A Fuzzy Ambient Intelligence System for Third Generation Home Telecare

Telecare systems use a combination of sensors and communication technologies to detect remotely the requirements of people in need of medical care or other assistance. We can distinguish three generations, with the first generation using pull-cords or pendant alarms to summon help when needed and the second generation using sensors to detect emergencies. Third generation telecare systems adopt a more pro-active approach, aiming to give early warning of possible emergency situations. This requires monitoring the ”well-being” of a person defined in terms of their physical, mental, social and environmental status. By detecting changes in the daily activities of the person, it is possible to detect changes in their well-being which may not be immediately observable, but can be detected over a longer period of time. Advances in sensor design and the continuing increase in processing power make it possible to implement such an ambient intelligence system. This talk will cover the background to telecare and describe a third generation telecare system which has been tested in the homes of elderly clients. The system involves a customised sensor network, able to detect a person’s movements and use of appliances and household items, coupled with a sophisticated fuzzy data analysis process. The talk will focus particularly on the intelligent processing which enables the system to take low level data (e.g. kitchen sensor activated; cold water run for 20 seconds; kettle switched on for 60 seconds; fridge door opened) and answer questions such as ”is the occupant eating regularly”, ”is the occupant’s social interaction increasing or decreasing”, ”has the occupant’s sleep pattern changed significantly in the past few months” etc. This is a very substantial inference and learning task and the nature of the data and queries make a soft computing approach a natural choice.