AI and machine learning

Key Research Areas

Generative AI for Shape Design: Our group has actively explored using generative AI models for fast exploration of new aircraft configurations. These technologies allow us to explore vast design spaces with improved efficiency and creativity, pushing the boundaries of traditional aircraft design.

Surrogate-Assisted Methods: To optimise the exploitation of available data through AI/ML methods, we take advantage of probabilistic surrogate models, such as Kriging, Bayesian Neural Networks, etc.  We are able to blend different levels of accuracy, incorporating expert knowledge to improve generalisation capabilities of developed models. These find applications in efficient global optimization, uncertainty quantification and propagation, etc. and accelerates the design process, facilitating more rapid iterations and refinements.

Design Optimization under Uncertainty: We also explore how our design can be made robust, by incorporating uncertainty quantification in the design process. We explore ways to ensure that aircraft designs can tolerate uncertainties and perform reliably under a variety of operating conditions. We also explore dimensionality reduction techniques for efficiently exploring the design space.

Growing Team and Expanding Focus: Our research group is dynamic and expanding, with a focus on nurturing talent and exploring new frontiers in scientific machine learning. We are always interested in passionate researchers and students who are keen on developing and applying complex probabilistic machine learning techniques not only in aircraft design but also in broader systems design contexts.

We encourage collaborative projects and are keen to form partnerships with other academic groups and industry stakeholders. Through our research, we aim to contribute significant advancements to the field of engineering design, ensuring practical impacts and technological innovation.