Self-Assembly in Physical Autonomous Robots - the Evolutionary Robotics Approach

Elio Tuci, Christos Ampatzis, Vito Trianni, Anders Lyhne Christensen, Marco Dorigo
2008 Workshop on the Synthesis and Simulation of Living Systems  
This research work illustrates the details of a methodological approach to the design of homogeneous neuro-controllers for self-assembly in physical autonomous robots in which no assumptions are made concerning how agents allocate roles. Artificial evolution is used to set the parameters of a dynamical neural network that when ported on two physical robots allows them to coordinate their actions in order to decide who will grip whom. The neural network directly controls the state of all the
more » ... ators. To the best of our knowledge, this work is the first example in which physical robots manage to self-assemble without relying on a priori injected morphological or behavioural heterogeneities. The results shed a light on the minimal requirements necessary to achieve self-assembly in autonomous robots.
dblp:conf/alife/TuciATCD08 fatcat:lwg6ihxhg5c7dbw7aozyd6cmqe