Ultra-wideband monitoring sensor with pattern recognition

Elderly-care support is a very important issue in many developed countries because of the arrival of increasingly aging society. Actually, fall accidents often take place in care facilities when the elderly attempt to leave the bed alone. So, a sensor estimating their state and motion with ultra-wideband are gaining a lot of attention because of preventing those accidents. In this paper, we propose an algorithm to estimate their motion on a bed accurately with support vector machine (SVM) which is one of pattern recognition. This paper presents the measurement results conducted in a care/ nursing home about the effectiveness of the proposed algorithm.