Elderly people present a unique health situation-- often being vulnerable to falls and forgetfulness. Falls are a major cause for old age morbidity, costing the health industry over $30 billion per year. Considering it was reported by the Administration of Age that almost 47% of women aged 75 or older lived alone in 2010, this is concerning.(http://www.aplaceformom.com/blog/2013-4-1-dangers-ofseniors-living-alone/) Faide is a connected smartwatch application that aims to act as a personal health assistant. Using the pebble’s accelerometer, it can detect when an elderly user has fallen down. This prompts the watch to give the user 10 seconds to respond whether help is needed, or if no response is given, immediately sends for help by contacting caregivers through text and email. A webapp acts as a visualization dashboard for the caregivers (and the user) to see how the user has been doing-- how many falls? medication taken on time? weight changes? In the later development we hope to set it up so the caregivers can also set notifications and medicine reminders from this interface. Another feature of the app is the use of geo-fencing to alert the user that they are next to say a gym, or weighing machine. Using the smartwatch, they can quickly make data entries about their weight; for example, by looking at large weight fluctuations on the dashboard, caregivers might suggest the intake of a water pill.

We use the Pebble smartwatch which is waterproof and can safely be worn all throughout the day. We developed an algorithm to detect falls automatically through empirical experimentation using accelerometer data and testing. When it detects a fall, it seeks a confirmation from the user. If s/he fails to respond or seeks help, a message is forwarded to the webapp. The webapp in turn uses Twilio and SendGrid to send text messages and emails to the subscribed caregivers. We also demonstrate the use of geofencing by using an Estimate Beacon kept next to a bathroom scale. The phone detects its presence and sends a notification to the Pebble to manually enter the user’s weight through the watch.

Share this project: