Environmentally induced task partitioning in competing bio-robots

Competition in living beings is always shaped by learning, which, even in simple creatures, like insects, plays a fundamental role in enhancing the basic inherited capabilities. Feeding, as well as courtship, involves environment exploration and exploitation, which mainly includes local competition: when such life-saving behavior leads to a global benefit for the colony, this can be considered as a form of global cooperation, even if the single agent is not aware of this global aspect. So the boundary between cooperation and competition is rather subtle. Some key results of recent research activities based on a group of 2 robots endowed with a simplified structure of an insect brain model are here generalised to include more robots and different scenarios. The results obtained consist in the role Specialization of each robot, induced by local competition. This helps in decreasing the overall energy spent to fulfill a given task from the whole group, via environmentally induced learning of behavioral sequences. The description of the neural network and learning mechanisms used for robot specialization and sequence learning will be presented in the paper. Some interesting simulation results will discussed and remarked to show the potentiality of the approach.