Research within the School is carried out by academics from a wide range of academic disciplines covering Computing, Electrical and Electronic Engineering, Mathematics, Statistics and Physics. A feature of the research is its cross-disciplinary nature. Topics include notions of Complex Systems, Networks, and analysis of large data sets (Big Data).
A common theme for much of the research within the School is the systematic studying of large amounts of data using a range of techniques including data mining, machine learning, techniques from statistical physics and topology. The research ranges from fundamental to applied, with novel applications of benefit to the local and global community.
Cogent Labs Research
Cogent Labs is a world-leading applied research group dedicated to analysis and development of sensing-based sociotechnical systems. It has a dual focus: robust, deployable pervasive sensing systems for real-life applications at scale; and effective packages for empowering users to maximise the benefits of those systems.
Serious Games Research
Our Serious Games research is setting international standards for excellence on the theme of applying effective digital technologies for learning and behaviour change. With expertise in serious games for health, education, and the environment; the Institute has a very successful track record of securing funding as a partner in a wide portfolio of projects funded through the European Union, European Regional Development, Prime Minister’s Initiative Fund, Higher Education Funding Council, Engineering and Physical Sciences Research Council and a range of international industrial partnerships. Projects include application of serious games to science and technology teaching, patient awareness of Chronic Kidney Disease and games to treat attention deficit disorder (ADHD) in children.
Mathematics, Physics and Statistics Research
This area of research covers fundamental and applied problems in mathematics, statistics and physics, in particular the properties of complex systems and their applications. Research interests particularly focus on areas of Statistical Physics and Fluid Dynamics.
Our strengths include the application of techniques from statistical physics and critical phenomena to novel areas of application. Public transport networks are analysed mathematically to enable a local logistics company to decide how best to transfer much of its European logistic network from road to rail. This improves the efficiency of the company’s transport requirements and helps reduce road congestion and carbon footprint. Similar methods are currently being used to analyse big data provided by Centro on UK public transport routes to investigate the resilience of traffic flow to traffic jams.
The School is also involved in research identifying and explaining environmental change in its broadest sense. This typically requires the use of long-term and/or spatially extensive data to identify change against a background of variability that may mask or hinder detection.
Data sources may be digitised already, but could also be based on alternative or novel sources such as paper records, museum collections, and digital or photographic images.
This work is done in collaboration with several national and international institutions, including the Woodland Trust and Universities in Germany and Poland. The research conducted has been extensively used to inform the Intergovernmental Panel on Climate Change and features regularly on national TV, Radio and Press.
The machine learning techniques have been applied in a number of important areas of science and technology.
For example, our research in the area of autonomous manufacturing addresses the challenge of improving the efficiency and competitiveness of UK manufacturing.The project provides a launch pad for pushing manufacturing to the next level of data driven autonomous operation, self-configuration, management and intelligent fault detection resulting in improvements in manufacturing efficiency, quality assurance and overall productivity.
Another application addresses the issue of corporate tax evasion by employing a graph-based method to portray the properties of tax evasion to describe suspicious relationship trails with the same antecedent node behind an interest affiliated transaction.
The school benefits from a High Performance Computer (HPC) consisting of 1400 CPU cores, and 36 NVidia Kepler K20 GPUs for a total computing power of about 60 Tera Flops.
For researchers working on experimental methods in parallel programming, a separate HPC system called Nostromo offers the same facilities on a smaller scale, upon which many different operating systems, libraries and paradigms can be tested and developed.