A survey on situation-aware ambient intelligence systems

Due to the demographic change in the population the costs for healthcare and nursery will drastically increase in the coming years. Ambient Intelligence (AmI) technologies will never replace personal care, but can help to decrease costs by relieving care personnel and allowing elderly to stay independently at their own home for a longer time. Improvements in sensing technology lead to well-engineered and context-aware devices, that ease ones daily living activities, but there is still a gap according to data fusion. Current AmI applications face issues, such as: (i) proper situation classification, (ii) reduction of interpretation faults, (iii) action triggering, and (iv) projecting the future situation evaluation. There exists a huge amount of prototypes and systems, motivated by different requirements (use cases for processing environmental data, vital data, health data), which still have open issues when a global view on the data is needed for gaining situation awareness (SAW). This paper gives an overview of existing projects and seminal developments within the scope of AmI. After a survey on various AmI projects a common reference architecture was figured out and evaluation criteria focusing on SAW were defined. The criteria were used to classify selected projects and identify obstacles that avoid gaining SAW. In the end open issues and potential improvements are discussed and a perspective for further developments is given.