Recently an autonomous system that meets the growing demand for solving a problem in adiversifying environment has been desired, in order to secure more robustness and flexibility. As a first step towards establishing such an autonomous system, we developed Life Algorithm (LA). Life Algorithm is fundamentally an extension of Genetic Algorithm (GA). While the conventional GA approach is based on genes and is homogeneous in its essence, LA is something more like a genome and realizes more diversity. A group of genes with different evaluation function is considered and these constitute agents. Thus, more individuality and more interaction between these agents can be taken into account and more robustness to cope with the widely and quickly changing environment can be realized. This paper demonstrates the usefulness of an LA approach by illustrating the example of power plant operation through simulation and an experiment of a solar cell system.