 |
|
I am Professor of Intelligent Systems in the Department of Computer Science, at the University of Wales, Aberystwyth. I have been founder and Director of the Centre for Advanced Software and Intelligent Systems and a co-founder Director of the European Network: MONET.
My research interests concern new methods of learning for robots. In particular, I am interested in why and how human infants can achieve so much learning - and robots so little! During the first year, infants learn at fantastic rates; they gain control over complex sensory-motor systems, build up an understanding of three-dimensional space, learn to recognise subtle patterns, and start to communicate and interact with others. If only our robots could grow like this!
Despite the long history of Artificial Intelligence, we still have very primitive robots - especially in terms of what they can learn. I believe the reason for this is that we do not give enough attention to development, which is the key to human growth and learning. The new field of developmental robotics is now being established, with the aim of recognising processes of growth and change in human behaviour and building models, mechanisms and algorithms that will exhibit growth of competence and intelligence similar to that observed by psychologists.
I have worked on this issue, on and off, for many years and am now in a very active stage with several on-going projects. A previous project called DVL , (2002-06, EPSRC funded) used data from psychology to explore developmental learning methods in sensory-motor systems, and this project laid some foundations for current work.
The REVERB project is funded by EPSRC and is in collaboration with a neuroscience group at Sheffield and several other universities and aims to reverse engineer some of the action selection machinery in the brain. REVERB uses a 7 DoF robot arm and a fast panning camera system.
The ROSSI project is a recent FP7 3-year project and is looking at the nature of the internal mental models necessary for robots and humans to interact and cooperate. Clearly, robots wont have mental models like humans but what degree of shared world understanding is needed to allow useful interaction? ROSSI is exploring both robot grasping with a three fingered tactile robot, and human grasping and manipulation with data capture equipment.
The IM-CLeVeR project is another new FP7 funded project and is aimed at applying the methods and concepts of developmental robotics to a fully equipped humanoid robot, in this case the iCub robot from IIT in Italy. This is a 4 year project, involving 7 partners and 3 iCub robots.
I am also excited by the prospect of building brain-like machines that may offer interesting alternatives to the standard computing model. I am a member of committee 5 of the Grand Challenge programme, "The Architecture of Brain and Mind" and my entry is in section 5, see, e-Brain, and GC for the programme.
My other research interests include Model-Based Reasoning with applications in engineering. I originated a novel modelling method for electrical circuit analysis and this developed into a major research and application area. I am currently developing a many valued extension to the method.
Disclaimer: Information provided on this page and other pages by me, Mark Lee, is under my own personal responsibility and not that of the University of Wales, Aberystwyth. Similarly, any opinions expressed are my own and are not to be taken as those of UWA.