Wind Propulsion Systems (WPS) have gained significant attention as a means of decarbonizing shipping. Limitations in available deck space, emissions reduction targets, and regulatory compliance have led to a wide array of potential WPS configurations, each exhibiting distinct aerodynamic performance and requiring unique optimum sail trims for each unit due to complex interactions. This variability challenges existing aerodynamic models and optimization efforts for maximizing fuel savings. To address this, we present a novel methodology that, for the first time in WPS aerodynamic performance prediction, combines Computational Fluid Dynamics (CFD), independent sail trim optimization, and Machine Learning (ML) to develop surrogate models — Gaussian Process Regression and Feedforward Neural Networks — that rapidly predict aerodynamic performance with CFD-equivalent accuracy. These surrogates capture aerodynamic interactions across various WPS configurations, including unit number, deck arrangement, independent sail trim, hull characteristics, and wind conditions. While employing established ML techniques, our approach is novel in its resource-efficient generation of a comprehensive aerodynamic database, derived from the first in-depth independent trim optimization of a DynaRig case study. Our approach enables the modeling of complex, non-linear interactions that traditional interpolation methods fail to capture. Results show that the developed surrogate models achieve CFD-level accuracy, with an average error below 1 while significantly reducing computational time. This ML-enhanced framework facilitates extensive, rapid WPS design optimizations, supporting efficient integration into performance prediction programs (PPPs) and maximizing fuel savings and emissions reductions tailored to specific routes and wind conditions.Machine Learning; CFD-Simulations; Aerodynamic Performance; Wind Propulsion Systems; Green Shipping; Independent Sail Trim Optimization.
Our research highlights the current state and trends of artificial intelligence (AI) adoption in Denmark’s chartering, particularly in the dry bulk and tanker segments. Companies in the dry bulk sector are leading AI adoption, with the tanker segment closely following and adoption rates in our sample appear higher than national averages reported by consultancies. Most firms are in either the experimental phase or transitioning toward more integrated AI systems, often opting for hybrid models that allow them to maintain internal control over key processes. Factors such as company size and maturity also influence the pace and approach to AI adoption.AI is seen as a tool to enhance rather than replace jobs in the early stages of shipping operations, especially in pre-fixture activities. However, there is greater potential for automation and job substitution in the post-fixture phase, particularly in tasks such as contract (CP) management.
On the supply side, the market for maritime AI and software solutions is highly competitive and fragmented, with many providers offering diverse products. Recent consolidation trends reflect different strategies: some companies, like are specializing in core offerings, while others, like are diversifying into both SaaS and pure software models. These consolidations are not only intensifying competition but also fostering partnerships between rivals—a dynamic known as coopetition. Interestingly, some shipping firms are entering the software market themselves, signaling innovation in business models. Machine learning (ML) technologies are primarily used in pre-fixture tools (like email management and tracking), while generative AI is increasingly applied in post-fixture functions, particularly contract management.
We present a machine learning model for calculation of wave added resistance. The model training is performed using a large set of pre-calculated added resistance curves covering a broad range of ship hulls and operational conditions, i.e. forward speed, draft and relative wave heading. The underlying hydrodynamic model is the classical strip-theory where the wave added resistance is computed according to a modified version of Salvesen’s formulation. It is concluded that the developed data-driven model is able to produce a non-linear mapping between a set of operational conditions as well as the ship’s main particulars to the wave added resistance coefficient.
Current practice for maritime search and rescue (MSAR) adheres to predetermined full-coverage patterns for finding targets. These do not account for key success factors for MSAR missions such as the dynamic location of targets, updates on situational awareness during mission execution, and search vehicle kinematics. Consequently, current practice cannot incorporate realistic MSAR operational conditions into path-finding, increasing the likelihood of mission failure. To address this issue, a novel, flexible path-finding framework is proposed for generating a path while dynamically updating the probability of a target based on the path's trajectories. The solution approach implements the A* algorithm, which can accommodate the dynamics of a vehicle and guarantees the optimality of the final path with respect to the target objective function. Experiments show that a more than 50% improvement in the time needed to guarantee a certain probability of finding a target is exhibited compared to the parallel sweep coverage path-finding approach.
In this article, we develop a deep neural network model to estimate the wave added resistance. The required data to train the model is generated using strip theory calculations over a wide range of hull geometries and operational conditions. The model is efficient as it only requires the ship’s main particulars: length, beam, draft, block coefficient, and slenderness ratio. In addition, we present an application of this model in a vessel performance framework. This will be used for predicting propulsion power and analyzing the degree of biofouling on ships from the company Ultrabulk2. The study shows that the developed deep neural network model produces reliable results in predicting the added wave resistance coefficient in comparison to strip theory calculations. Also, the developed ship propulsion and biofouling analysis display satisfactory output for monitoring hull performance under actual ship operational conditions.
An adaptive machine learning framework is established for an implicit determination of the performance degradation of a ship due to marine growth, i.e., biofouling. The framework is applied in a case study considering telemetry data of a cruise ship operating predominantly in the Caribbean Sea. The dataset encompasses seven years including three dry-docking intervals and several in-water cleaning events. The COVID-19 period receives special focus due to the drastic change in the operational profile. A main outcome of the study is a comparison of the derived performance estimate to the corresponding results of the industry standard ISO 19030. Additional aspects of the present study include the use of special regularization techniques for incremental machine learning and the increase of transparency through the implementation of prediction intervals indicating model uncertainty. Overall, it is found that the developed machine learning framework shows good agreement with the industry standard underlining its plausibility.
This paper presents an assessment of three methods used for sea state estimation via the wave buoy analogy, where measured ship responses are processed. The three methods all rely on Machine Learning exclusively but they have different output; Method 1 provides bulk parameters, Method 2 yields a point wave spectrum and the wave direction, while Method 3 gives the directional wave spectrum in non-parametric form. The assessment is made using full-scale data from an in-service container ship in cross-Atlantic service. Training and testing of the methods are made using data from a wave radar, and the three methods perform well. An uncertainty measure, equivalently, a trust level indicator, based on the variation between the post-processed outputs of the methods is proposed, and this facilitates determination of estimates with small errors; without knowing the ground truth.
Background: Autonomous ships have the potential to increase operational efficiency and reduce carbon footprints through technology and innovation. However, there is no comprehensive literature review of all the different types of papers related to autonomous ships, especially with regard to their integration with ports. This paper takes a systematic review approach to extract and summarize the main topics related to autonomous ships in the fields of container shipping and port management. Methods: A machine learning method is used to extract the main topics from more than 2000 journal publications indexed in WoS and Scopus. Results: The research findings highlight key issues related to technology, cybersecurity, data governance, regulations, and legal frameworks, providing a different perspective compared to human manual reviews of papers. Conclusions: Our search results confirm several recommendations. First, from a technological perspective, it is advised to increase support for the research and development of autonomous underwater vehicles and unmanned aerial vehicles, establish safety standards, mandate testing of wave model evaluation systems, and promote international standardization. Second, from a cyber–physical systems perspective, efforts should be made to strengthen logistics and supply chains for autonomous ships, establish data governance protocols, enforce strict control over IoT device data, and strengthen cybersecurity measures. Third, from an environmental perspective, measures should be implemented to address the environmental impact of autonomous ships. This can be achieved by promoting international agreements from a global societal standpoint and clarifying the legal framework regarding liability in the event of accidents.
Harbour vessel emissions are growing concerns in the maritime industry regarding environmental sustainability. Accurate emissions prediction can stand in monitoring and addressing the issue. This study proposes a machine-learning approach using Artificial Neural Network (ANN) for predicting harbour vessel emissions. The approach shows superiority over the bottom-up method introduced by the 4th IMO GHG Study regarding prediction accuracy. Actual emissions data from onboard measurements are used for training ANN models and as references for evaluating the methods. Compared to the bottom-up method, the improvement in error reduction can be up to 30% for predicting nitrogen oxides and 54% for carbon monoxide when only using ship-related factors as input variables. By adding selected meteorological factors in the experiments, the prediction accuracy enhancement can achieve up to 48% for nitrogen oxides and 62% for carbon monoxide. The proposed ANN approach could assist relevant stakeholders in improving emissions prediction and operations optimisation.
We present a hybrid linear potential flow - machine learning (LPF-ML) model for simulating weakly nonlinear wave-body interaction problems. In this paper we focus on using hierarchical modeling for generating training data to be used with recurrent neural networks (RNNs) in order to derive nonlinear correction forces. Three different approaches are investigated: (i) a baseline method where data from a Reynolds averaged Navier Stokes (RANS) model is directly linked to data from an LPF model to generate nonlinear corrections; (ii) an approach in which we start from high-fidelity RANS simulations and build the nonlinear corrections by stepping down in the fidelity hierarchy; and (iii) a method starting from low-fidelity, successively moving up the fidelity staircase. The three approaches are evaluated for the simple test case of a heaving sphere. The results show that the baseline model performs best, as expected for this simple test case. Stepping up in the fidelity hierarchy very easily introduces errors that propagate through the hierarchical modeling via the correction forces. The baseline method was found to accurately predict the motion of the heaving sphere. The hierarchical approaches struggled with the task, with the approach that steps down in fidelity performing somewhat better of the two.