General Terms Design MAS FOR PERVASIVE COMPUTING The technological progress -concerning chip density, processor speed and network bandwidth, to cite some -makes it possible to conceive new classes of applications and systems, which can be generally referred as pervasive computing. Intelligent / Smart Environments are prominent cases of pervasive computing, whose aim is to remodel the environments where people live and act, considering the new services (such as energy management, health care,

more »

... ertainment) that can be provided by embedding (intelligent) software in network-enabled subsystems, such as sensors, actuators, mobile devices and general-purpose computers. In this way, some of the tasks currently performed by humans can be automated -possibly with improvements in efficiency or quality of service [2] . Pervasive computing scenarios account for two forms of intelligence: intelligence of the individual parts, which must autonomously execute their own specific tasks despite the openness and unpredictability of the environment; and social/collective intelligence, required to provide services that necessarily involve the coordinated interaction of the autonomous parts. Since traditional software engineering approaches are known to be inadequate to face the complexity of such scenarios, new paradigms and models are required. Accordingly, approaches based on the multiagent system (MAS) paradigm *