Modeling Context-Awareness in Agents for Ambient Intelligence: An Aspect-Oriented Approach

Ambient Intelligence (AmI) systems are inherently context aware, since they should be able to react to, adapt to and even anticipate user actions or events occurring in the environment in a manner consistent with the current context. Software agents and especially the BDI architecture are considered to be a promising approach to deal with AmI systems development. However current agent models do not offer a proper support for developing AmI systems because they do not offer support to model explicitly the interaction between the agent, context sources and effectors, and the context-awareness features are scattered in the system model. To solve these problems we propose an aspect-oriented agent metamodel for AmI systems, which encourages modularity in the description of context-aware features in AmI systems. This metamodel achieves better results than other metamodels in separation of concerns, size, coupling and cohesion.