Pressure sensor array infused mattress with neural network posture classification : a thesis presented in partial fulfilment of the requirements for the degree of Master of Engineering in Mechatronics at Massey University, Albany, New Zealand. EMBARGOED until 30 June 2024.

Thumbnail Image
Open Access Location
Journal Title
Journal ISSN
Volume Title
Massey University
The Author
Sleep is an essential physiological process and is vital for human health and well-being. There is an undesirable trend of poor sleep hygiene, especially insufficient sleep. Sleep monitors can help bring to light this essential nighttime activity. The Comfort Group® (TCG), Australasia’s largest mattress manufacturer seeks to embed a sleep posture monitor into a mattress as this is a platform used in almost every home. A full body posture monitor can provide richer information about the body than most movement-based sleep monitors on the market. There is particular interest in embedding a Pressure Sensor Array into a mattress because of the amount of sleep information that can be extracted from a pressure image. The renewed interest in this already-proven technology comes with the growth of computing power and machine learning algorithms. Today, features such as body shape, limb position, joint angles and even breathing rate can be inferred from the sleeper. However, they are not cost-effective for the consumer market. Here, we demonstrate a cost-effective pressure sensor array that is embedded into a consumer mattress. Basic posture detection of left-lateral, right-lateral and supine is shown using an artificial neural network. An accuracy rate of 99.1% is achieved. Having a cost-effective mattress-infused platform for the consumer will increase sleep hygiene in society and open the doors to a larger dataset for further analyse. For the scientific community, this larger dataset has the potential to produce a higher fidelity of insights into societal sleep.
Embargoed until 30 June 2024