In future distributed systems, the appearance of mechanisms that should be able to realize robust, predictable and self-adaptive behaviour for existing networked systems can be seen as a critical enabler for further growth. If, in addition, such mechanisms are related to options for achieving cooperation of several (heterogeneous) network elements and/or domains, with the aim of reducing energy wastage and minimizing energy consumption, then the challenge becomes critical. To this aim, the present paper discusses an innovative approach, focusing upon the development of energy-aware distributed and cooperating systems for monitoring and control, centred on appropriate wireless sensor networks for the provision of radio and environmental context information. In particular, we discuss the concept of a “Self-Growing Network” that is able to efficiently incorporate autonomic and collaborative capabilities, intending to increase the level of the global network energy efficiency by using cognition-based features. Self-growing capabilities are enablers for a novel type of network flexibility. They allow optimising a heterogeneous collection of network nodes or sub-networks to be dedicated to a specific optimisation “purpose” temporarily and on-demand.