Advanced Fall Detection System

Falls represents a major public health risk. A fall not assisted in time can cause functional impairment with a significant impact on independence and life quality. This concerns not only elderly people but also persons who work at height and extreme sports like riding, boarding and skiing.

Existing solutions are either costly (based on image recognition for indoor environment) or only suitable for a specific need. The solution proposed here take advantage from a Smartphone App and the power of the cloud to provide an unified solution for detecting falls situation.

A smartphone app coupled with SmartEdge Agile hardware (acting as a wearable device) permits Fall detection. When the device detects a fall, it executes predefined emergency protocol :

- Reading safety recommendations with a Text To Speech module inviting the user to deactivate the alert (if it is a false alert or non serious fall). After 60 seconds, the app sends an automated text message to emergency services with the home location and the live location of the user and sends the same message to the user configured friend/family member

The application is very simple :

At startup, the app connects to Brainium sever and subscribes to Asynchronous alert messages using the MQTT ( Eclipse Paho Android Client). The main screen displays the connection status et the live location of the device.

The user can configure the application with the "Settings screen".

When the app receives a notification, it displays a timed message inviting the user to cancel the alert and gives safety recommendations by voice. With no response after 60 seconds. The apps executes the alert protocol by sending to emergency services and friend/family phone a text message informing them of the situation and giving the home location and the live location of the user.

The fall motion has been trained using Brainium AI module by simulating a series of fall and capturing different sensors values (velocity, acceleration , pressure...)