ECS

Dr Seth Bullock, Head of Science and Engineering of Natural Systems (SENSe) Group, Director of the Institute for Complex Systems Simulation

When Dr Seth Bullock first considered a degree in computer science he was a little put off by the image of a subject associated with geeks and nerds destined for jobs in back rooms and basements where they would spend more time dealing with machines than people. Now, some years later, he exudes passion for the subject and is on a mission to help complete what he sees as its transition from the back room to the front line as computing claims an increasingly pivotal role in industry, science, commerce, education, entertainment, and health.

He was introduced to computers at the age of eight when his dad brought home a Sinclair ZX80 kit computer that they worked together to build.

‘Like a lot of people my age, I guess I was hooked at an early age’, he said. ‘I was part of that first generation that learned to write computer programs at home for fun’.

At the University of Sussex, Seth avoided a straight computer science degree by studying for a BA in Psychology and Computer Models, an interdisciplinary course that combined psychology, philosophy and computer science. Lectures on artificial intelligence and cognitive science drew his attention to the ways in which scientific disciplines could overlap and to the potential for both radical progress and sticky problems at such intersections.

‘Since then, I have always been interested in working across disciplines,’ he said.

At the same time, Seth became more aware of the extent of computer science’s image problem.

‘The department was full of young, male computer science students who bought into the idea of working as a coder or a “sys admin”, whereas the cognitive science cohort stood out as much more diverse,’ he said. ‘But over the three years of my degree program, there was a steady trickle of “cogsci” students leaving or changing course, (mostly to psychology), disillusioned with the computing element of their degree.’

There is still a mis-match, he claims, between the massive and growing diversity within computer science as a subject, and the relative lack of diversity in computer science as a community. ‘There’s a need to bring in not just more women, but people who would not necessarily think of themselves as computer scientists.’

‘Today, computing is everywhere,’ he said. ‘It needs to come out of the back room and be wider so that it can take a full part in key projects.’

Seth went on to do a PhD in Artificial Intelligence and Computer Science at the University of Sussex from 1993 to 1997. During his studies, through working on evolving simple insect-like robots, he became interested in another discipline: biology. This led him to explore whether simple computer simulations could be used to understand the evolution of animal behaviour.

‘My PhD got me more and more fascinated by biology’, he said. ‘I became very interested in how computer simulations could be used properly. Computer scientists and biologists are very different, with different training and different ideas – standards for dealing with the place where, say, artificial intelligence and evolutionary biology meet are still being developed.’

Seth got an opportunity to explore the relationship between computer science and biology further when he went to work, first, at the Max Planck Institute for Human Development in Berlin in 1997, and, then, in the University of Leeds’ School of Computing in 1999. He joined the University of Southampton’s School of Electronics & Computer Science (ECS) in 2005 to help establish the new Science and Engineering of Natural Systems (SENSe) research group.

‘I came here because it is one of the best places to do computing research and the School had made a commitment to my area’, he commented. ‘My research is now part of the School’s vision.’

The new SENSe group will develop ECS’ research at the interface between biological and computational systems. The group’s twin aims are to use computational approaches to further our understanding of biological and other natural systems, and to exploit this understanding of natural systems in order to develop novel computational tools and techniques.

Some of this work takes place within a field called complexity science, which Seth believes could revolutionise the way that engineered systems are designed and deployed. For instance, the government’s attempts to establish large-scale IT systems to handle passports, pensions, traffic, etc., are examples of battles with complexity. He thinks that one of the reasons these projects have tended to come unstuck lies in our expectation that traditional divide-and-conquer engineering approaches will successfully scale up to cope with larger and more interconnected systems.

‘For complexity science to be judged a success, it will have to provide the tools to build these systems’, said Seth. His group has already embarked on an ambitious £1.5m, three-year project: ‘Spatially Embedded Complex Systems Engineering’. Funded by the Engineering and Physical Sciences Research Council (EPSRC), the project has brought together experts in neuroscience, artificial intelligence, geography and complex systems, to understand the role of spatial organisation and spatial processes in complex networks. Through this, they will explore the possibility of engineering large-scale networks based on natural systems.

Seth believes that his group’s approach to computer science will make the discipline a more viable subject and enable IT to become integrated into the core values of business. In working backwards and forwards across the interface between natural and computational systems, the group expects to discover and articulate general design principles – ways of thinking and talking about complex systems.

‘Natural systems are rarely hierarchical, formal or ordered,’ he said. ‘The promise that nature holds is that enormously sophisticated systems can be robust, flexible, and even intelligent, despite being made-up of parts that are individually fallible, put together in a slightly sloppy fashion, operating without central control or direction – it all depends on how you organise the bits. Complexity science is in helping us to understand how this can be the case, and computer science will need to learn lessons from it.’