Low-Temperature Polymer Electrolyte Membrane Fuel Cells (LT-PEMFCs) have recently emerged as a promising solution for sustainable ship energy systems. However, enhancing durability is essential to enable their broader adoption in the maritime sector. Durability enhancement depends on a thorough understanding of degradation mechanisms and accurate prognostics, both of which are highly application-specific. The current literature lacks a comprehensive understanding of LT-PEMFC degradation under maritime operating conditions and its integration into reliable prognostic models. To address this gap, this review provides an overview of LT-PEMFC durability and prognostic models from the perspective of maritime applications. Through a comparative analysis of studies across various sectors, we identify and discuss maritime-specific degradation drivers, including ship load profiles, sodium chloride contamination, vibrations, and wave-induced inclinations. Building on this analysis, we critically evaluate existing prognostic models and their suitability for lifetime prediction in maritime applications. This review proposes durability enhancement strategies based on current knowledge and highlights key research gaps requiring further investigation. In addition, it outlines promising prognostic methodologies and identifies technical challenges for application to maritime LT-PEMFCs. In this way, this work lays the foundation for enhancing LT-PEMFC durability in maritime environments and supporting its broader adoption for zero-emission ships.

The maritime industry is transitioning from traditional fossil fuel-based power and energy systems to cleaner and more efficient energy practices. Battery adoption is driven by three main factors: regulations, techno-economic development, and improved infrastructure for onshore charging with new methods of charging batteries for maritime applications. While regulations have driven the maritime industry toward sustainable energy practices, advancements in technology and large-scale battery adoption have made a variety of battery chemistries more accessible, affordable, and longer lasting. This has resulted in the adoption of batteries not just as a means to meet regulations but also as an alternative technology in combination with traditional fossil fuels. This means that having batteries on board offers several advantages, such as reducing emissions and the operational hours of other onboard power sources, among many others. The benefits of batteries on board are categorized into environmental and operational performance, covering a total of 11 distinct advantages. The wide variety of ship types and operations means these benefits are realized in diverse ways, with batteries used differently across various vessel types. A total of 15 distinct maritime battery functions are identified, each with its benefits and applications within the maritime industry. These functions range from using batteries in a battery electric mode, where batteries serve as the sole power source on board, to acting as a spinning reserve during dynamic positioning applications. A hypothetical tugboat system is used to demonstrate various battery functions. This article discusses maritime batteries’ drivers, benefits, and functions, highlighting their role in the maritime industry.

In recent years, Artificial Intelligence, particularly Machine Learning, has achieved remarkable success in solving complex problems. However, this progress has also revealed the emergence of unexpected, poorly understood, and elusive phenomena that characterize the behavior of machine intelligence and learning processes. These phenomena often challenge researchers to interpret them within the boundaries of existing Machine Learning theoretical frameworks, thereby motivating the development of new and more comprehensive theoretical foundations. One such phenomenon, known as grokking, refers to the sudden and substantial improvement in a model's performance following a prolonged period of stagnant or even regressive learning. In this paper, we argue that it is possible to provide insights into grokking by leveraging the existing theoretical foundations of Machine Learning, in particular concepts from Statistical Learning Theory, such as norm-based and stability-based generalization bounds. We further show how these theories can help reconcile the phenomenon of grokking with established principles of learning and generalization. Furthermore, we demonstrate the practical applicability of these insights through concrete examples.

As navies and maritime organisations transition towards low-emission propulsion systems, spark-ignited (SI) gas engines capable of operating on sustainable, low-reactivity fuels are gaining renewed interest. These engines, while offering potential for fossil-free operation, present significant challenges under transient conditions due to complex interactions between throttle control, fuel regulation, and combustion stability. Accurate dynamic modelling is critical to integrate these engines into resilient naval power systems and to support the development of advanced control strategies. This study evaluates several high-fidelity mean value first principle engine modelling (MVFPEM) approaches for simulating the dynamic gas path behaviour of a large, high-speed, SI marine engine under rapid load changes. Models with varying levels of complexity, including simplified and full turbocharger implementations and different gas path volume resolutions, were calibrated using a single measurement campaign and validated against measured transient data. Several methods for turbocharger performance mapping (Stapersma, Casey & Robinson, and Jensen) were evaluated for their applicability in predicting the engine behaviour in dynamic operating scenarios. The results highlight that models incorporating three control volumes and full turbocharger dynamics achieve the highest accuracy, particularly during rapid load increases and recovery phases. Simplified models fail to capture turbocharger inertia and pressure transients, limiting their applicability to investigate naval propulsion or electric power generation plant behaviour under transient load conditions. This work provides guidance on selecting and validating engine models for marine applications and reinforces the role of high-fidelity MVFPEMs in the design and simulation of future naval energy systems.

Modern power grids are becoming increasingly complex with the integration of heterogeneous distributed energy resources, underscoring the need for accurate and efficient Power Flow Analysis to ensure stability, reliability, and market operations. Existing methods generally rely on iterative numerical techniques (INT) or machine learning (ML). While INT is physically consistent and highly accurate, it can be computationally expensive and vulnerable to slow or non-convergence. ML methods offer faster solutions but often require extensive data, suffer from limited extrapolation capabilities, and lack physical consistency. Physics-informed ML (PIML) bridges these gaps by embedding domain knowledge before, during, and after training. However, current PIML approaches typically do not leverage this full range of opportunities. In this paper, we propose a novel PIML framework for Power Flow Analysis that integrates physical insights at all three stages (pre, in, and post-processing) to achieve superior accuracy and efficiency. Notably, we introduce a new post-processing technique that partitions the power network into its mesh and radial components: the mesh portion is handled via PIML, while the radial portion is efficiently solved with a convex optimization approach informed by the PIML outputs. This approach is efficient with radial topologies, especially in power distribution networks where the radial part is predominant. Experiments on realistic power networks demonstrate that our method outperforms state-of-the-art approaches in both accuracy and computational performance.

This study presents a framework for designing and optimizing ship energy systems including weather-driven speed variability and navigation safety constraints. Navigation risks including resonance, surf-riding, and successive high-wave impacts, are calculated using five years of hourly weather data. Random speed variations (up to ±5%) are applied to a baseline speed profile to capture operational uncertainty, and safety-based speed reductions (up to 40%) are applied when required. Course changes are excluded. Treating navigation risks as constraints, operating profiles are generated for different weather conditions. For a conceptually retrofitted cargo ship, hydrogen fuel cell and battery capacities, and their power distribution, are optimized for each operating profile to minimize lifetime energy system cost and assess the effects of weather-induced power variation. Results show that speed and weather variability can significantly change power demand, requiring fuel cell capacities between 700 and 1500 kW. The most common configuration is a 1200 kW fuel cell system with 180 kWh of battery capacity, covering 39% of laden profiles, while full power coverage requires 1500 kW. Lifetime cost outcomes exhibit a 5th–95th percentile spread of −10.3% to +11.1% relative to mean cost. The results demonstrate the significant influence of weather variability on system sizing and cost.

Low total lifetime cost is essential for the adoption of zero-emission ship energy systems, which must meet operational power demands while complying with onboard safety regulations. However, many studies rely on a simplified, averaged or insufficiently representative load profile and treat system design, operation, and integration feasibility separately, which can distort lifetime cost assessments and result in practically infeasible retrofit concepts. This study investigates how a hydrogen-based ship energy system can be optimally sized, operated, and arranged onboard to minimize total lifetime cost while satisfying operational constraints and stability requirements for a general cargo vessel retrofit. A representative power profile is synthesized from one year of operational data using a probability-based downsampling method and then used in a mixed-integer nonlinear lifetime cost optimization with discrete placement and ballast decisions, solved using the SCIP solver. The optimal retrofit comprises 1.4 MW of fuel cells, 180 kWh of batteries, and a 146 m³ liquefied hydrogen (LH₂) tank, requires 171 t of ballast to satisfy trim and vertical stability constraints, and is primarily driven by fuel costs, which account for 74% of the total lifetime cost. Overall, the results indicate that the viability of hydrogen-based ship retrofits primarily depends on LH₂ storage integration constraints and hydrogen price assumptions, and that the proposed framework provides a practical basis for lifetime cost assessment of feasible retrofit designs.

As Polymer Electrolyte Membrane Fuel Cells (PEMFCs) emerge as a promising technology for transport decarbonization, the development of durability assessment protocols tailored to specific applications, such as maritime operations, is becoming relevant for the identification of stressors and lifetime enhancement. This study presents a preliminary experimental campaign aimed at introducing a methodology to assess the degradation of PEMFCs subjected to Accelerated Stress Test (AST). In particular, the methodology encompasses the utilization of electrochemical characterization and, in this work, the fuel cell operating profile has been chosen to mimic the operation of a small passenger vessel. The tests were carried out on two single Membrane Electrode Assemblies (MEAs) for 500 h. One membrane was subjected to the AST, and a second sample, tested under constant load operation, served as a reference. Periodic electrochemical characterization was conducted to assess performance degradation through polarization curves, electrochemical impedance spectroscopy, and cyclic voltammetry. The electrochemical analysis of degradation was conducted through a dual-method approach combining model-free and model-based methods for impedance analysis, as well as catalyst active area evaluation from voltammograms. Results show that the dynamic operation characteristic of the passenger ferry increases degradation compared to constant operation, evidenced by increased ohmic and interfacial resistances and losses in catalyst active area. This work provides a framework for developing application-specific durability protocols, enriched with multi-method diagnostic approaches to assess PEMFC degradation under realistic maritime conditions. Such methodologies support the development of durability enhancement strategies tailored to maritime applications, allowing a broader application of the technology in the sector.

Batteries have emerged as a promising solution across diverse vessel segments, offering benefits in operational efficiency, cost reduction, and emissions reduction. This study investigates the specific requirements of batteries onboard 7 vessel types, such as tugboats, ferries, cruise ships, yachts, fishing vessel, vessels with cranes, and dynamic positioning vessels, through an in-depth analysis of load profiles and operational needs. By identifying 24 potential operational requirements, ranging from battery electric operation to silent operations and load smoothing, a mixed-integer linear programming model is used to optimize the power and energy allocation for each requirement. This framework enables a generalization of battery requirements for various vessel segments and enables the assessment of three lithium-ion battery chemistries: Lithium Iron Phosphate, Nickel Manganese Cobalt Oxide, and Lithium Titanate Oxide. The results indicate that different vessel types prioritize either high energy density batteries or those capable of delivering high power relative to energy capacity. To guide battery selection, a decision tree is presented that matches battery types with specific vessel needs. Lithium Titanate Oxide batteries are well-suited for applications requiring frequent, high power cycles, especially where fast charging is needed. Lithium Iron Phosphate batteries are best for energy-intensive operations, while Nickel Manganese Cobalt Oxide batteries perform well in both high power and high energy applications. This study offers a practical approach, an inventory of battery requirements, and guidance on selecting the chemistries best suited to various vessel types and operational needs.

The dynamic behaviour of floating offshore wind turbines (FOWTs) involves complex interactions of multivariate loads from wind, waves, and currents, which result in complex motion characteristics. Although methods for analysing global motion responses are well-established, the time- and location-dependent kinematics remain underexplored. This paper investigates the instantaneous centre of rotation (ICR), a point of zero velocity at a time instance of general plane motion. Understanding and strategically positioning the ICR can reduce the dynamic motion in critical structural locations, enhancing the performance and structural robustness of FOWTs. The paper presents a method for computing the ICR using time-domain simulation results and proposes a statistical analysis approach suitable for design studies. Building on prior research, it examines the sensitivity of the ICR to external loading and design features, providing insights into how these factors influence motion response and how the motion response influences the statistics of the ICR, structural loads, and other performance metrics of interest. The study explores two FOWT configurations, a spar and a semisubmersible, identifying design variables that most effectively control the ICR statistics and identifying the ICR statistics most correlated with the responses of interest. Finally, through two case studies, we demonstrate how to apply these new insights in a practical design scenario. By adjusting the design variables most correlated with the ICR (fairlead vertical position and centre of mass for the spar and mooring line length and offset column diameter for the semisubmersible), we successfully modified the designs of the floating support structures to reduce the loads in the mooring lines, tower base, and blade roots, improving the ultimate strength and fatigue characteristics compared to the original designs.

Methanol is a promising alternative fuel, which can assist in reducing emissions in heavy-duty (HD) dual-fuel (DF) compression ignition (CI) engines. In medium and large bore marine engines, DF operation is achieved through either direct injection (DI) or port fuel injection (PFI) of methanol with diesel acting as a DI pilot fuel for ignition. However, the injection of methanol presents a significant challenge due to its high latent heat of vaporization and decreased lower heating value (LHV) compared to diesel. Therefore, for the same energy content operation, methanol requires around eight times the amount of heat to evaporate completely in comparison to diesel, which results in lower in-cylinder temperatures. This charge cooling effect leads to a strong negative temperature gradient influencing ignition and flame propagation. This paper aims to quantify the cooling effect of methanol in a heavy-duty dual-fuel direct injection compression ignition (DICI) engine environment. The presented methodology uses computational fluid dynamics (CFD) simulations to model methanol sprays with validation originating from the engine combustion network (ECN) Spray D experimental data. The CFD models operate within the Lagrangian–Eulerian framework in CONVERGE-CFD using the Reynolds Averaged Navier Stokes (RANS) turbulence modeling. Compared to diesel, injecting methanol with the same energy content exhibited up to 100 K more decreased temperature within the mixture. Consequently, this cooled mixture may pose challenges to combustion stability due to the intense temperature gradients. Nonetheless, lower mixture temperature decreases NOx emissions, which can prove beneficial for high methanol energy fractions in dual-fuel DICI engines.

A key factor towards zero-emission shipping is the adoption of electric propulsion with hybrid power sources. The heterogeneous power sources of modern electric vessels require optimal energy management systems, as conventional rule-based control in hybrid energy systems may result in suboptimal solutions with limited flexibility. Advanced optimal control strategies offer a promising avenue to address this issue. This paper presents a novel control strategy based on the Equivalent Consumption Minimization Strategy for a dual-fuel full-electric vessel operating with diesel engines and hydrogen fuel cells taking into account both fuel cost and NOx emissions. The effectiveness of the developed controllers is evaluated against a benchmark derived from state-of-the-art strategies in a simulation study using real-world data. The results highlight the controller's performance, as well as the operator's choice by selection of weights for the objectives. The proposed control strategy achieves nearly 2 % fuel savings compared to a single-objective rule-based controller. It also exploits the potential for up to 45 % reductions in NOx emissions. When both objectives are combined, the controller still delivers over 0.5 % fuel savings while reducing NOx emissions by nearly 15 %. If a financial cost is assigned to emissions, the total operational cost savings increase to more than 4 %.

Methanol is considered an alternative fuel for the shipping decarbonisation, the properties of which, however, impact the marine dual-fuel engines ignition and combustion characteristics, especially at low load conditions. This study aims at parametrically optimising a marine dual-fuel engine operating with methanol high energy fraction at low loads to achieve knock-free combustion with the highest efficiency and lowest emissions. Computational Fluid Dynamics (CFD) modelling in the CONVERGE software is employed for the investigated large-bore marine four stroke engine considering four injection strategies including single, two stage and stratified injection. The Reynolds Averaged Navier Stokes (RANS) approach is employed to represent turbulence, the Lagrangian-Eulerian approach is used for the spray formation, and the SAGE detailed chemistry solver is used for modelling combustion. The CFD model was first developed and validated for the engine diesel mode. Subsequently, the validated model was expanded to accommodate the direct injection (DI) of both methanol and diesel fuels. Parametric runs are performed considering the compression ratio (CR) in the range 14–17 and the temperature range at inlet valve closing (T_IVC) 360–400 K. The results reveal that acceptable combustion efficiency and high thermal efficiency are achieved with CR and T_IVC above 17 and 380 K respectively for single injection, above 16 and 380 K respectively for double injection, as well as above 14 and 360 K respectively for stratified injection. Stratified injection is proposed to improve engine performance and reduce NOx emissions. This study provides insights to achieve stable and efficient operation of methanol-fuelled marine engines at low loads, and as such it contributes to the maritime industry decarbonisation.

Relying on a traffic simulator is often necessary when multiple traffic conditions need to be replicated, either to predict specific quantities of particular interest or to assess more general network properties, such as efficiency or resilience, under different scenarios. While potentially delivering a high level of fidelity, producing such a simulation may come at a prohibitive computational cost when it is applied in a real context, making this approach unsuitable for real-time applications in most cases. In this regard, the goal of the present work is twofold. Firstly, we aim to surrogate a traffic simulator with a data-driven approach in order to produce real-time traffic predictions that are also sufficiently accurate and effective. Secondly, we want to determine the minimum amount of data, i.e., the smallest number of sensors deployed on the network, that still allow to obtain predictions within a predefined bound. The effectiveness of the approach is evaluated on the full-scale urban network of Rapallo, Italy, in which we employ the AIMSUN NEXT simulator targeting the morning peak hours, i.e. between 7:00 a.m. and 9:00 a.m. In the paper, multiple state-of-the art ML algorithms are tested to assess their effectiveness as surrogate models under the considered problem.

The electrification of ship power systems plays a center role in the mobility transition towards sustainable transport solutions. It allows the integration of various power sources, energy storage systems, and intermittent generation. The integration of an increasing number of components with distinct characteristics shapes the notion of a shipboard microgrid which benefits from a modular approach in its design to reduce costs and uncertainties. DC distribution facilitates the modular design by simplifying the control, and, combined with power electronics interfaces, increases the controllability of power flows in the system. To handle the increasing system complexity, this work proposes a distributed and predictive control approach, addressing the modular topology of future shipboard power systems and leveraging load power forecasting. Investigations show that a distributed, predictive energy management reaches a similar performance as a centralized implementation. For a modular shipboard power system, the proposed method decreases both fuel and degradation costs with increasing performance gains for longer prediction horizons.

In contemporary industrial applications, predictive models have been pivotal in bolstering production efficiency, product quality, scalability, and cost-effectiveness while promoting sustainability. These predictive models can be constructed solely based on domain-specific knowledge, exclusively on observational data, or by amalgamating both approaches. They are commonly referred to as Full-, Zero-, or Partial-knowledge-based predictive models, respectively. Full-knowledge based models are highly explainable, point-wise accurate, data efficient, computationally demanding in the prediction phase to achieve high accuracy. Zero-knowledge based models are poorly explainable, data hungry, highly accurate on average, computationally demanding in the construction phase but inexpensive the prediction phase. Partial-knowledge based models focus on taking the best of the two worlds. To maintain a focused scope and avoid redundancy with existing literature, our review will primarily delve into the key industrial applications within a narrow context, encompassing sectors such as extraction, chemical processes, manufacturing, transportation, energy, and construction. Our approach entails several steps. Initially, we conducted a meta-review to pinpoint gaps in previously published surveys on Full-, Zero-, and Partial-knowledge-based predictive models for industrial applications. Subsequently, we present a formal analysis of the subject matter, supplemented with illustrative examples to offer valuable insights. The core of our work comprises a review of existing research categorized by specific industrial applications. Finally, we outline the unresolved challenges and future prospects in this burgeoning field of research. We contend that our work serves as a valuable resource, catering to the needs of both young researchers seeking a solid foundation for their studies, industrial practitioners aiming to grasp core concepts and applications, and senior researchers seeking potential real-world applications for their findings.

Optimizing the hull form is crucial for enhancing vessel performance, requiring a careful exploration of design variations once an initial concept is chosen. To address this, a well-defined parameterization and search space must be established, followed by a numerical optimization process. This phase involves balancing three key requirements: improving performance indicators like resistance (and consequently emissions), ensuring physical plausibility (e.g., designs that meet stability standards), and maintaining computational efficiency. While existing methods partially address these challenges using surrogate models to optimize performance, conducting a-posteriori checks for physical plausibility, and reducing the search space a-priori, they fall short of fully integrating these aspects. In this work, we introduce a two-fold approach to address these challenges. First, we integrate a key physical plausibility criterion directly into the optimization loop: the International Maritime Organization (IMO) Intact Stability Code. For the first time, we use a surrogate model to include this constraint within the numerical optimization process. This enables the generation of designs that inherently meet stability requirements. Second, we reduce the computational demands by applying data-driven methods to limit the search space in a problem-specific manner. This targeted reduction reduces the computational load without compromising the overall optimization performance. We test our proposal by optimizing the KCS hull form. Our results demonstrate two main achievements. First, we find that current optimization pipelines fail to meet the IMO stability guidelines, whereas our method achieves compliance by design. Second, our approach reduces the computational burden by 30 % without sacrificing performance, representing a significant leap forward in practical hull design optimization.

The electrification of shipboard power systems carries an increasing variety in power sources, energy storage systems, and power converters. DC distribution is gaining relevance due to efficiency increase, space savings, and high controllability. The dominant primary control method voltage droop control, which offers easily implementable and scalable power sharing and voltage stabilization. However, compared to terrestrial microgrids, shipboard power systems have low line impedances and highly fluctuating loads. Most primary loads are power-controlled, introducing a non-linearity that leads to a weak damping and unstable operation points. To handle this non-linearity, this study proposes an energy-based control approach as an alternative to the voltage-based scheme. The controller is further extended by integral feedback loop to achieve fast voltage restoration. A low-bandwidth communication is leveraged for an adaptive power sharing control that facilitates an efficient allocation of the load among parallel units under varying conditions. The proposed control structure is deployed on an I/O board embedded in a hardware-in-the-loop testbed. It is shown that the energy-based controller operates stably and achieves a reduced voltage deviation from the nominal voltage in various load conditions compared to the conventional voltage-based method.

Renewably produced methanol is a promising fuel for internal combustion engines in long-range transportation thanks to its scalability, liquid storage, and favorable combustion properties. However, the distinction between different injection and ignition strategies for methanol engines and the resulting combustion mechanisms has not been consistently defined. Moreover, diffusion combustion strategies are favored over premixed strategies in large engines because of higher methanol energy fractions, disregarding the advantages of premixed approaches, such as reduced nitrogen oxide emissions and retrofitting opportunities. To address ambiguity in terminology, this paper proposes a classification framework for injection and ignition strategies and applies it to methanol-fueled internal combustion engines. Subsequently, this review focuses on experimental studies of methanol-fueled heavy-duty engines, which are crucial for transitioning to renewable and sustainable energy in long-range transportation. This research summarizes the impact of the reviewed injection and ignition strategies on combustion characteristics, engine performance and emissions to identify key trends. Furthermore, this review highlights how specific design and operating parameters influence premixed dual-fuel combustion, offering insights into optimizing performance and emissions. While mono-fuel and premixed dual-fuel strategies with methanol can significantly promote methanol use in heavy-duty engines and reduce harmful emissions like nitrogen oxides, a rise in unburned hydrocarbon emissions may also be expected, necessitating further research in this area. Additionally, methanol injection location in premixed dual-fuel schemes affects its cooling effect, influencing volumetric and thermal efficiency. Overall, this study deepens our understanding of methanol's impact on heavy-duty engine performance, highlighting critical challenges to be addressed for advancing sustainable transportation.