Edutainment Robotics: Applying Modern AI Techniques

Abstract

With the development of cheap robotic tools, it has become possible to allow children to use robots as a toy and in educational environments. For the purpose of increasing children's awareness and knowledge about technology, we have developed tools that allow them to interact with robots in an easy and straightforward manner, e.g. exemplified through our design and realisation of RoboCup Junior. Some of these techniques arise from the fields of evolutionary computation, adaptive systems, agents, and artificial neural networks and we show how they can be used in edutainment robotics in order to provide easy access to the robot technology. The user-guided approaches that we developed include user-guided behaviour-based systems, user-guided evolutionary robotics, user-guided co-evolutionary robotics, and morphological development. All these techniques are applied to allow children to develop their own robot behaviours in a very easy and fast manner