Adaptive swarm robotics with social/evolutionary learning
Abstract: In this talk, I will give a brief overview of evolutionary swarm robotics, which is the use of evolutionary algorithms to endow a swarm of robots with a capability to learn new behaviours. While originally loosely inspired from the process of biological evolution, these algorithms face challenges that are usually not addressed in robotics, but arise from the interactions of robots with their environment, and with each other. I will show how robots can socially learn to coordinate in an unknown environment based solely on local interactions, as well as how robots can learn to perform morphological computation, which means to use one's own body, rather than one's brain, to solve problems such as coordinated motion and search for resources. I will also give a glimpse on how issues arising from adaptive swarm robotics echoes those met during the study of evolution and social learning of cooperative behaviours in nature.
Predictive Coding for Cognitive Development in Humans and Robots
Cooperative control of environmental extremes by artificial intelligent agents