Aurelien Ghiglino

Thanks to Unmanned Air Vehicles, Urban Air Mobility, and the drive for net zero, there is the motivation and opportunity for real innovation in aerospace – particularly to evolve beyond traditional aircraft design. As a PhD student at Stanford University, Aurelien writes generative machine learning-based models to design and optimize unconventional aircraft configurations. The aim is that these tools will lower the difficulty of working with unconventional designs and allow aerospace engineers to perform rapid iteration of designs at the conceptual design stage. Aurelien uses the SUAVE tool, based on traditional aircraft design techniques, to analyse a design’s many systems (aerodynamic, stability, propulsion) in a coupled, multidisciplinary analysis, with a particular focus on balancing the cost and environmental impact of future aircraft designs. The use of various optimization techniques, such as genetic algorithms, allows families of near-optimal aircraft designs to be obtained. From this, Aurelien builds databases of fictional but reasonably optimal aircraft that can be used to train generative machine learning models. Other parts of his research involve finding new methods of integrating physical system behaviours into machine learning techniques.

During his undergraduate studies, Aurelien extensively researched the issue of noise pollution for aerofoils and propellers, particularly relating to broadband trailing edge noise.