Adaptive oriented clustering in difficult environment of collective robotics

Purpose - The collective intelligence emerging from behaviors of social insects has become an inspiration source that is impossible to avoid; guiding researchers in various domains to solutions of insolvent problems by traditional approaches. These behaviors are made possible because of the interactions individual-individual and individual-environment, representing support on which cooperative work within the same group is based and allowing emergence at macroscopic level of sophisticated achievements. Many models were inspired by this new and very promising vision, to find simple rules, leading mobile, autonomous robots with limited capacities in their environment to realize tasks, like those of: browsing, collecting or self-assembly. In this context, the purpose of this paper is to suggest a method, making global behavior evolve within an homogeneous agent-robots community to accomplish heap-formation task based on appointment principle in changing environment which can be very difficult. Control device, comparable to the functioning of cellular automaton containing sensory-motor rules, is then used to arbitrate between some given elementary attitudes with which each agent-robot initially is equipped. Design/methodology/approach - Evolutionary approach using genetic algorithm based on reverse emergence principle seeks, then, for cellular automaton whose arbitration succeeds to realize this adaptive oriented grouping task. Findings - Rules as simulation results obtained according to reactive model of multi-agent systems are provided, compared with those found at the ants and commented. Originality/value - Discovered rules are adaptive; it means when training ground becomes more difficult, agent-robots become more flexible by decreasing thresholds conditioning rules application. If environment state continues to turn into harsh, robots are able to seek for another direction to start new heap formation somewhere else. Such zones are like Saharan region, airports or supermarkets.