What is your role?
I lead an academic research group in the department of Aeronautics at Imperial College London; working in the field of Computational Fluid Dynamics (CFD). I also I teach Mathematics to 1st Year undergraduates.
What do you enjoy most about your role?
Firstly, the independence and freedom to play and explore. Being able to take a new idea, test it and see what happens. Secondly, my job enables me to work with outstanding individuals, and in outstanding teams, both here at Imperial, and internationally. I love working with excellent people, and making things happen!
What is your background?
Both of my parents are artists so when I chose science over art I was probably considered to be the black sheep of the family; although I would argue that creativity is key to driving innovation in mathematics/physics/engineering etc. I did my undergraduate degree in Physics and PhD in Aeronautics, both at Imperial College. My postdoctoral scholarship in Aeronautics and Astronautics was at Stanford University. I came back to Imperial College as Lecturer and now I am a Senior Lecturer.
What are you working on at the moment?
Our main focus is on development of next-generation CFD software that can harness the power of future hardware platforms to solve currently intractable fluid flow problems. In particular we are developing an open-source flow solver called PyFR (www.pyfr.org). We are also working to translate this technology to industry via projects such as Hyperflux with CFMS and Zenotech. It is really exciting to see our latest academic advances being drawn through to industry, where they will hopefully have real and significant impact.
I also have a strong interest in simulation of biological flow; in particular we are working to design novel Arterio Venous Fistulae (AVF) configurations for dialysis with reduced failure rates. The ultimate aim of our current work is to develop a bespoke patient specific AVF design capability, which leads to significantly improved clinical outcomes for patients with kidney failure who rely on AVF for dialysis.
What are you looking forward to in your field/area of expertise/industry? Any predictions/advice?
I think one of the biggest challenges over the next 5-10 years in the computational engineering sector will be developing algorithms and software that can effectively leverage the capabilities of modern hardware. Whilst available FLOPS (Floating-Point Operations per Second) are increasing, memory bandwidth is not keeping pace, and the FLOPS are being offered up in a very parallel fashion. Existing algorithms, many of which were developed in a serial FLOP bound era, will find it increasingly difficult to exploit these extra FLOPS.
Further, I also think algorithm/software/hardware co-design will become increasingly important if you want to develop the most performant solution.
My advice: learn to code! A computer is a tool – make it work for you. Being only a user and relying on prewritten software will eventually create boundaries for your work.
If you would like to feature in 'Shining the Spotlight', click here.