AI Accelerated Engineering Blogs

The Future of Maritime Travel: The Next Phase in Ship Design Evolution

Blog: Ship Hull Design

Ship designers face challenges like emissions regulations, efficiency optimization, and future-proof propulsion. AI advancements enable real-time prediction of flow fields and scalar quantities, accelerating product development. A pilot boat project used parametric models and simulations with CAESES and Simcenter STAR-CCM+. Machine learning models based on geometric parameters facilitated rapid design processes. This blog discusses the project, including the creation and accuracy of the Reduced-Order Model (ROM).

Probabilistic Gaussian Processes – an approach to stable response surface modelling for noisy processes

Probabilistic Gaussian Processes – an approach to stable response surface modelling for noisy processes

Response surface models based on Gaussian Kernels, such as Kriging, require the estimation of hyperparameters to fit the model to the training data. Commonly, this estimation is made by exploiting the maximum marginal likelihood (MML) function. However, this is not an easy optimization problem due to the topology of the MML function. An alternative – and more robust – approach to hyperparameter estimation can be based on Bayesian statistics.

Ready to explore how AI can fast-track your problem-solving?

Let's talk.

Get in Touch