Wearable Multistage Fall Detection and Emergency Alert System for Assisted-Living Older Adults

Abstract

Falls are becoming increasingly serious issues for elderly people. In fact, one out of three individuals aged 65 and older will sustain at least one fall per year. Unfortunately, existing fall detection systems still have to overcome problems of false alarms and sensitivity in order to be effectively employed. This article introduces a new smart wearable device with an advanced multi-stage fall detection algorithm that successfully distinguishes between falls and ordinary actions. This system is based on the ESP32 DevKit V1 microcontroller, MPU6050 inertial measurement system (accelerometer and gyroscope), SIM800L GSM module, buzzer, pushbutton, and LED indicator, powered by 3500 mAh rechargeable battery. The fall detection system works by sequentially analyzing free-fall, impact, orientation change, inactivity after fall, and final standing position with regard to the user-calibrated reference. The results show that the accuracy of readings in tests reached 96.0%, with false alarms rate equal to 4.0% based on 150 simulated experiments. In addition, it was observed that system could send emergency SMS alerts within 12 seconds of confirming a fall. Another benefit is that the Wi-Fi dashboard facilitates remote management of contacts as well as monitoring contacts in real time. This inexpensive prototype offers an innovative means for ensuring a high level of safety among the elderly and providing them with an independent lifestyle, with the future upgrades in plans, such as the addition of GPS and memory capabilities for storing contacts.