Cellular robotics: simulation and HW implementation

Aspects of self-organization are presented in this paper. Computer simulations as well as a real prototypical implementation are used to illustrate the proposed approach. Results of simulations are presented to compare different strategies of self-organization enabling a system of autonomous robots to form a chain between two landmarks in a completely unknown environment. This chain implicitly represents a path between any two points of the environment without an explicit representation of free space (no single robot has a global map of the environment). The experimental part, even if restricted to a few robots, demonstrates that the set of stimuli-action processes used in the simulations are indeed feasible on real systems.<
>