What is a Wrist Fall Detector
Wrist Fall Detector technology enables individuals to wear medical alert systems on their wrist, which automatically send an alarm signal if they fall. This signal may notify an emergency response center of need or simply let the user know they are alright so they can resume normal activities without worry. This solution may be particularly suitable for people who struggle to press the button on a traditional fall detection pendant when they fall, or cannot press it at all.
Falls can be extremely hazardous when they happen in a domestic setting where help is not immediately accessible. A person who falls and cannot get up can quickly become dehydrated, starve, experience muscle tissue breakdown or hypothermia, develop pressure sores, or develop other serious health problems before receiving assistance. Furthermore, those who remain lying on the floor for an extended period could risk breathing fluid into their airways from lying still for too long and contract pneumonia or suffer adverse side effects from any medication they're taking to reduce falls.
One can now find various devices on the market that can detect falls and call for assistance immediately, including bracelets, watches and pendants from medical alert systems. They're typically worn around the wrist or neck; many companies are now testing wireless versions that don't require wearing any pendant.
These products often utilize sensors such as accelerometers to detect motion and identify falls, then notify a medical alert company's monitoring center that someone needs help. Many devices include a help button in case wearers cannot press their alert button themselves; although many companies provide disclaimers that no device can ever be 100% accurate, medical alert products remain an invaluable safety measure and could save lives.
falling alarm Though these technologies are relatively new, their technology is rapidly progressing and more companies are coming out with innovative options - some without pendants at all - for monitoring movement in rooms.
Current studies relating to elderly individuals using wrist-based fall detection lack data, with prior work having focused on head, chest or waist wearables that were uncomfortable and associated with stigma. Wrist wearables on the other hand are easily integrated into daily life and worn multiple ways - less likely to cause discomfort or look like medical devices than their predecessors.
This paper introduces a fall detector designed for integration into smartwatches. Using a low computational FS-1 feature set and taking advantage of limited memory and hardware resources on a smartwatch's limited memory and resources to provide a responsive yet accurate fall-detection algorithm. Testing was carried out on an innovative wrist-based dataset which has never before been publicly released; evaluation revealed good performance as well as false positive resistance comparable to results obtained using more sophisticated approaches in previous work.
