‘Elderly Monitoring System with Sleep and Fall Detector’ by Abdulakeem Odunmbaku, Amir-Mohammad Rahmani, Pasi Liljeberg, and Hannu Tenhunen
Best paper at HealthyIoT 2015, 2nd EAI International Conference on IoT Technologies for Health Care
Western population is growing old. While we don’t usually describe the aging prognosis as a ‘demographic crisis’, our current trajectory certainly has a lot of people worried. The population of the 60 and over age group is expected to reach 1.2 billion by the year 2025, and 2 billion by 2050, and regardless of how accurate these prognoses are, the trend is an undisputable fact. And it is about to put extreme pressure on our health care infrastructure. As the authors point out (and tackle the issue directly with a system design), IoT-powered technology will need to be one of the pillars that takes a load off of human resources – if the elderly are to be taken care of.
Two main activities in the elderly population have been put forward in previous research as activites that are relatively easy to track, but have significant value for diagnosis and crisis detection. Sleep monitoring and fall detection are two aspects which have been a focal point of many applications in the past already. Since low quality of sleep can lead to extensive health problems (such as high blood pressure), it can be a telling ambient indicator. Furthermore, a simple fall can result in a permanent handicap for an elderly person, especially when not treated immediately. Efficient fall detection can thus be a crucial element in preventing critical conditions for the elderly.
While assistive technologies have focused on either one or the other in the past, the authors of this paper set out to combine the two into a single device – one that is non-obtrusive, autonomous, and accessible.
The authors have chosen accelerometer to be the central component of their system – with very intuitive reasoning in terms of fall detection – and under the principle that the brain activity during sleep is equal to the motion produced by the body during sleep. The device’s requirements for both of these cases is simple – it needs to be attached to the body. An internet-connected smart watch was used for the prototype of this design, which also keeps the cost of the device relatively low – not having to develop a custom device from scratch.
If you wish to take a closer look at the back-end architecture of the system, the method of testing, and the results, you can get the full paper on ResearchGate.
This year’s edition of the conference, HealthyIoT 2016, is accepting submission until July 10th!