Modeling Situation-Aware Ambient Assisted Living Systems for Eldercare

Abstract

The development of ambient assisted living (AAL) systems, which are tailored to health or elder care, requires specific methods and tools. AAL systems make often use of wireless sensor networks, machine learning algorithms and sensory devices. Since wireless sensor networks and their sensors are inhomogeneous, it became apparent that such systems need to cope with different hardware platforms, different programming languages, unreliable wireless communication, energy constraints, data analysis algorithms, recognition of situations, and deployment options. Developers to date tend to use a bottom-up approach: hardware components dictate the development of AAL systems and thereby restrict the range of use cases that can be realized; domain experts by contrast would prefer a top-down approach and model the system’s functionality independently from the hardware platform. Currently available software development environments and tools do not adequately support domain experts and developers to accomplish these tasks efficiently. This paper presents methods that support domain experts in their top-down approach, as well as technically experienced developers in their bottom-up approach. The implemented tools enable a model-driven software development process (from platform-independent modeling to generating AAL application code) and thus facilitate programming AAL systems.