Guaranteeing the continuity and the quality of services in network plants is a key issue in the research area of asset management. Especially when the plants are located in a wide area where machines are not continuously monitored by the operators. In particular, the pervasive adoption of smart sensors could be able to develop intelligent maintenance system through an elaboration of data coming from the machines: this data could be processed by diagnostics algorithms to warn preventively the fault status of the components or machines monitored. The algorithms’ structure is contained in a multiple system of agents that have different tasks to manage both the single machine and the information exchanged within the whole system. This paper aims to present an application of Artificial Immune System defining, for each plant section, the kind of agents employed and the related sensors that must be adopted to collect the useful data. In order to provide a practical example, the structure of an Artificial Immune System has been implemented in a wastewater treatment plant where the agents are tested with noteworthy results. © 201649285560