ENGYS participates in AI4TwinShip2: Development of interactive CFD-AI models for real-time hull design and optimisation of marine vehicles

Share on:

ENGYS is pleased to announce that it has been awarded funding for the development of the AI4TwinShip2 research project, aimed at improving the design and optimisation of ship hulls using advanced artificial intelligence (AI) tools applied to computational fluid dynamics (CFD). Conceived as a continuation of the parent project AI4TwinShip, AI4TwinShip2 has the main objective of consolidating and integrating the operations of a numerical methodology for the real-time calculation of merchant ship hull resistance and selected CFD results of interest in the naval sector. This methodology will make use of AI models trained on databases of CFD results, with the aim of improving the efficiency, performance and environmental sustainability of ships.

Several technologies will be used within AI4TwinShip2 to achieve the objectives identified for increasing the technology readiness level (TRL) of the numerical methodology. HELYX-Marine, an open-source CFD software developed by ENGYS, will be adopted to generate the CFD results database related to resistance analyses. PhysicsNeMo, an open-source Python toolkit developed by NVIDIA, will be used for training and deploying the AI model through machine-learning algorithms, while rbfCAE, developed by RBF Morph, will be employed as a morphing tool for the geometric parametrisation of hull forms under investigation. An interactive dashboard will also be developed for the real-time prediction of CFD results of interest for ship hulls not used during the training of the AI models.

AI4TwinShip2 will focus on consolidating the procedures investigated in the parent project, implementing components of the HELYX-Marine graphical user interface (GUI) to facilitate the generation of the database of relevant results and the training of the AI model, and developing an interactive dashboard to perform inference on hulls not processed during AI model training.

The AI model will provide accurate predictions of hydrodynamic characteristics, thereby enabling optimisation of the hull shape and reducing fuel consumption and environmental impact. By providing real-time results characterising the hydrodynamic behaviour of the ship, these models will also constitute a key component in the creation of digital twins of ships.

Total costs: 270.088,00 €, Funding: 121.539,60 € (of which 40% European Union, 42% Italy and 18% Friuli Venezia Giulia Region).

Click here to download the project poster.

Clicca qui per scaricare questo articolo in italiano.

Read more

Keynote Speaker Confirmed for the Upcoming ENGYS UGM 2026 in Madrid, Spain

ENGYS Partners with RBF Morph to Deliver Advanced Shape Morphing and FSI Capabilities via rbfCAE

ENGYS Participates in the Aerodynamic Development of the Shell Triple 10 Challenge Concept Car 

Search our Resources
Use the Search bar below to search a specific term in our website.