Runtime models for self-adaptation in the ambient assisted living domain

Ambient Assisted Living systems (AAL) must fulfill several challenging requirements as, for instance, the ability to render their services at a quality level that is high enough to enable an independent living. This requires a sound understanding of the current situation of the users, their environment, and, the availability of required resources. Further, AAL systems need the ability to adapt and extend the system behavior, as the demands for living assistance substantially differs between different individuals (differential aging) and changes during a person's life. Both are requirements for a sound adaptation support at runtime that require adequate models. In this paper we identify stereotypical adaptation scenarios in the AAL domain, elaborate on components and their respective models to support the adaptation scenarios, and discuss the evolution of these models.