Motion recognition using 3D accelerometer sensor network for Mobility Assistant Robot

Abstract

The authors proposed a 3D Accelerometer Sensor Network for Mobility Assistant Robot. The Mobility Assistant Robot is composed of a Mobility Robot, a Smart Robot and ZigBee Accelerometers. The Mobility robot is used to assist walking and standing up. It serves as both table and walker so user can take both advantages. The Smart Robot communicates with user by their gestures and utterances. A ZigBee accelerometer is attached on user's waist to record user's conditions such as room position, posture, walking, getting up, and falling down. The experimental results showed the proposal method recognizes user's falling down with 98.4% recognition rate. With 3D accelerometer, the Mobility Assistant Robot identifies the person who is the nearest and quickly informs the user's falling down conditions.