Theme 5: Learning
One of the most challenging aspects of robotics is how we build robots that can learn or adapt their behaviour to new situations.
Learning can take two forms: supervised learning, in which the robot is 'taught' new skills, or unsupervised learning in which the robot learns from trial and error, or its own experience. A quite different form of adaptation is 'evolutionary', in which we evolve robots to improve their design or behaviour using a process of simulated evolution.
The later part of the project also provides the opportunity to learn from what we have achieved in terms of a multidisciplinary and multi-sector group of experts engaging with the public across this broad and far reaching set of themes. We will review what we have learned from this project, and what we have learned from robots.