IS05 - Artificial Learning for Marine Engineering: Methods and Applications

M. Diez (CNR-INM) and L. Mainini (Politecnico di Torino)
The complexity of marine engineering problems (from the analysis of complex physical phenomena to optimal design and control of marine structures and vessels; from marine exploration to autonomy and robotics applications) calls for suitable scientific computing frameworks able to provide cost-effective and reliable solutions. Cutting-edge methodologies of machine learning (ML) and artificial intelligence (AI) have shown their potential in providing effective solutions to these problems. However, commonly ML/AI techniques require significant amounts of data to learn from; in addition, the responses are often affected by lack of interpretability and their reliability is of difficult characterization. These features often constitute a limitation to the acceptance of these techniques for scientific computing in engineering applications since many engineering problems are associated with high-regret and safety-critical decisions for which the collection of reference data points is usually expensive. The scientific community is dedicating efforts to address this limitation, reduce the quantity of data required by the models, and improve interpretability and reliability of the predictions, therefore paving the way for a broader adoption and acceptance of ML/AI in engineering applications [1,2]. The objective of the invited session is to offer a place for discussion on capabilities, challenges, and open issues for the application of ML/AI to marine engineering. We invite contributions on different approaches and applications of ML/AI in marine engineering with the aim of achieving cross-fertilisation of approaches and new ideas.
Computational Mechanics, Machine Learning, Optimization, Artificial Intelligence, Scientific Computing, Control, Decision Making