From the magazine – Ship design is engineering in its purest form; a balance of physics, economics, and practical considerations. For decades, the process has relied on applied science and empirical models, refined through experience. But while the industry has embraced digital tools, certain early stages of design remain stuck in a dependency on the design intuition of the naval architect.
This article originally appeared in SWZ|Maritime’s “Smart” special of February 2026. It was written by Diego De León, Researcher & PD candidate on data-driven early ship design at NHL Stenden Hogeschool, and Herbert Koelman, Lector Maritime Innovative Technologies at NHL Stenden Hogeschool and Founder of SARC.
This process is usually still taught through the design spiral, a theoretical framework that describes this as an iterative cycle, where each loop refines the ship’s parameters, balancing requirements. In theory, it’s a dynamic, evolving process that helps make sure all of the design parameters are correlated and in check.
In practice? It’s often clunky, rigid and tedious; especially as the recalculations take time and expert work, a scarce resource in most cases, meaning the framework tends to be abandoned for a more direct, yet rough approach to design. This series of issues is amplified by the administrative and sales overhead that includes variations in requirements due to client demands and changes in mission expectations, among others.
Also read: SWZ|Maritime’s February 2026 issue: Designing the digital maritime future
A gamechanger: A digital design platform
The Smart European Shipbuilding (SEUS) project represents a structured effort to digitally integrate the fragmented stages of ship design. Funded under the EU’s Horizon program, EU Horizon SEUS focuses on integrating computational tools into shipbuilding processes to enhance efficiency, reduce costs, and improve competitiveness. The project establishes an advanced digital framework that interlinks computer-aided engineering (CAE), computer-aided design (CAD), computer-aided manufacturing (CAM), and product lifecycle management (PLM) software into a unified platform.
EU Horizon SEUS brings together academia (Norwegian University of Science and Technology, University of Turku, NHL Stenden), industrial software developers (Cadmatic, Contact Software, SARC), and shipyards (Ulstein, Gondan Shipbuilders) to bridge the gap between theory and development and practical, in site application. With a strong focus on guiding and basing the development of the platform on the direct needs of the industry and the reality of the shipbuilding and ship design workflow.
Tools matter
The ship design process relies on conventional software tools that serve two primary functions: drafting and evaluation. CAD systems enable the creation and manipulation of digital ship models. These tools allow naval architects to define physical geometries and spaces. Their deterministic approach ensures that the digital representation remains accurate, modifiable, and ready for further analysis.
In technical industries, CAE platforms complement CAD by providing evaluation through simulations and calculations. These tools assess critical performance metrics with some examples from the maritime industry including hydrostatics and stability, hydrodynamic resistance and propulsion performance, structural integrity, and specialised software modules support niche, but essential tasks like piping and classification society regulatory compliance checks.
These conventional tools form the backbone of modern ship design, providing the precision, reliability, and repeatability that underpin the entire process helping to translate creative vision into engineered reality.
Also read: TODDIS paves way for data-driven shipping and ship design
The SEUS toolset for early ship design
It starts with SARC, a maritime software and research company specialising in ship design and ship operation tools, blending naval architecture expertise with digital solutions. Its flagship software suite, PIAS, provides a design platform with multiple specific ship design tools, built to bridge the gap between a client’s demand and detailed engineering.
Large area hull shape modifications have been a challenge in ship design from the beginning. Most hull design tools rely on patchwork surfaces: Rigid, interconnected panels that do not reflect the way a hull is thought about in naval architecture. Fairway, the hull shape module from PIAS, takes a different approach, using a network of curves instead.
This method removes traditional constraints like topological rules as curves can intersect freely, creating cells with any number of corners; regularity demands, curves can stretch across the hull in any direction, adapting to the design’s needs; and direct surface control, every point and curve sits on the hull, eliminating indirect adjustments. Furthermore, closed surfaces are automatically generated between these curves.
The real gamechanger? Spatial deformation, a technique proven in the entertainment industry that warps the design space itself, letting designers reshape the hull intuitively. Unlike older methods that break continuity, spatial deformation maintains smoothness, making it ideal for early-stage exploration where flexibility matters most, allowing for specific modifications without compromising the integrity of the model.
Increasing the potential for this type of tool, the continuity of the surface makes it ideal for use in computational fluid dynamics (CFD) simulations, and the parametrisation of the hull shape opens the door to implement easy hull shape modifications based on the feedback from the fluid simulations. Local modifications without disturbing the surface of the model are made easy with spatial deformation.
While most propulsion tools treat components in isolation, the PIAS Propulsion and Machinery module works to integrate them into a unified system of interconnected equations. Building on PIAS’ proven resistance and propeller modules, the tool links mechanical, electrical, hybrid and even wind-assisted propulsion setups through physics-based relationships, all governed by core principles like RPM synchronisation, power conservation, and torque-RPM relationships. Unlike standalone methods, it solves the entire propulsion chain as one, ensuring that changes to engines, gearboxes, or propellers propagate realistically through the system evaluating a complete picture of a static condition of the propulsion train.
The tool’s power lies in its practical flexibility. Whether predicting a vessel’s speed from a hybrid diesel-electric setup or determining the power required to hit a target speed, the tool models interactions while factoring in propeller efficiency, hull resistance and losses in the propulsion train. The complexity of the equations stays behind the scenes; the user simply defines the problem, and the tool delivers actionable insights, turning propulsion design from a latestage challenge into an early-stage opportunity.
The PIAS Layout module is getting an upgrade to accelerate internal design with “chunks”: predefined, scalable sections of a ship (engine rooms, accommodations, cargo holds) that snap together like modular blocks. Instead of drawing every compartment from scratch, designers pick from a library, drop chunks into the hull to create a rough subdivision that fits the vessel. Need adjustments? Matching to requirements ensures spaces meet volume, area, or distance rules, whether fixed, flexible, or range-based. A smarter layout design, turning a time-consuming puzzle into a drag-and-drop workflow. And this flexibility for direct and easy modifications of the layout opens the door to the possibility of a new generation of ship design tools.
A new generation of tools?
The current ship design process relies heavily on deterministic methods, computational tools that translate a designer’s vision into precise, rule-based models and simulations. These approaches are indispensable for validation and certification, but they operate within strict boundaries. While effective for refining known solutions, they struggle to explore vast design spaces or adapt to emerging design challenges.
This is where data-driven approaches offer a complementary path. Unlike conventional tools, which require explicit inputs and fixed rules, data-driven methods learn from patterns, generalise behaviours, and uncover new possibilities. Allowing them to do things as:
- Design exploration: Generative algorithms can produce multiple design proposals based on learned patterns, offering designers a broader view of the design space.
- Adaptive optimisation: Data-driven algorithms can go beyond conventional optimisation methods by adjusting the design to evolving environments via methods like reinforcement learning.
Crucially, these approaches do not replace traditional engineering tools. Instead, they extend them, bringing a component of computational intuition. The result? A process where deterministic precision meets exploratory potential.
Also read: MIWB wins Horizon grant for major European shipbuilding project
D3: SEA
Data Driven Design: Ship Engineering Application (D3: SEA) is the push from EU Horizon SEUS for the integration of these new computational technologies into early ship design. While other developments have been done for optimisation, like ShipHullGAN for hull shapes, the first prototype focuses on the generation of the subdivision for the general arrangement, as it has always relied on experience and iteration. Naval architects sketch compartment layouts, adjust dimensions, and manually balance competing needs between cargo space, machinery placement, fuel storage, and crew areas while ensuring the ship meets operational and regulatory demands. But as the industry shifts towards alternative fuels, the complexity of layout design has increased.
Enter data-driven layout generation. Instead of starting from a blank slate, this approach leverages generative algorithms to propose multiple valid compartment arrangements. The system doesn’t just offer one rigid template; it generates diverse, feasible layouts from the ship’s dimensions and operational requirements. By leveraging conventional software tools as filters, it intends to present configurations that are not only fast to generate, but also practically sound.
Faster exploration, smarter decisions
Of course, this can’t replace the naval architects, but it can empower them. The tool acts as a creative partner, rapidly generating layout alternatives so designers can focus on evaluating trade-offs rather than drafting compartments from scratch. Need to accommodate a new fuel or reconfigure cargo spaces? The system proposes multiple solutions, allowing designers to explore more options early, when changes are easiest to make. Too many designs to choose from? The same evaluation methods the algorithm uses to evaluate the generated designs can be tweaked to better limit the design space.
The real value lies in speed and adaptability. Instead of committing to a single layout prematurely, designers can test ideas quickly, adjusting adjusting for changing requirements. The tool handles the repetitive drafting, ensuring a good ship design.
Also read: Digital experts unite for Maritime Master Plan
Revisiting the ship design method
Early ship design has long relied on static documents and manual processes, sketches, spreadsheets, and PDFs that capture decisions, but lock them in place. Even as CAD and CAE tools have become standard, their potential remains untapped, as designs are still shaped by document-based workflows and disjointed digital steps. The result? A process where ambiguity slows progress, and exploration is limited by the time it takes to redraw, recalculate, and recheck every iteration. Now, the industry stands at a turning point. Proven digitisation like CAD and CAE and emerging data-driven tools are converging, demanding a shift from document-centric to model-centric workflows.
The proposed system doesn’t just add new tools, it rebuilds the foundation to leverage what already exists. Instead of treating early design as a series of fixed hand-offs, it creates a living digital model that evolves with the project, turning uncertainty into opportunity and static files into dynamic, actionable designs.
The proposed system builds on this transformation. Instead of fighting ambiguity or manual iterations, it turns them into advantages:
- Clear and readily available objectives, targets and constraints anchor the model as a foundation.
- Uncertain or evolving requirements become exploration space for algorithms, not roadblocks.
- Existing, proven tools and data-driven tools work together: Traditional methods refine known solutions, while generative algorithms propose feasible alternatives, all within a single, evolving digital model.
- Designers remain in control, shaping proposals and feeding back constraints, whether from clients, physics, or practical limits.
- The SEUS data integration platform ensures that the digital model keeps coherent via a single source of truth and automatic data pipelines between different software tools.
The result isn’t just a static baseline, but a living model that carries forward, adapting as requirements clarify. By the time detailed engineering begins, the design is already optimised, validated, and ready to evolve further. This isn’t just an upgrade to existing theoretical methodologies, it’s a paradigm shift to a model-centric methodology. In an era where ships must be smarter, greener, and more adaptable, ship design needs to keep pace.
Picture (top): The Smart European Shipbuilding (SEUS) project represents a structured effort to digitally integrate the fragmented stages of ship design.
REFERENCE
- [Khan et al., 2023] Khan, S., Goucher-Lambert, K., Kostas, K., and Kaklis, P. (2023), Shiphull-gan: A generic parametric modeller for ship hull design using deep convolutional generative model, Computer Methods in Applied Mechanics and Engineering, 411:116051
