Using fuzzy technology in ambient intelligence environments

Ambient Intelligence (AmI, shortly) gathers best results from three key technologies, Ubiquitous Computing, Ubiquitous Communication, and Intelligent User Friendly Interfaces. The functional and spatial distribution of tasks is a natural thrust to employ multi-agent paradigm to design and implement AmI environments. Two critical issues, common in most of applications, are (1) how to detect in a general and efficient way "context" from sensors and (2) how to process contextual information in order to improve the functionality of services. In this work we experiment a framework where hybrid techniques (distributed fuzzy control, mobile agents, fuzzy rules induction algorithms) are mixed to gain flexibility and uniformity. © 2005 IEEE