and exploiting the theoretical framework of the multi-agent Abstract State Machines (ASM) [12], we here show how to model and analyze in ASM the behavior of a self-adaptive system by representing MAPE-K loops explicitly and naturally formalized in terms of agents’ actions (ASM transition rules). Most existing formal approaches to SA assume a centralized point of control and none of them, except [24], [35], [1] (see Sect. II), provides a semantic description of the adaptation logic from the perspective of feedback control loops.