Technology Futures | 1 February, 2017
SYSTEMS THAT TELL THE STORY OF EVERYDAY HUMANITY
Professor Dimitrios Nikolopoulos is Head of the School of Electronics, Electrical Engineering and Computer Science (EEECS) at Queen’s.
He is an internationally-acclaimed expert in system software and hardwaresoftware co-design of large-scale systems – and he is a man who is not afraid of a challenge.
He explains, ‘Back in my graduate days, you had data used by scientists and engineers only but now it’s data used by everyone – for your financial transactions, your traffic navigation system – pretty much anything. We generate vast amounts of data and to process this we need advanced computing technology that we don’t have on our desktops, our workstations or even in our research labs.
‘The pace at which we’re generating data is much faster than the pace at which we’re improving our systems and their capacity. And because we don’t have the capacity, we throw 80-90 per cent of the data away. Imagine – humanity is generating all this data on a daily basis by travelling, by working, by interacting with others and with the environment and we throw most of it away. So that’s where I come in – to build new software and hardware systems that allow you to host more data and process more data per second.’
LEAD INVERSTIGATOR IN GROUNDBREAKING RESEARCH
Over the past number of years, Dimitrios has been a lead investigator in competitively funded projects with an overall budget in excess of £40m. Holder of the Royal Society Wolfson Research Merit Award, he is Acting Director of the Centre for Data Science and Scalable Analytics Systems, one of the new Centres at the Queen’s Institute of Electronics, Communications and Information Technology (ECIT), where his expertise is leading ground-breaking research.
He says, ‘Before, there were bits and pieces of knowledge engineering, data engineering, high performance computing – but there was no integration of these research themes into one coherent team to address the global challenge of the information explosion. Now there is. ‘We have a good spectrum of expertise, starting from extracting knowledge from the data, understanding the data, right through to building the hardware and the software and the networks to move the data. All of this gives us the opportunity to solve problems of scale, magnitude and complexity that we have not been able to solve just through our individual research efforts.’
He gives an example of such a problem. ‘Say there’s a virus somewhere. It starts spreading and you want to understand how this is happening. In order fully to do so, you need to somehow model the travel patterns and the movement and social interaction of every single person in the world.
‘Take a global event like the Olympics where you had millions of people travelling. They came from so many different countries to Brazil and then they went back and resumed their daily lives. This is extremely difficult to model computationally because people don’t behave the same way everywhere in the world.
‘You have to understand roughly how different people of different social backgrounds, from different countries and different continents behave on a very large scale and then try to simulate that on a computer. It’s an enormously challenging problem.’
INDUSTRIAL CONNECTIONS AND A GLOBAL IMPACT
Dimitrios sees his research having huge benefits for Queen’s and its global impact. ‘Our scientists here – in any discipline – rely on our ability to analyse data in order to understand trends, to understand societies, communities, and how our natural environment evolves.
‘And we need to understand what our scientists need – the doctors who are researching cancer, the food security experts, the astrophysicists. Out of all this data, what exactly is the information we have to extract, how soon do they need it and what’s the most economical method to provide it for them?’
And there are important connections with the industrial world. ‘In this School, every one of my colleagues has not only strong links with industry but measurable impact, whether through technology transfer under licence or spin-outs.
‘They always think of what industry needs and by extension what the customer needs and what society needs. They don’t solve research problems just out of curiosity. It’s because someone out there is waiting for a solution.’