“Swarms” of living organisms behave intelligently as if the entire group were a single individual with a will. This behavior emerges from local interactions among the components of the swarm (hereinafter referred to as "autonomous individuals") and is referred to as “swarm intelligence.” This study aims to elucidate the design principles of heterogeneous swarm intelligence systems, in which autonomous individuals with various characteristics find appropriate roles in a fluctuating environment, create order, and continue to demonstrate high functionality.