Simulation-Based Energy Optimization Through Maneuvering Prediction for Complex Passenger Ships: Results from the SimPleShip-SigMa Project

Abstract

The decarbonization of shipping and the transformation towards digitally assisted or automated ship operation require new methods to analyze, predict, and optimize energy demand during maneuvering. The SimPleShip-SigMa sub-project of Hochschule Wismar developed and validated a comprehensive simulation-based framework combining real-time capable fast-time simulation of ship motion, detailed thermodynamic engine modeling, and hybrid data exchange via Functional Mock-up Units (FMU/FMI). The approach enables consistent coupling between navigation-related and machinery-related simulations and supports energy-optimized decision-making on the bridge. Operational relevance and validation of use cases were supported through collaboration with Carnival Maritime GmbH, providing practical feedback on large passenger-ship operations. The study presents the architecture of the simulation environment, the implementation of energy- and emission-prediction models, and the result of validation runs and simulator-based trials. The developed method was applied to a virtual cruise-ship scenario representing a confined coastal environment similar to the Geiranger Fjord. The work builds upon earlier research on simulation-augmented maneuvering and extends it toward a modular digital-twin concept linking hydrodynamic and thermodynamic models. The paper concludes with an outlook on applying the system for crew training, on-board support, and gradual automation of sustainable ship operations.

1. Introduction

1.1. Motivation

Maritime transport is facing the simultaneous challenges of environmental protection, economic efficiency, and workforce transformation. Although international shipping contributes only about 2–3% of global CO₂ emissions, local effects, particularly in harbor and coastal areas, are substantial [1,2,3,4]. The International Maritime Organization (IMO) is undertaking strong action to meet ambitious goals to reduce emissions [5,6,7] and European Commission is supporting those activities further through its own initiatives [8,9]. Among the manifold measures to reduce fuel consumption and improve energy efficiency [10,11], in parallel, the introduction of automation and digitalization technologies is transforming bridge and engine-room operations [12]. A key opportunity lies in integrating these developments to realize both energy-efficient and safe maneuvering practices.

For ships already in service, retrofitting with new propulsion or fuel technologies is often economically infeasible [9,13,14]. Therefore, operational measures that optimize existing systems provide immediate potential for emission reduction [15]. Such measures include speed optimization, trim and draft adjustment, or the optimization of maneuvering procedures [16,17]. Previous studies have shown that intelligent use of propulsion systems can lead to significant energy savings without structural modifications [18,19,20,21,22].

1.2. From Assistance to Automation

In regard to design and operation of engine and propulsion systems installed onboard, a wide range of assistance systems to support decision-making and enhanced systems automation have been applied to support energy efficient ship operation, i.a., [23,24,25,26,27]. Over the past decade, the Institute for Innovative Ship Simulation and Maritime Systems (ISSIMS) at Hochschule Wismar, together with ISSIMS GmbH, have developed simulation-based tools that support this operational optimization. Early developments, such as the Simulation Augmented Maneuvering Design and Monitoring (SAMMON) system, based on Rapid Advanced Prediction and Interface Technology (RAPIT) enabled fast-time prediction of ship motion for training and planning [28]. The German research project MEmBran (Modeling Emissions and Fuel Consumption during Ship Maneuvers) extended these concepts by linking them with detailed engine-process models to calculate fuel consumption and pollutant emissions during transient maneuvers [16,29].

Building on these foundations, the SimPleShip (Simulation Platform for Digital System Analysis and Energetic Operational Optimization of Complex Passenger Ships) aimed to develop an integrated digital framework that connects ship-handling simulators, engine-room simulators, and analytical models through standardized interfaces. Its sub-project, SigMa (“Simulation and integration of power demands of the ship propulsion system during maneuvering and transit”), led by Hochschule Wismar, focused on the simulation-based prediction and optimization of maneuvering energy demand.

1.3. Research Aim and Contribution

The objective of this work is to describe the methodological advances and achieved results of the SigMa sub-project, situating them within the broader scientific discourse on energy-optimized and automated maneuvering. The paper provides:

1.

An overview of the research framework and developed simulation architecture;

2.

A description of the integrated models and interfaces to connect nautical and technical domains;

3.

Empirical and simulation-based results demonstrating achievable efficiency gains;

4.

A discussion of implications for automation, crew training, and decarbonization strategies.

2. Background and Related Work

2.1. Operational and Simulation-Based Energy Optimization

Numerous research initiatives and technological development activities have addressed energy efficiency in ship operation, distinguishing between technical and operational measures [30,31,32,33,34,35,36,37]. While technical approaches involve propulsion redesign, hybridization, or the use of alternative fuels [38,39,40,41], operational measures not only focus on technical issues but merely look into operational measures related to, among others, improved navigation, trim optimization, or efficient maneuvering strategies [42,43,44]. Bouman et al. (2017) reviewed 150 studies and reported that operational optimization alone can achieve up to 60% CO₂ reduction per transported unit [45,46]. This includes speed- and weather-optimized routes, as well as other operational measures. The significance of maneuvering phase, especially in coastal and port environments, has been repeatedly emphasized, as engines operate under highly transient conditions with increased emissions of NO_x, SO_x, and soot [22]. Complementary to simulation-based planning, recent work proposes lightweight ship–shore data platforms that cleanse, classify, and report operational efficiency (inclunding CII) for smaller operators [46,47,48]. Data-classification/IoT management frameworks validated against AIS and machinery logs outline an organizational layer for ongoing efficiency supervision on board [49,50].

Simulation has emerged as an essential tool to analyze and improve these operational regimes. The Fast-Time Simulation (FTS) concept enables prediction of ship motion several minutes ahead, supporting decision-making during complex maneuvers. When combined with models of engine processes, this allows for the prediction of not only trajectories but also of associated fuel and emission outcomes.

2.2. Maneuvering Assistance Systems

The integration of aspects of energy efficiency into ship handling and maneuvering is still a rather young trend in research and technical developments. Applications are still in a rather early phase and most of them are in connection to the use of dynamic positioning equipment [50,51]. The SAMMON (Simulation-Augmented Maneuvering Design and Monitoring) software developed at ISSIMS introduced an intuitive human–machine interface connecting mathematical ship models with navigational displays. Using the underlying RAPIT kernel, SAMMON performs highly accelerated dynamic simulations of ship motion, achieving a speed-up of up to 1000× real time. Its modules, Planning and Monitoring and Conning, have been adopted in maritime education and by several fleet operators.

Earlier simulator experiments indicated notable energy savings when maneuvers were pre-planned and executed with SAMMON support and demonstrated substantial energy savings through prediction-based maneuvering. In controlled simulator trials with experienced navigators, energy demand at the propellers was reduced by up to 35% when maneuvers were pre-planned and executed with SAMMON support [19,20,52]. These findings underline the direct link between human decision-support and sustainable operation [22,53,54].

2.3. Integration of Engine and Emission Models

The MEmBran project (Modeling Emissions and Fuel Consumption during Ship Maneuvers) [22] provided the basis for coupling ship-handling and engine simulations. Using testbed measurements from medium-speed diesel engines, detailed thermodynamic models were developed to describe transient combustion, fuel injection, and pollutant formation. Both physical and data-based model approaches were implemented. The latter employed artificial-neural-network (ANN) architectures trained with measurement data to predict NO_x, PM, and CO₂ emissions during transient load changes. These models achieved sufficient accuracy for integration into FTS environments, enabling near real-time emission forecasts [53,54].

2.4. Gradual Automation of Maneuvering

Parallel research outside the SimPleShip project at the University of Rostock and Hochschule Wismar explored the transition from assisted to automatic maneuvering. The concept employs a gradual automation chain, beginning with human-in-the-loop prediction support and progressing toward full trajectory control. Experimental work with the research vessel DENEB (Federal Maritime and Hydrographic Agency, BSH) demonstrated automatic berthing using model-based control strategies while preserving manual override options. These studies illustrate how assistance technologies can serve as intermediate steps toward autonomous and energy-optimized operation [22,54].

2.5. Recent Developments and Outlook

The latest investigations summarize the state of multiple-prediction and step-ahead simulation concepts [55,56]. Concurrently, voyage-level studies show that under time-varying environmental and regulatory conditions, including wind, waves, currents and ECAs, multi-objective co-optimization of speed, route, and trim provides the operational backdrop against which multiple- and step-ahead maneuver predictions should be assessed [57,58]. These methods extend single-prediction approaches by allowing several alternative maneuver plans to be computed and compared simultaneously. The SimPleShip framework builds upon these developments by coupling them with energetic and emission-based evaluation functions, thereby closing the loop between maneuvering design and energetic optimization.

3. Objectives and Research Framework of SimPleShip-SigMa

3.1. Project Context

The joint research project SimPleShip was funded by the German Federal Ministry for Economic Affairs and Climate Action (BMWK) under the coordination of the FVTR GmbH. Partners included the University of Rostock and Hochschule Wismar. Within this consortium, the sub-project SigMa focused on the integration of nautical and machinery simulations for operational energy optimization.

3.2. Scientific Objectives

The main scientific objectives of SigMa within SimPleShip were:

• To integrate dynamic ship-handling models into a comprehensive virtual full-ship simulation environment enabling consistent analysis of maneuvering and energy processes.
• To establish an FMI-based framework for coupling hydrodynamic ship models with thermodynamic and engine models provided by the project partners FVTR and the University of Rostock—Chair of Technical Thermodynamic (LTT).
• To analyze and validate the coupled simulation results using real-ship measurement data and representative operational scenarios, including cruise-ship maneuvers in restricted and fjord environments.
• To demonstrate the potential of simulation-based energy optimization for maritime operations through simulator studies and workshops involving nautical experts and ship operators.

3.3. Conceptual Framework

Figure 1 illustrates the conceptual architecture of SimPleShip. At its core, the framework combines three interacting layers:

• The Nautical Layer, representing ship motion dynamics, control surfaces, and environmental influences;
• The Technical Layer, describing engine, propulsion, and auxiliary systems through thermodynamic process models;
• The Data and Interface Layer, facilitating communication between both domains using FMU/FMI standards for model exchange and co-simulation.

This architecture allows both offline and online modes: offline for planning and analysis, online for real-time assistance or hardware-in-the-loop applications.

3.4. Use-Case Definition

To ensure practical relevance, the consortium selected a cruise-ship reference scenario modeled after operations in the Geiranger Fjord, Norway. The scenario includes confined-water navigation, dynamic environmental influences, and a combination of diesel-electric propulsion with azimuth thrusters. It provides an ideal testbed to study the trade-offs between safety, time, and energy efficiency during complex maneuvers.

Additionally, representative engine-room configurations were modeled for hybrid propulsion systems combining diesel-generator sets with battery modules, enabling exploration of future low-emission operational strategies by our project partners.

3.5. Expected Impact

By linking realistic simulator environments with predictive models, the project aimed to deliver:

• A validated digital twin of maneuvering and machinery behavior for complex passenger ships;
• Quantifiable reductions in energy demand and emissions during maneuvering;
• Improved training tools that raise crew awareness of energy efficiency and a foundation for semi-autonomous energy management in forthcoming vessel generations.

4. Methodology

4.1. Simulation Environment and Tools

The SimPleShip framework was implemented using the simulation infrastructure available at the Maritime Simulation Centre Warnemünde (MSCW) of Hochschule Wismar. This infrastructure combines two core simulator types:

• Full-Mission Ship-Handling Simulator (SHS)—a DNV-certified system reproducing realistic bridge configurations and the hydrodynamic behavior of various ship types. It allows the execution of complex maneuvering scenarios under environmental influences for both research and training purposes.
• Engine and Power-System Simulation Environment—developed and provided in collaboration with FVTR GmbH and the University of Rostock (LTT), comprising thermodynamic and hybrid-energy models implemented in a Modelica-based framework. These models enable simulation of energy conversion and charge processes and were coupled with the hydrodynamic ship models through FMI/FMU interfaces.

In addition, SAMMON (Simulation Augmented Maneuvering Design and Monitoring) software was used to visualize and analyze maneuvering processes in offline mode. It enables fast-time simulation and prediction of ship trajectories and ship states, thus supporting the planning, evaluation, and optimization of maneuvers under different environmental and technical conditions.

Both simulation environments—SAMMON and the Modelica-based engine framework—are founded on data-based and experimentally verified ship models. Their parametrization relies on extensive measurement campaigns, including real-ship trials and simulator-based validation, ensuring high physical fidelity and reproducibility of hydrodynamic and thermodynamic processes. This provides a scientifically robust basis for energetic assessments and for analyzing the effects of operational strategies on fuel consumption and emission.

4.2. Data Sources and Measurement Integration

4.2.1. Shipboard Data Acquisition

To ensure realistic simulation parameters and boundary conditions, operational data were obtained from a cruise ship used as the testbed. The initial data acquisition concept was originally developed using the hybrid ferry Berlin. The ship was chosen because of its comprehensive automation and monitoring infrastructure, allowing the acquisition of representative measurement data under real operational conditions. From our project partners and the ship operator, a comprehensive dataset originated from the on-board automation system (IMAC/IAS) was made available. The raw dataset comprised more than 13,000 signal channels covering propulsion systems, energy management, navigation, and auxiliary machinery. However, a direct live interface to the ship’s IMAC automation system could not be implemented due to strict onboard safety and data-access restrictions. To maintain realistic validation data, an alternative procedure was applied, combining synchronized video records of bridge displays and manually logged control signals. This workaround still provided sufficient fidelity for model calibration and verification within the SimPleShip framework. In cooperation with the partners, these channels were systematically reviewed, and a reduced subset was selected that contained all relevant parameters for model coupling and validation, such as propulsion power, engine torque, fuel consumption, pressures, temperatures, and control signals. The selected signals were then harmonized and formatted according to a unified tag list agreed upon among the project partners to enable consistent data exchange between simulation environments and measurement datasets.

This harmonized dataset provided the essential foundation for verifying and parameterizing both the hydrodynamic and thermodynamic simulation models within the SimPleShip framework.

The acquired data served two purposes:

1.

Model validation, by comparing simulated and measured responses for equivalent maneuvers;

2.

Scenario definition, providing realistic boundary conditions (wind, current, water depth) for simulator tests.

4.2.2. Data Harmonization

Since the measurement data originated from different recording and automation systems on board and from various project partners, a standardized data format had to be created to ensure compatibility across all simulation environments. For this purpose, a dedicated data access and conversion tool was developed, which allowed the merging, synchronization, and preprocessing of the recorded datasets. The tool combined bridge and engine-room data, synchronized their timestamps, and removed invalid or redundant signals. Special attention was given to aligning the sampling rates and to maintaining the integrity of time-dependent parameters such as engine load, propeller thrust, and control commands. The resulting harmonized dataset enabled consistent analyses between nautical and technical systems, particularly for the evaluation of energy demand, dynamic load changes, and emission-relevant parameters. For this purpose, a dedicated data-harmonization tool was implemented at Hochschule Wismar. The software automatically merged and synchronized datasets from bridge and engine systems, aligned sampling rates, and exported a unified tag structure according to the project-wide data exchange format.

4.3. Ship Dynamics Modeling

The hydrodynamic core model, developed with software SIMOPT v2.2 of ISSIMS GmbH, reproduces the ship’s motion with up to six degrees of freedom (6-DOF). For the analysis of maneuvering dynamics and energy behavior, the full 6-DOF ship model, representing surge, sway, heave, roll, pitch, and yaw motions, was applied. At the fundamental level, the dynamics follow Newton’s second law for the translational degrees of freedom and the Euler rigid-body equation for rotation.

$$m·\mathit{a}=\sum \mathit{F}$$

(1)

$$I \dot{\omega }+\omega \times \left(I\omega \right)=\sum \mathit{M}$$

(2)

where m is the (lumped) mass of the ship including the entrained water, a the acceleration, ∑F the total external and internal forces, I the inertia tensor about the center of mass, ω the angular velocity, and ∑M the total moments acting on the hull. To obtain forces (and, analogously, moments) from measured or simulated velocities, we use the integral relation.

$${\int }_{t_0}^{t}F\left(\tau \right)d\tau =m\left(v\left(t\right)-v\left(t_0\right)\right)$$

(3)

which in discrete time yields the practical estimate

$$\mathit{F}\left(t_i\right)\approx m{\displaystyle \frac{v\left(t_i\right)-v\left(t_{i-1}\right)}{\Delta t}}$$

(4)

These identities are employed within the SIMOPT core to ensure that actuator transients and environmental forcing translate consistently into physically based force and moment balances during fast-time maneuver prediction. The comprehensive model ensures realistic dynamic responses under complex environmental and operational conditions while maintaining sufficient computational performance for fast-time simulation and energy-prediction studies. Environmental influences, such as wind and current forces, are represented by empirical coefficients derived from reference vessel data and validated by simulator calibration compared to full-mission simulations. The model incorporates dynamic behavior of control actuators, including rudders, azimuth thrusters, and bow thrusters, each described by first-order delay elements that reflect the physical response times of these systems. Engine commands from the technical simulation layer are converted into propeller thrust using quadratic characteristic curves derived from measured propulsion data of the reference vessel.

The hydrodynamic and propulsion characteristics used in this study were identified for a reference vessel to ensure a consistent and reproducible representation of the investigated use case. Transfer to other ship types requires a parameter-transfer procedure that combines geometric scaling (principal dimensions, propulsion arrangement) with re-calibration of selected empirical coefficients based on a limited set of maneuvering and measurement data. Operational degradation effects, such as increased resistance due to hull fouling and performance changes due to engine aging, can be represented through corresponding adaptations of model coefficients (e.g., resistance/force and efficiency-related parameters) within the same framework. For hybrid configurations, the State-of-Charge (SoC) visualization is intended as an operator-facing indicator of the available energy reserve, supporting operational decisions rather than changing the underlying hydrodynamic model.

4.4. Engine and Thermodynamic Modeling

The technical layer is based on thermodynamic process models of medium-speed marine diesel engines, which were developed and provided by the project partners FVTR GmbH and the University of Rostock. These models come from previous research within the MEmBran project and were adapted by the partners for integration into the SimPleShip framework via a Modelica-based simulation environment. The Modelica implementation enables the coupling of engine and power-system dynamics with the ship-handling simulation, allowing the analysis of interactions between propulsion control, energy consumption, and overall system efficiency. This integration is essential for the entire SimPleShip framework, as thermal and electrical energy flows are interdependent: the heat output of the engines affects auxiliary systems such as cooling and heating circuits, while zero-emission operating modes can be evaluated by including battery-supported hybrid configurations within the same simulation environment. This comprehensive coupling ensures that both energetic and operational effects are consistently represented across the hydrodynamic and thermodynamic domains, enabling realistic assessment of different maneuvering and energy-management strategies.

4.5. Coupled Co-Simulation Procedure

Within the SimPleShip project, the simulation environments were operated independently and linked only through defined data interfaces. The SAMMON software was used to design and analyze maneuvering plans in offline mode. These plans were generated as time-series datasets describing control commands and ship-response parameters for specific maneuvers. The SAMMON environment was connected via an FMU interface to the Modelica-based technical models developed by the project partners. This allowed for the energetic and thermodynamic consequences of the planned maneuvers—such as fuel consumption, load dynamics, and emission behavior—to be computed and evaluated. The maneuver plans considered here are to be understood as pre-planning in the sense of good seamanship. They provide the basis for a standardized, reproducible evaluation. In the FMI coupling mode, SAMMON (nautical domain) and Modelica (technical domain) can, in principle, be operated in parallel as co-simulation components; however, this paper focuses on offline-based plan generation with subsequent energy analysis. The Full Mission Simulator serves as a practical substitute for onboard trials, as the plans are executed and evaluated identically under realistic bridge conditions. An analysis of the effects of communication step size and delays on online assistance will be added as a next step in the future. The data exchange between the nautical and technical simulation domains was realized through a dedicated FMI-based coupling setup developed jointly by Hochschule Wismar, FVTR GmbH, and the University of Rostock. Figure 2 illustrates this integrated workflow, including the planning and transition tools, as well as the FMU interfaces connecting the partner environments. The schematic visualization helps to understand how maneuver plans created in SAMMON were transferred into the Modelica-based energy models and back into the analysis environment for validation and optimization.

The technical implementation of the coupling was realized in the Dymola simulation environment provided by FVTR GmbH. Within this setup, Functional Mock-Up Units (FMUs) were embedded as modular interface blocks allowing bidirectional data transfer between the nautical and thermodynamic domains. Figure 3 shows the realized FMU configuration inside Dymola, which served as the central component for the co-simulation between the ship-handling and engine-system models.

In a separate step, the same maneuvering plans were transferred to the Full-Mission Ship-Handling Simulator (SHS), where they were executed and reviewed by experienced nautical officers. This process enabled a qualitative and quantitative validation of the simulation results under realistic bridge conditions, comparing predicted and observed ship behavior. Through this modular workflow, SimPleShip achieved a consistent evaluation chain linking fast-time maneuvering prediction, technical energy modeling, and human-in-the-loop validation within the framework.

4.6. Experimental Setup and Validation Strategy

4.6.1. Simulator Trials

The validation of the developed methods was conducted at the Maritime Simulation Centre Warnemünde (MSCW) using the Full-Mission Ship-Handling Simulator (SHS). Within this environment, the maneuvering plans created in SAMMON were executed by experienced nautical officers under realistic conditions. The focus of these validation sessions was to demonstrate the practical applicability of the predicted maneuvers and to compare the simulated ship response with the behavior observed during the simulator trials. Representative maneuvers such as harbor approaches, turning operations, and berthing sequences were used to verify the plausibility of the coupled simulation results. Rather than conducting extensive test series, the emphasis was placed on qualitative validation and demonstration, showing that the energy-related and operational effects computed in the Modelica environment corresponded consistently with the maneuvering behavior observed in the simulator. In addition, environmental conditions such as wind, current, and restricted water depth were parameterized based on Norwegian meteorological and hydrographic data (BarentsWatch.no “https://www.barentswatch.no/bolgevarsel/?lang=en (accessed on 26 July 2024”). These boundary conditions were implemented in the simulation scenario to verify the model response under realistic operational stress and to ensure that energy-related effects were not limited to idealized scenarios.

4.6.2. Quantitative Evaluation

Energy consumption was evaluated as the time-integrated propeller power demand:

$$E_P={\int }_0^T{P}_{prop}\left(t\right)dt$$

(5)

The evaluation of energy demand and emissions was based on the Modelica engine simulations combined with the harmonized measurement data from the reference vessel. Instead of a detailed statistical analysis, the focus was on practical consistency checks between simulated and recorded signals. The propeller power demand was integrated over time to estimate the total energy required for each maneuver and the results were compared with representative operational data. Predicted fuel consumption and emission trends (CO₂) were assessed for plausibility using characteristic load and engine-speed profiles. The maneuvering plans generated in SAMMON were analyzed offline with the coupled Modelica models to evaluate their energetic impact and then replayed in the full-mission simulator by experienced navigators. This approach provided a realistic understanding of how energy-efficient maneuvering strategies perform in practice, without requiring a fully quantitative test campaign.

4.7. User-Interface and Decision-Support Components

An energy-feedback extension was developed for the SAMMON planning environment to visualize the energetic effects of maneuvering decisions. Within this interface, the predicted ship trajectory is displayed together with estimated power demand, fuel consumption, and emission indicators derived from the coupled Modelica models. This prototype allows nautical officers to explore alternative control sequences and immediately see how different maneuver plans affect the ship’s energy profile. The approach follows a human-centered design philosophy, aiming to strengthen the crew’s understanding of energy-efficient ship handling and to improve awareness and acceptance of assistance functions.

5. Results

5.1. Energy-Oriented Maneuver Planning in Confined Waters

The main goal of the SimPleShip implementation was to explore how simulation-based assistance can contribute to more energy-efficient ship maneuvering under realistic operational conditions. For demonstration, a confined-water scenario in a coastal area similar to the Geiranger Fjord environment was used. The virtual cruise vessel (approx. 350 m length, twin azimuth propulsion, bow thrusters) performed approach and berthing maneuvers under variable environmental influences, such as wind and current.

Figure 4 presents a sample snapshot of the three-dimensional simulation environment representing the Geiranger Fjord scenario used for the demonstration runs. The 3D visualization reproduces realistic topography, coastline contours, and depth variations, providing the spatial context for the maneuver planning and validation activities described in this section. Within the SAMMON software, offline maneuvering plans were prepared and evaluated using the coupled Modelica engine models to estimate the energetic impact of different control strategies. Figure 5 presents one of the maneuvering plans developed within the SAMMON environment, focusing on the use of bow thrusters as the primary control elements while the POD drives provide only supportive thrust. This strategy enables resource-efficient operations, as the combined power demand of all three bow thrusters remains below 9 MW even at full utilization. The approach is particularly suitable for battery-assisted or hybrid operation modes, where precise but low-power control is required during departure or low-speed maneuvers.

The same plans were later executed by experienced nautical officers in the full-mission simulator to compare predicted and observed vessel behavior. The results showed that the assistance-supported maneuvers were smoother and exhibited reduced peak loads on propulsion systems compared with conventional manual operation. Although no large-scale statistical campaign was carried out, the simulations consistently indicated lower overall energy demand and more stable control inputs. Figure 6 compares the total energy demand of three representative maneuvering strategies. The diagram illustrates how assistance-supported maneuvers resulted in smoother power profiles and reduced total energy consumption compared with conventional control sequences. The results confirm the trend observed during simulator trials, where prediction-based planning produced more energy-efficient motion with fewer load peaks on propulsion units. Participants also reported better situational awareness, and a clearer understanding of how their control actions influenced energy consumption.

5.2. Engine and Emission Model Validation

The thermodynamic and data-based emission models developed by FVTR GmbH and the University of Rostock were verified using on-board operational data obtained from the IMAC monitoring system of the reference vessel. The datasets provided realistic load and response characteristics under transient conditions, enabling the adjustment and validation of model parameters. This ensured that the simulated engine torque, fuel consumption, and emission behavior reflected actual ship operation and could be reliably used within the SimPleShip framework. These results confirmed the suitability of the Modelica-based models for integration into fast-time or co-simulation environments.

5.3. Practical Findings from Simulator Studies

5.3.1. Human Performance and Behavioral Effects

During simulator demonstrations, officers using the energy-feedback functions in SAMMON tended to plan fewer but more deliberate maneuvering points, resulting in smoother steering actions and less frequent thruster use. This behavior mirrors earlier observations from training studies, where prediction-based assistance promoted more conscious and energy-aware decision-making. Feedback gathered after the sessions described the interface as intuitive and helpful for understanding the energetic consequences of different control strategies, underlining the relevance of human-centered design. The subjective feedback collected after each session indicated that crews considered the prediction interface “intuitively understandable” and “useful for assessing maneuvering consequences.”

5.3.2. Integration of Hybrid Power Systems

In additional simulation studies, the coupled framework was extended to include a simplified hybrid-power configuration.

By introducing a battery module into the Modelica simulation by the LTT, short-term energy buffering during high-load situations could be represented. The hybrid setup demonstrated how energy peaks on diesel generators can be reduced and how such models could be used to assess future low- or zero-emission propulsion concepts such as battery or fuel-cell systems. This illustrates the flexibility and scalability of the SimPleShip approach—any power-generation model that follows the FMI standard can replace or complement the current engine model. The inclusion of hybrid elements demonstrates the scalability of the SimPleShip framework toward future zero-emission vessels using batteries or fuel cells. The flexible co-simulation setup allows substitution of the engine FMU with any other power-source model following the FMI standard.

5.3.3. Workshop Demonstrations and Industry Feedback

At the final SimPleShip workshop at MSCW in 2025, the coupled simulation environment was presented to representatives from maritime research and industry. Parallel sessions showcased the energy-optimized maneuvering system and the hybrid power management simulation. Feedback highlighted the educational and research value of the system. This workshop outcome may serve as a spotlight pilot-study with first indicative statements providing rough tendencies of user opinions and views. Those statements seemingly tend to support, that even without automation, the visualization of energy feedback increased the crew’s awareness and acceptance of energy-efficient operation principles. Participants agreed that human expertise remains central to operational optimization and must be considered in future system design—an insight consistent with earlier studies on crew involvement and acceptance.

6. Discussion

6.1. Comparison with Related Research and Publications

The results of SimPleShip align with and extend a well established line of research on simulation-based optimization and maneuvering assistance developed at Hochschule Wismar and its partners. In contrast to many existing studies that focus primarily on nautical trajectory/maneuver optimization or, separately, on machinery and energy analysis, our approach couples both domains within a modular FMI-based setup. The technical novelty is based on the exchangeable multi-domain architecture (via FMUs) and on the validation design, which links plan-based evaluation to human-in-the-loop replay in a full-mission simulator. This enables consistent comparisons of different propulsion and power-system variants (e.g., hybridization) under identical nautical boundary conditions without changing the integration logic. The current boundaries of the contribution—reference-vessel calibration and the absence of long-term field validation—are stated explicitly already.

Earlier work carried out in other RTD projects [29] provided the scientific foundation by numerically modeling emissions and fuel consumption for transient maneuvering. The present project advances that approach by embedding the thermodynamic engine models within a coupled simulation framework that, innovatively, for the very first time, links technical and nautical domains. This setup enables the energetic evaluation of maneuvers directly within the ship-handling simulation environment.

Earlier studies, such as those presented in [52], have highlighted the educational potential of simulation-augmented maneuvering tools. Their focus was mainly on navigational optimization, improving steering sequences, and trajectory prediction to reduce fuel consumption. SimPleShip builds on these findings and further improves the approach by integrating hydrodynamic motion models with engine and power-system simulations, thereby demonstrating that combined analysis of complex motion and machinery provides a more comprehensive view on energy efficiency.

Finger and Schaub et al. [22,52] presented the embedding of thermodynamic engine-process models directly within the SAMMON simulation environment. This approach required extensive adaptations to the software structure in order to represent different engine types and operating modes consistently within the fast-time simulation framework.

In contrast, the SimPleShip-SigMa implementation followed a modular concept, where the thermodynamic models developed by the project partners were linked via FMU interfaces rather than embedded. This made it possible to exchange or extend engine and power-system models more flexibly and to integrate them with the nautical simulation domain. The experience gained from this approach underlines the importance of standardized model interfaces for building interoperable, multi-domain digital twins.

From an automation viewpoint, Damerius et al. [56] and Schubert et al. [59] provided detailed discussions on gradual transitions from manual assistance to automated control and emphasized the continuing importance of human expertise. The current findings reinforce this perspective: even with accurate prediction and optimization tools, the navigation officer’s understanding, acceptance, and interaction remain critical for achieving sustainable operational improvements.

Finally, Schaub et al. [60] introduced data-driven models for particle-emission prediction. Their integration with physical thermodynamic models in the current work shows how such hybrid methods can be applied in practical simulation. Combining physical measurements with computational efficiency marks an important step toward digital-twin-based optimization of ship operations.

6.2. Implications for Digital Twin Development

The SimPleShip results illustrate that the coupling of nautical and engine models enables the provision of a reliable foundation for developing a ship-operation digital twin. The architecture realized in the project addresses several core aspects of such systems:

• Standardized model interfaces: Data exchange between the hydrodynamic and thermodynamic domains was achieved through FMI-based FMU coupling, ensuring consistent communication between independently developed simulation environments.
• Synchronized data processing: Although the simulations were executed sequentially rather than in real time, the harmonized datasets allowed time-aligned evaluation of maneuvering and energy processes within a shared framework.
• Model calibration and validation: Using measured data from the reference vessel enabled the parameter adjustment and verification of simulation models, improving their representativeness of real-world ship behavior.

This approach supports scenario analysis, energy evaluation and model-based condition studies within a single integrated framework. In future developments, the same modular structure could be extended toward real-time data exchange.

6.3. Human-Centered Automation and Training

The human element remains at the core of maritime operations. Simulator demonstrations within the project showed that assisted decision support can improve the consistency and efficiency of maneuvering actions without removing the officer from the control loop. These observations confirm earlier findings [16,22,55,57,61], where early familiarization with simulator environments and virtual-reality tools was found to enhance training effectiveness and operational confidence. The present study further highlights that acceptance of assistance systems increases significantly when crews are actively involved in the planning process and understand the rationale behind system recommendations. Therefore, future automation concepts should combine algorithm-based optimization with clear, user-friendly interfaces that make system recommendations easy to understand and follow. The simulator provides a practical environment to test and improve these concepts under realistic conditions and helps to ensure that new automation solutions remain user-orientated and suitable for real operations.

6.4. Contribution to Decarbonisation Strategies

While technical innovations such as alternative fuels and electrification dominate public discussion, operational optimization offers an immediately applicable route to reduce emissions from existing fleets. The findings confirm that more efficient maneuvering and informed operator behavior can noticeably reduce energy demand without technical modifications to the vessel. Comparable simulator studies have shown potential savings in the range of 15–25%, depending on ship type and operating conditions. The authors underline that the reported range is based on defined simulator experiments and scenario studies and should not be interpreted as a universal, fleet-wide savings figure. It is bound to the specific boundary conditions of ship type and propulsion configuration, maneuver characteristics, environmental influences, and the setups of the scenario and the simulator. Establishing a robust, statistically valid statement for real-world operation requires long-term onboard measurements and repeated operational profiles across different operating states but goes far beyond the project’s period. This field validation is planned as one of the next steps of our research. Scaled to the passenger-ship sector, such improvements could lead to substantial reductions in CO₂ emissions and fuel costs. In addition, the project’s modular simulation framework allows the evaluation of future propulsion concepts—such as hybrid systems or alternative fuels like methanol and ammonia—by simply replacing the engine FMUs with suitable power-plant models. Beyond propulsion and maneuver dynamics, ‘hotel’ loads, particularly HVAC, offer significant efficiency levers. Dynamic, zone-resolved HVAC Digital Twins have shown that setpoint and ventilation strategies can markedly reduce electrical demand on cruise vessels, complementing our maneuver-time optimization and broadening the pathway to decarbonization [62].

This flexibility supports scenario studies aligned with the IMO decarbonization roadmap and complements ongoing research.

6.5. Methodological Limitations and Future Challenges

Although the results are encouraging, several methodological limitations should be acknowledged:

• Sensor Data Availability: Continuous on-board measurement of fuel consumption and emissions is still limited, meaning that model calibration currently depends mainly on engine testbed and logged IMAC data.
• Model Generalization: The simulation models were tuned for specific reference vessels. Developing generic ship-type models will require larger and more diverse datasets from different ship classes.
• Human Factors: The behavioral observations were based on sessions with more than thirty participating nautical officers and engineers. While this represents a solid empirical basis, an extended evaluation across multiple training institutions would be valuable to further quantifying acceptance and learning effects.

Future research will focus on improving model adaptability through automated parameter estimation, expanding the data basis for different ship types and conducting joint simulator studies with international partners to strengthen the statistical and behavioral findings and addressing the complexity of ship operation and maritime transportation [61]. Moreover, activities to improve training of seafarers in energy-efficient ship operation, as i.a. required by [63] and studied and exemplarily conducted in [64,65] will be another important focus of future research.

7. Conclusions and Outlook

7.1. Summary of Achievements

The SimPleShip project has demonstrated that a simulation-based integration of nautical and technical fields can significantly support energy-efficient operation of complex passenger ships. The research achieved three major outcomes:

1.

Development of a modular co-simulation framework linking ship-handling/maneuver planning and engine-system models via FMI/FMU interfaces. This enables consistent analysis of maneuvering dynamics and energy processes across both domains.

2.

Validation of hybrid physical and data-driven models for fuel consumption and emission estimation using testbed and IMAC measurement data, achieving sufficient accuracy for scenario-based energetic evaluation.

3.

Demonstration of human-centered optimization methods, showing that prediction- and planning-supported maneuvering can noticeably reduce energy demand and improve situational awareness during complex operations.

The results confirm that operational optimization, when supported by intelligent simulation and visual feedback, offers an immediately applicable and cost-effective pathway toward maritime decarbonization.

7.2. Outlook for Research and Application

The project results open several directions for further development and application:

• On-board decision support: The modular architecture could be deployed on real vessels to provide energy-advice and predictive feedback during operations.
• Integration with hybrid and alternative-fuel systems: The flexible FMU-based framework allows replacement of engine or energy-storage modules (e.g., methanol, battery, or fuel cell) to study future propulsion scenarios.
• Assisted automation: Coupling predictive simulation with control algorithms may enable semi-autonomous maneuvering functions that optimize both safety and energy use while keeping the officer in the loop.
• Education and training: Incorporating the prediction and energy-feedback tools into simulator courses promotes energy awareness and supports the training of environmentally responsible bridge officers.

Overall, SimPleShip provides a replicable methodology and validated model foundation for sustainable ship operations and an important step toward the digital and green transformation of maritime transport.