Monitoring and alerting system for home-alone elderly

• Main Author: Kork, Jing Yi
• Other Authors: Yvonne Lam Ying Hung
• Format: Final Year Project
• Language: English
• Published: 2015
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Applications of electronics
• Online Access: http://hdl.handle.net/10356/64660
• Institution: Nanyang Technological University

There are several kinds of sleeping disorder with severity ranging from mild to life threatening. Sleep Apnea is a form of sleep disorder where an individual experiences pauses in breathing throughout his sleep. These pauses typically last for 10 seconds or more. Despite efforts to breathe, it often leaves the individual gasping for air. Often, the individual is unaware of his inability to breathe during that period. While usually the individual is able to self-recover from an apneic attack and regain his breathing ability, there are times where he does not regain his ability to breathe, hence endangering his life. Therefore, medical professionals recommend tracking sleep disorder individuals actively and monitoring their condition so that treatment can be applied promptly to reduce the risk of fatality. Traditionally, patients who may be facing any sleep disorder have to go through polysomnography in a sleep laboratory where many sensors are attached to the individual. This is time and labour consuming. Therefore, in the recent years, researchers are looking into snore-based analyses that are non-invasive, require less manual support and are less costly. This project explored the options of monitoring the sleep pattern, specifically through the detection of snores and identification of the various sleep states at different period in time. The initial stage of this project focused on detecting the sleep states through monitoring movements and brain waves. In this stage, a vibration sensor installed on a Raspberry Pi was successfully developed to detect movements. However, the results collected from the brain waves using the Myndplay brain band were insufficient to conclusively identify the different sleep states. In the second stage of the project, a snore detector was added on a Raspberry Pi and it was able to accurately differentiate between a snoring sound and the surrounding sound such as human voice. The snore sound can also be viewed remotely on a web server. Lastly, in order to aid in providing treatment promptly, this project also creates an alert signal, in the form of an attached LED light, when it detects an abnormal pattern. In conclusion, other alerting the user when required, the prototype developed in this project introduces a more cost-efficient and convenient way of monitoring at home thereby aiding in the early identification of sleep disorder, such as sleep apnea.