Abstract |
Ambient Intelligence (AmI) is an emerging research field that aims to support and enhance the activities of everyday life. This new paradigm gives rise to opportunities for novel, natural and more effective interactions with computing systems. In this context, we propose a holistic framework for modeling and programming AmI environments. Through the proposed work, we aim (a) to define flexible and intuitive software abstractions for modeling all the different entities that comprise an AmI environment (architectures), (b) to suggest systematic, well-defined processes that identify and highlight all the necessary steps for developing AmI environments from the design to the deployment phase (methods), and (c) to provide programming libraries and tools that fully support and streamline the application of the defined architectures and methods (tools). The contribution of this work is twofold, as the focus is on the architectural and theoretical aspects of building AmI environments, but also on the practical issues that stem from the development process. In this regard, our approach is structured over the provision of services for exposing the resources of AmI environments through a unified homogeneous access layer, and the definition of smart objects that provide well-defined abstractions, dynamic interaction capabilities and extended composition potential for implementing AmI-related functionalities. The utilization of the concepts and tools proposed in this thesis improves existing practices (i) by maximizing the potential for incorporating heterogeneous technologies, (ii) by allowing for the comprehensive modeling of all aspects of functional AmI environments, (iii) by streamlining the whole implementation process, and (iv) by enabling the simulation of the target environment’s behavior. Ultimately, these mechanisms allow for succinct implementations, minimizing the required effort for building, deploying and extending fully functional AmI environments.
|