructure, ship operations, and digital twins.
Methodology Research Design This study adopts a quantitative research design integrating panel data econometric analysis and AI-based predictive optimization modelling.
The objective is to evaluate the cost-effectiveness of artificial intelligence (AI)-based scheduling systems in maritime energy management, particularly for cargo and container vessels operating in major global shipping lanes between 2015 and 2024.
The research applies descriptive, inferential, and simulation-based quantitative approaches to validate the efficiency gains, fuel savings, and emission reductions achieved through AI scheduling algorithms compared with conventional operational scheduling systems.
The methodological framework follows the structure used by SzczepaEska-Przekota & Przekota .
for macroeconomic modelling of maritime performance, and integrates energy optimization modelling as developed by Taghavi et al.
and Buonomano et al.
for techno-economic assessment of coordinated energy management in multi-vector systems.
Data Collection Data were collected from automatic identification system (AIS) datasets, ship energy management logs, and digital voyage data recorders from ten major global ports (Singapore.
Rotterdam.
Busan.
Hamburg.
Tanjung Priok.
Shanghai.
Dubai.
Santos.
GdaEsk, and New Yor.
Complementary datasets were obtained from: .
IMO Greenhouse Gas Study .
UNCTAD Maritime Statistics Database.
World Bank Port Infrastructure Index .
9Ae2.
Peer-reviewed datasets cited in Deja & Ulewicz .
DragoviN et al.
, and FrkoviN et al.
Data reliability was ensured using cross-source triangulation, where each data entry was validated against at least two independent sources.
Table 3.
Variable Definition and Measurement Variable Symbol Measurement Unit Data Source Expected Effect Energy Cost Efficiency ECE % reduction vs.
Vessel logbook.
IMO Positive Fuel Consumption tonnes/day AIS data Negative Voyage Duration AIS data Negative AI Scheduling Adoption AISA 0Ae1 .
Survey & log system Positive Index COCC Emission Intensity CEI gCOCC /ton-mile IMO dataset Negative Digitalization Index 0Ae100 scale World Bank Positive Operational Cost USD/voyage Company record Negative International Journal of Economics.
Business and Accounting Research (IJEBAR) Page 309 International Journal of Economics.
Business and Accounting Research (IJEBAR) Peer Reviewed Ae International Journal Vol-9.
Issue-3, 2025 (IJEBAR) E-ISSN: 2614-1280 P-ISSN 2622-4771 https://jurnal.
stie-aas.
id/index.
php/IJEBAR Table 3 demonstrate that variables form a comprehensive model linking technological adoption (AISA.
DI) with operational and environmental outcomes (ECE.
FC.
CEI.
OC).
The expected directions of effect suggest that greater integration of AI and digital systems in maritime operations enhances energy efficiency, cost-effectiveness, and emission performance, thereby validating the research hypothesis that AI-based scheduling contributes positively to sustainable maritime energy management.
Econometric Model The study employs a panel data fixed-effect model (FEM) to quantify the influence of AI-based scheduling on cost and energy efficiency.
The model tests HCA : CA = 0, indicating no significant effect of AI scheduling on energy cost efficiency.
Significance was tested at = 0.
AI Optimization Simulation An AI-based scheduling model was developed using reinforcement learning (RL) combined with multi-objective optimization (MOO).
The architecture was based on the MARL framework proposed by Cuong et al.
and Kaklis et al.
for digital twin applications in maritime .
AI engine trained with 70% of dataset, validated on 30%.
Performance metrics: MAE (Mean Absolute Erro.
RMSE (Root Mean Square Erro.
, and optimization gain (%) in energy cost and emission reduction.
RESULTS AND DISCUSSION
The descriptive analysis was conducted using a panel dataset of 4,280 vessels operating across ten global ports during the period 2015Ae2024.
The sample covers various vessel types including bulk carriers .
%), container ships .
%), oil tankers .
%), and passenger liners .
%).
Table 4 Presents the summary statistics of the primary variables Variable Mean Std.
Dev.
Min Max Observation Energy Cost Efficiency (ECE, %) 42,800 AI Scheduling Adoption Index (AISA) 42,800 Fuel Consumption .
on/da.
42,800 Voyage Duration .
42,800 COCC Emission Intensity .
COCC /ton-mil.
42,800
Operational Cost (USD/voyag.
128,400 34,900 62,000 213,000 42,800 Table 4 demonstrate that AI-based scheduling systems were adopted by approximately 47% of vessels by 2024, reflecting an accelerated transition toward digital energy management systems, consistent with Ichimura et al.
and Wu et al.
who reported a similar trend in maritime digitalization.
The Fixed Effect Model (FEM) was selected based on the Hausman Test (NA = 43.
62, p < 0.
, confirming that vessel-specific effects were significant and non-random.
Variable Constant Table 5.
The main regression results Coefficient Std.
Error t-statistic p-value Hypothesis International Journal of Economics.
Business and Accounting Research (IJEBAR) Page 310 International Journal of Economics.
Business and Accounting Research (IJEBAR) Peer Reviewed Ae International Journal Vol-9.
Issue-3, 2025 (IJEBAR) E-ISSN: 2614-1280 P-ISSN 2622-4771 https://jurnal.
stie-aas.
id/index.
php/IJEBAR Variable Coefficient Std.
Error t-statistic p-value Hypothesis AI Scheduling Adoption Index (AISA) 000 HCA supported Fuel Consumption (FC) 014 HCC supported Voyage Duration (VD) 007 HCE supported Digitalization Index (DI) 005 HCE supported COCC Emission Intensity (CEI) 003 HCI supported RA Adj.
Durbin-Watson Table 5 demonstrate that the model explains approximately 68.
2% of the variation in energy cost efficiency (ECE), demonstrating that AI-based scheduling exerts a strong positive influence on maritime operational sustainability.
Specifically, vessels adopting AI-based scheduling exhibited an average 4.
33% higher energy cost efficiency relative to non-AI vessels.
This aligns with Taghavi et al.
and Buonomano et al.
, who found that machine learning-based energy coordination can reduce overall cost by 5Ae7% in multi-vector maritime systems.
The significant negative effect of fuel consumption (FC) and COCC intensity (CEI) on energy efficiency corroborates the findings of Zanobetti et al.
and FrkoviN et al.
indicating that optimization of energy inputs directly enhances maritime decarbonization.
The reinforcement learning (RL)Aebased scheduling simulation achieved notable performance improvements across several key indicators compared to conventional scheduling algorithms.
Table 6.
AI Optimization Simulation Results Metric Fuel Cost (USD/voyag.
Baseline (Traditional Schedulin.
128,400 AI-Based Scheduling 115,600 Improvement
(%)
9 Ie Voyage Duration .
1 Ie COCC Emission Intensity .
COCC /tonmil.
5 Ie Energy Cost Efficiency (ECE, %) 8 Ic Optimization Gain (Model MAE) 1 Ie Table 6 demonstrate that The AI-based model outperformed the traditional system by reducing average fuel cost per voyage by nearly 10% and emissions by 13.
Model accuracy, measured by Mean Absolute Error (MAE) and RMSE, improved by 33% and 28% respectively, confirming the robustness of the proposed predictive scheduling system.
These findings are consistent with the reinforcement learning results of Cuong et al.
and digital twin integration proposed by Kaklis et al.
, both emphasizing AIAos contribution to operational optimization in maritime logistics.
International Journal of Economics.
Business and Accounting Research (IJEBAR) Page 311 International Journal of Economics.
Business and Accounting Research (IJEBAR) Peer Reviewed Ae International Journal Vol-9.
Issue-3, 2025 (IJEBAR) E-ISSN: 2614-1280 P-ISSN 2622-4771 https://jurnal.
stie-aas.
id/index.
php/IJEBAR Figure 1.
Performance Impact of AI-Based Maritime Scheduling and the trend of increasing AI Adoption Rate and Energy Efficiency Figure 1 illustrates the Performance Impact of AI-Based Maritime Scheduling, highlighting improvements in energy efficiency, reductions in operational costs, decreases in voyage delays, and contributions toward achieving the IMO 2050 emission reduction targets.
Depicts the trend of increasing AI Adoption Rate and Energy Efficiency Improvement in the maritime sector from 2019 to 2024, demonstrating a positive correlation between digitalization and energy efficiency The findings confirm that AI-based scheduling significantly improves cost efficiency by optimizing engine load distribution, voyage planning, and weather routing.
This supports earlier International Journal of Economics.
Business and Accounting Research (IJEBAR) Page 312 International Journal of Economics.
Business and Accounting Research (IJEBAR) Peer Reviewed Ae International Journal Vol-9.
Issue-3, 2025 (IJEBAR) E-ISSN: 2614-1280 P-ISSN 2622-4771 https://jurnal.
stie-aas.
id/index.
php/IJEBAR evidence from SzczepaEska-Przekota and Przekota .
that macroeconomic efficiency in maritime operations is directly affected by technological sophistication.
Moreover, the positive correlation between the Digitalization Index (DI) and energy efficiency indicates that digital maturity accelerates sustainability outcomes, as also emphasized by Hamidi et al.
in the digital maturity model for blockchain readiness in maritime logistics.
The efficiency improvement of 19.
8% recorded in this study aligns closely with Buonomano et al.
, who demonstrated a 20Ae25% gain through digital simulation-based optimization in ship design.
Similarly.
Taghavi et al.
highlighted that AI-integrated energy management enables costeffective coordination between fuel storage, propulsion, and auxiliary systems.
However, despite these gains, challenges remain.
Gu et al.
emphasized that maritime supply chain resilience depends on synchronization between AI optimization and real-time data interoperability, an issue still constrained by network latency and inconsistent IoT sensor reliability .
ee also Elgharbi et , 2.
AI-based scheduling not only reduces fuel and energy cost but also substantially cuts carbon emissions, in line with the IMOAos 2050 decarbonization goals.
The observed 13.
emission reduction is comparable with results obtained by FrkoviN et al.
and Perna et al.
on the integration of renewable energy systems in ship operations.
This shift also enhances operational resilience, as reduced voyage time and predictive maintenance minimize risk of delay and environmental incidentsAiechoing the accident control perspective of Hanafiah et al.
and Maternovy & Materna .
The econometric estimation demonstrates that for every 1% increase in AI adoption, there is an approximate 0.
43% increase in energy cost efficiency, holding other variables constant.
Over a ten-year horizon, this implies potential savings exceeding USD 12,800 per vessel annually, depending on voyage frequency.
This result resonates with the VAR macroeconomic findings of SzczepaEska-Przekota & Przekota .
, confirming that maritime innovation is a statistically significant driver of national economic The study provides strategic insights for ship operators and policymakers:
Integration of AI systems in voyage scheduling and engine management should be prioritized in green investment portfolios.
Digital twin and reinforcement learning frameworks can serve as real-time energy monitoring systems, supporting sustainable operational decisions.
Port authorities and maritime regulators must develop standardized data-sharing frameworks to enhance interoperability, as proposed by Lind et al.
in the Sea Traffic Management .
Education and workforce training, as recommended by Karahalios .
, are essential for accelerating digital literacy and ensuring the ethical deployment of AI technologies at sea.
The results reinforce the theoretical link between digital transformation and sustainability performance in maritime operations.
They substantiate the argument of Al-Okaily et al.
that Industry 4.
0 technologies exert a synergistic effect on supply chain sustainability.
The integration of AI scheduling with energy management thus constitutes a core mechanism of Maritime 5.
0, bridging environmental performance with economic rationality.
Despite the robustness of findings, several limitations are acknowledged:
Data availability was limited to large ports, potentially biasing results toward high-capacity International Journal of Economics.
Business and Accounting Research (IJEBAR) Page 313 International Journal of Economics.
Business and Accounting Research (IJEBAR) Peer Reviewed Ae International Journal Vol-9.
Issue-3, 2025 (IJEBAR) E-ISSN: 2614-1280 P-ISSN 2622-4771 https://jurnal.
stie-aas.
id/index.
php/IJEBAR AI model generalization may differ for short-route coastal shipping where energy variability is lower.
Weather anomalies and port congestion events were not fully simulated in RL algorithms.
Future research should explore hybrid AIAeblockchain energy coordination, integrating digital twins with IoT-based real-time monitoring as suggested by Surucu-Balci et al.
and Zhang et al.
Expanding analysis to regional economies .
Southeast Asia, as noted by Ruthbah et al.
, 2.
may further reveal adaptive models for emerging maritime markets.
The study empirically validates that AI-based scheduling enhances maritime energy cost efficiency, reduces fuel and emission levels, and shortens voyage duration, confirming the transformative role of artificial intelligence in sustainable maritime management.
This finding contributes a novel quantitative framework to the techno-economic discourse on maritime decarbonization, aligning with Taghavi et al.
Buonomano et al.
, and Zanobetti et al.
, and extending their scope through integration of econometric and reinforcement learning models.
Conclusion AI scheduling reduces energy expenditure by up to 20%, directly impacting profitability and voyage planning efficiency.
Emission intensity is lowered by 10Ae15%, supporting IMO 2050 Integration of AI.
IoT, and digital twins ensures real-time adaptive energy management.
Encourages standardization of digital energy frameworks across ports and maritime regulatory This research comprehensively examined the role of Artificial Intelligence (AI)-based scheduling as a transformative mechanism for achieving both cost efficiency and environmental sustainability in maritime energy management.
The integration of quantitative findings from recent studies (Deja & Ulewicz, 2021.
Taghavi et al.
, 2025.
Perna et al.
, 2.
and macroeconomic analyses (SzczepaEska-Przekota & Przekota, 2.
confirms that energy consumption and greenhouse gas emissions remain dominant operational challenges in global shipping, contributing approximately 3% of global emissions and consuming nearly 280 million tonnes of fuel annually.
The study underscores that AI-enabled scheduling models, when combined with digital twin systems and predictive energy algorithms, significantly enhance the optimization of voyage operations, reducing delay variance by 22Ae28% and energy consumption by 14Ae18%.
These results validate the empirical correlation between digital maturity and maritime operational efficiency, aligning with the strategic digitalization trajectory outlined by Ichimura et al.
and DragoviN et al.
Moreover.
AI-driven predictive analytics and reinforcement learning models (Cuong et al.
, 2025.
Bourzak et al.
, 2.
exhibit measurable cost savings ranging between 12Ae20% in operational expenditure, reinforcing their techno-economic viability for the maritime sector.
From an environmental perspective, the deployment of AIbased scheduling systems facilitates compliance with the International Maritime OrganizationAos (IMO) 2050 decarbonization roadmap, contributing up to 25% of the targeted emission The synergy between AI technologies, hydrogen-based propulsion (Perna et al.
, 2.
and shore-to-ship renewable electrification systems (FrkoviN et al.
, 2.
creates a sustainable framework for next-generation maritime energy systems.
The synthesis of multidisciplinary literature .
1Ae2.
further demonstrates that successful AI implementation requires integrated governance, robust data infrastructure, and upskilling of human resources to ensure adaptive and ethical use of intelligent decision-support systems (Wu et al.
, 2025.
Hamidi et al.
International Journal of Economics.
Business and Accounting Research (IJEBAR) Page 314 International Journal of Economics.
Business and Accounting Research (IJEBAR) Peer Reviewed Ae International Journal Vol-9.
Issue-3, 2025 (IJEBAR) E-ISSN: 2614-1280 P-ISSN 2622-4771 https://jurnal.
stie-aas.
id/index.
php/IJEBAR Thus, future research should extend toward the development of interoperable AIAeIoT frameworks and multi-agent reinforcement learning models capable of autonomously coordinating fleet energy behavior under real-time uncertainty.
In conclusion.
AI-based scheduling presents a high-potential, data-driven paradigm that unifies energy efficiency, cost optimization, and digital resilience within the maritime industry.
This approach not only enhances economic performance but also advances global commitments toward sustainable and decarbonized maritime transportation.
Acknowledgement The authors would like to express their highest gratitude to the Yayasan Pembina Kemaritiman Indonesia, the legal entity and founding of Universitas Maritim AMNI Semarang, established in 1964, for its generous financial and institutional support in the implementation of this research The foundationAos long-standing commitment to advancing maritime education, research innovation, and sustainable transport management in Indonesia has provided the essential framework and resources for this study on AI-Based Scheduling for Cost-Effective Maritime Energy Management.
The authors also wish to acknowledge the academic contributions and technical insights shared by the research and development division of Universitas Maritim AMNI Semarang, whose interdisciplinary collaboration in maritime operations, data analytics, and sustainable logistics has greatly enhanced the quality and rigor of this investigation.
This work stands as part of Yayasan Pembina Kemaritiman IndonesiaAos ongoing mission to strengthen IndonesiaAos maritime competitiveness through research-driven innovation and technological transformation in the blue economy.
References