Self-healing in ensembles’ adaptive collaborative patterns

Many of today's complex applications are composed of ensembles of service components which interact and collaborate with each other to satisfy user requests or to build higher-level services. In any case, we can recognize different collaboration patterns, which we have classified in a preliminary taxonomy considering autonomic features.In this paper we analyze the collaboration patterns from the self-healing point of view, in order to point out the support that each pattern can provide to self-healing features of the whole ensemble, and not only of the single component.Moreover, we will provide a concrete example by means of a case study based on autonomous robots that search for food in an unknown environment to fill a common nest.