An introduction will be given about the main principles of Swarm Intelligence, a new behavioral and computational metaphor inspired by social insects and flocking and shoaling phenomena in vertebrates. It will be shown how their application to artificial embedded platforms can lead to self-organized, fully distributed, and scalable collective control while maintaining individual simplicity and autonomy. The speaker will present methodologies currently under development in his research group for automatically designing, optimizing, and modelling collective systems that show the above properties. The discussion will be supported with a few case studies derived from a particular area in the field of sensor and actuator networks - collective mobile robotics. After having discussed results, advantages and limitations of the current methodologies, the speaker will outline a few lessons learned from the collective robotics area and speculate about their potential application to further problems and platforms, in particular those characterized by reduced mobility or total absence of self-locomotion. Finally, the talk will conclude with a summary of future challenges we will have to face in order to deploy and control large-scale, distributed, real-time, embedded systems for real-world applications.

Vortragssprache: englisch