Danger warning via fuzzy inference in an RFID-deployed environment

Abstract

The population structure of some developed countries has been changed. Aged population with fewer kids is the main problem. To relieve the burden for caring elder people and kids at home, automatic mechanisms for homecare are needed. In this research, we focus on home safety with the elders and the kids. We aim at developing a RFID-based system that can detect the movement of the user at home. When the user approaches a dangerous location or a dangerous object, the system issues a warning to the caregiver to prevent possible dangerous situations. An active RFID tag is placed at each location and near each object. The user carries an RFID reader which detects the signal strengths of all tags and transmits them to the system in real time. The system issues a warning to the caregiver when the dangerous degree for a dangerous location/object is above the predefined level. A dangerous situation can be prevented if the caregiver watches out beforehand. The dangerous degree is determined through fuzzy inference on the user age and the signal strengths which reflect the distance of the user to a dangerous location/object. Fuzzy membership functions and fuzzy rules are defined in this system. A feedback mechanism is also designed to provide personalized services, which simply modifies the default fuzzy membership function of the corresponding location or object. Experimental results demonstrate that the system is promising in this application.