Dr Kean Lee Kang

Interests

My research interests lie in designing and applying artificial intelligence and machine learning methods to new problem domains. This approach was first successfully performed on wavepackets in fluid dynamics, producing results that agree well with classical theories even though the algorithms are blind to the underlying physics, while simultaneously suggesting improvements to existing models of wave interactions in the boundary layer flow. In parallel with these cutting-edge data analysis tools, I also developed the computational models that generate the raw data for subsequent analytics applications. During my PhD, this involved direct numerical simulations (DNS) of the notoriously challenging partial differential equations known as the Navier-Stokes equations. At Airbus, these computations took on a more practical slant due to the needs of industry, and involved accelerated solution methods such as the Reynolds-Averaged Navier Stokes (RANS) and Parabolized Stability Equations (PSE). Computational fluid dynamics models of advanced laminar-turbulent transition control systems were developed, involving plasma, heat and roughness arrays.

Key projects included collaboration with a team of eleven industrial, academic and governmental organizations in the European Union Seventh Framework Project BUTERFLI (BUffet and Transition delay control investigated with European-Russian cooperation for improved FLIght performance), cooperation with a consortium of nine companies and universities in the Innovate UK national project ALFET (Advanced Laminar Flow Enabling Technologies), and participation as an Airbus industry representative at DiPART (Distributed Partnership in Research & Technology) to network with the UK flight physics community.

Now, my computational modelling and data analytics research is directed towards improving the quality of service provision for cancer patients, working in close collaboration with the Macmillan Cancer Support charity, and the Digital Health and Care Institute, which is part of the Scottish Funding Council's Innovation Centre Programme. This is facilitated by access to detailed datasets tracking the journey of cancer patients from a social care perspective. Computational algorithms are developed for service optimisation and predictive modelling using this evidence base, and interactive visualization software is co-developed with computer games technology to provide the end user with a rich and immersive interaction experience with the model and data. Furthermore, deep neural networks are trained to use actual client demographic and holistic needs assessment (HNA) data to make intelligent referrals to relevant agencies such as hospices or government departments.

View Publications