智慧型跌倒緊急通報物聯網機器人

Smart Fall Emergency Alarm IoT Robot

臺灣進入高齡化社會，根據國健署統計年長者105年事故死亡原因，跌倒位居第二名。許多年長者身邊無人照顧，若跌倒沒人發現，易造成憾事，研究主要目的為設計物聯網感測設備與管理平台，偵測年長者是否跌倒並通知家屬、照護者及救護單位。物聯網設備以ADXL345三軸重力感測器為核心，結合LoRa傳輸元件將資料傳送中繼平台後上傳至雲端，照護者可從平台接收年長者狀況，並於緊急狀況發生時，進行救護通知。研究實際模擬1000筆跌倒及正常行走數據後，分析設計出跌倒感測演算式，實際跌倒偵測準確率高達95％、誤判率2.1％、漏判率2.9％。

Taiwan entered an aging society. According to statistics of Health Promotion Administration, falling over ranked second from the death of the elders accidents in 2016. Many elders live by themselves. If they fall over and nobody knows, it is easy to cause regrets. The main purpose of the study is to design an IoT sensor device and management platform to detect whether the elders fall over and notify the family, caregivers and ambulance units. The IoT device is based on the ADXL345 three-axis gravity sensor. The LoRa transmission component is used to transmit the data to the platform and then upload it to the cloud system. The caregivers can receive the elders status from the platform and provide an ambulance notice when the emergency situation occurs. The research simulated the 1000 data of falling over and walking, and then analyzed them. The accuracy of detection rate was 95%, the false rate was 2.1%, and the loss rate was 2.9%.