Fall Detection With Wrist-Worn Watch by Observations in Statistics of Acceleration

It is common for older people to live alone, which can have tragic consequences if they an accident and can’t call help in time. This particularly acute aging society where falling one of the most accidents. According CDC, 1/4 over age 65 United States fall each year. The development IoT MEMS has made it possible detect falls time automatically help. presented detection system focuses on walk-fall-still pattern, collects accelerations through wrist-worn M5StickC-Plus watch, analyses data locally detects using algorithm based observations statistics acceleration second, then transmits alarm signal a remote healthcare real-time via WIFI. lightweight been proven be 90% accurate detecting falls, notify service staff accidents within 1 second. features comfort, lightness, timeliness make device more practical than similar products. low-cost, non-intrusive used care homes also suitable elderly living alone.