My primary research interest is in Artificial Intelligence, in the "understanding the nature of intelligence by attempting to synthesise it" sense, not the "solving hard problems" sense. The basic philosophy underlying my approach to AI is the idea (made popular by Brooks in the late 80s, but much older than that) that AI is a robot problem. Because of this I am interested in solving the kinds of problems that embodied agents situated in a complex, dynamic environments must solve. Ultimately, I want to build highly adaptive, layered robot control architectures that will provide abstract insight into the learning and control hierarchies that must be present in animals and humans.

My current focus is on Reinforcement Learning, because while discrimination and classification are important (and interesting and hard), the first and most important function of the brain is control. RL is the only machine learning paradigm that really addresses the issues underlying autonomy and motivated learning for agents embedded in an environment in which they are capable of taking action.

My thesis work is on autonomous robot skill acquisition. Skill acquisition refers to the ability to create new skills, refine them through practice, and apply them in new task contexts. The ability to retain and refine solutions to important subproblems and then employ them later lies at the heart of two important aspects of human intelligence: the ability to perpetually improve at difficult control tasks through learning by practice, and the ability to solve increasingly difficult problems. Although robots exist that can learn a skill given a reward function and problem specific state space, none exist today that display truly autonomous skill acquisition. I am developing methods to remove the remaining barriers to autonomous learning.

During some of my time at Edinburgh I worked on the Hydra project, which aimed at investigating the principles underlying the co-operation of relative simple units (think cells) to construction more complex units (think arms and legs) and to generate complex behaviour, using only local communication. That experience has left me with a secondary but enduring interest in Artificial Morphogenesis and Self-Assembly Robotics. I also used to occasionally dabble in Artificial Evolution and Evolutionary Methods, where I was interested in the properties of the evolutionary process, rather than the use of genetic algorithms to solve hard optimisation problems.

During my undergraduate years at Wits my research was mainly in Computational Geometry, but I don't do that at all anymore. I have, however, retained a strong interest in the problems facing Computer Science Education in South Africa.