Adrian R. G. Harwood
MEng (Hons), PhD
Who Am I?
I am a Senior Research Software Engineer and former University Lecturer working in Research IT at The University of Manchester, UK. I am responsible for leading the development of Mobile and GPU applications for research projects across the university.
I completed my undergraduate master’s degree in Aerospace Engineering (MACE) at The University of Manchester. During this time I worked for Rolls-Royce throughout the UK on the design of safety-critical systems including thrust reverser control for corporate aviation gas turbine engines. Having been award the prestigious Alumni Research Impact Scholarship, I completed my PhD in aero-acoustics in 2014, also at The University of Manchester, developing novel numerical methods and models associated with the scattering of acoustic waves in general environments. This research yielded a collection of practical methods for enhancing hybrid noise prediction.
As a Research Associate (and later Fellow) in the school, I led the development of the LUMA (Lattice-Boltzmann @ the University of Manchester) software project, a multi-grid, many-core Fluid-Structure Interaction solver written in C++. It is based on coupling the Lattice-Boltzmann Method and the Finite Element Method. LUMA is currently being applied in the fields of aerodynamics, bio-engineering and flow control to develop new engineering solutions through its simulations.
I became a permanent academic at the University in 2018 focusing on developing real-time physics simulation coupled with interaction. My research develops energy-efficient simulation tools for virtual engineering and incorporates exciting technologies such as mobile devices, depth-sensing cameras, image processing and virtual/augmented reality. I changed my academic position for a leadership role in professional services in 2019 where I now lead research software engineering projects in mobile and GPGPU applications, connecting me with research groups across the university.
My research themes explore a number of opportunities by developing scientific modelling and simulation software for emerging hardware platforms with a view of increasing software scalability as well as user engagement and interaction with simulation.
I have previously led the development of two in-house simulation applications (LUMA and its GPU-accelerated brother LUMIS). Both are fully object oriented, modular and benefit from the use of version control to allow distributed development and testing. LUMA is used by our research group and our collaborators on the UK's national super-computing facility ARCHER.
LUMIS is complete framework for virtual engineering using fluid flow simulation based on the original LUMA initiative. It targets real-time, interactive computational fluid dynamics (CFD) technologies. It is written in CUDA C/C++ and is being developed for multiple GPUs to perform near-real-time simulation in 3D.
I also have an Android application powered by the LUMIS kernel to allow users to carry a real-time CFD simulation in their pocket. I am continuously looking to identify roles within industry and teaching and learning for mobile simulation of this kind. I like to broaden my experience and knowledge through interdisciplinary projects and look to explore new and exciting applications of technology.
Areas of particular interest are:
HPC (GPU/Many-core CPU)
Mobile Computing (Android-based)
Engineering Visualisation (AR/VR)
Real-time Computational Acoustics