Fall detection watch for independent living of the elderly

author: Jani Bizjak, Department of Intelligent Systems, Jožef Stefan Institute
published: May 23, 2017,   recorded: April 2017,   views: 5
Categories

Slides

Related content

Report a problem or upload files

If you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.
Lecture popularity: You need to login to cast your vote.
  Bibliography

Description

The developed world is facing the problem of aging population. Better living conditions, improvements in food production, and advances in medicine have drastically increased living expectancy. On the other hand, natality is falling. This results in an upside-down demographic pyramid, and, as a consequence, an imbalance between the number of elderly that require some form of care or assistance and people who can assist them. Introduction of ICT technologies is promising to provide better care for the elderly and to prolong their independent living, at the same time reducing the burden of the carers and the health system.

We propose a state-of-the-art solution on a wristwatch that monitors user’s actions and automatically calls for help in case of detected danger. The solution proposed currently supports three different standalone android watches (commercial products) that have in-built GSM, GPS, and accelerometer. Optionally, two models also have a heart rate sensor that can increase the depth of information provided by the watch. One of the watches is also water resistant and can be charged by magnet charger – which is important; majority of elderly have bad eyesight and it is hard for them to charge the device over a standard microUSB without using glasses.

The main feature of the watch is an automatic fall detection algorithm that detects falls and other dangerous situations (example: car crash). The algorithm monitors accelerations and detects fall like motion when fall occurs. Because watch is usually worn on wrist, there can be many fall-like motions during different activities. To prevent false alarms, a 20s period after the perceived fall is also monitored. If there is no substantial motion afterwards, the system interprets this as a fall and calls for help. The rationale behind this implementation is that if the user can move after the fall, he can press one button to call for help manually. To detect other potentially dangerous situations, the watch also monitors general daily activities of the user and compares them to the last 10 days. This allows the system to detect unusual situations. For example, if the user is usually going for 2 hours walk after lunch each day and then one day he is sleeping whole day in bed, the watch will detect this and alert the carers that something unusual is happening.

People suffering from dementia sometimes leave home for a walk and get lost, perhaps ending up several kilometres away from their home and have no memory on how to get back. Thus, the watch also offers the ability to get the user’s location in case of such emergency. Due to privacy concerns, the user or his legal representative have to agree with this and location can only be retrieved in case of emergency.

The watch is managed through a web portal that has the access to all watch information. Most of the watch functions can be managed remotely, such as changing phone numbers in case of emergency (carer or a call center), modifying the thresholds for fall detection algorithm, or setting reminders and notifications. To preserver the user’s privacy, the portal distinguishes between multiple roles, i.e. informal or formal carer, administrator, etc. Each user (the elderly) can then be assigned a carer (either formal or informal) and only that carer can then access his information. Additional scheduling tools are also available for formal carers in order to ease the management of home visits.

See Also:

Download slides icon Download slides: ipssc2017_bizjak_fall_detection_01.pdf (727.3 KB)


Help icon Streaming Video Help

Link this page

Would you like to put a link to this lecture on your homepage?
Go ahead! Copy the HTML snippet !

Write your own review or comment:

make sure you have javascript enabled or clear this field: