6 A Flexible Architecture for Ambient Intelligence Systems Supporting Adaptive Multimodal Interaction with Users

Abstract. Ambient Intelligence (AmI) systems aim at augmenting real environments to create Smart Spaces where users are provided with pervasive virtual services. In order to allow users to seamlessly complete their tasks across a multitude of smart devices, and across different physical locations, the AmI infrastructure must be complemented with ubiquitous Intelligent User Interfaces able to adapt the interaction to their characteristics and needs. In this chapter we address three main issues related to supporting an adaptive interaction between the user and the Smart Space: how can we approximate in an AmI system the deep understanding of the context typical of human beings in the user interface of a Smart Space? How can we replicate the richness and flexibility of human-to-human communication? How do we design a flexible architecture which can effectively link the context sensing and multimodal communication functionalities? The proposed approach is twofold: on one hand, we use a neuro-biological model of awareness as a blueprint for the design of a flexible architecture fo