A smart sensor based on rules and its evaluation in daily routines

Abstract
Our purpose is to automatically detect in real time the fall of the person from a single device, worn on the trunk in the "para sagittal" plane. The device includes 2 accelerometers, an 8 bits RISC microcontroller, a buzzer and a push button. It is serially linked to a RF-modem. It was tested on 10 healthy young persons in 15 different situations and exhibits good specificity and sensibility. Improvements can be made on the decision algorithm using non standard theories.