Derivation of Night Time Behaviour Metrics using Ambient Sensors

Abstract

Sleep problems have been shown to have significant negative impact on health. As such it is important to examine night time behaviour to objectively determine when sleep disturbances arise. Due to the large night-tonight variability in sleep quality for older adults, it is important to objectively measure behaviour over a significant period to establish trends or changes in patterns of sleep. In this paper we present a means of ambiently monitoring sleep through the use of sensors installed in each of sixteen independent living apartments. We investigate the effect of time outside the home and movement within the home on sleep. These measures are validated against comparative measures from two actigraph datasets. The first consisting of five adults, two of whom are healthy subjects and the other three adults have previously fallen, gathered over a period of between two and four nights. The second consisting of three older adults recorded over seven nights in their own homes. Results relating time outside the home and movement within the home to sleep are presented for three individuals spanning a period of between 630 and 650 days.

Inferring health metrics from ambient smart home data

“…This allows more complex algorithms which require data from multiple sources to run. For example, the determination of room-level location of the individual derived using an algorithm based on data from a mixture of PIR, door contacts and light switches as well as some deterministic logic [12].…”

“…2) Percentage of time spent in each room/location: A deterministic model was generated using data from the PIR, door and window contact, and light switch sensors to infer the location of the individual within the apartment [12]. The locations (l) of interest are the kitchen/living room, main bedroom, en suite, water closet, second bedroom, hall and outside the house as defined by l= l(k) where l is the location of the individual at sample k.…”

Older Adults’ Attitudes to Self-Management of Health and Wellness through Smart Home Data

“…We have gathered large amounts of data from sensors in the 16 apartments, and significant validation and analysis of data has taken place [5]. Algorithms have been developed to model patterns of daily behaviour and wellbeing [5], [11], [12]. A full description of the system architecture to gather and analyse data from GNH, as well as some of the applications providing feedback to residents is presented in [5].…”

“…The data focused on 3 areas of wellbeing, sleep, activity and physiological monitoring, which have been shown to be important in maintaining healthy ageing [11], [18], [24][25][26]. …”

Study on the Effect of the Ambient Temperature toward the Quality of Sleep

“…The other method that is available is the use of video and audio recording together with Passive Infra-Red (PIR) sensor to detect the user sleep status [9]. The study in [10] presented a research on ambient sleep monitoring which using sensors that are installed In this study, our focus is to illustrate the relationship between temperature and sleep quality and for that purpose, the use of PC instead of smartphone is preferable because while we are in sleep, we did not require too much of mobility ergo PC with higher processing capability is more suitable than smartphone that allowing user to do a lot of processing and video recording. The developed sleep monitoring system is a combination of sensors and BLUNO microcontroller board to monitor and record the ambient parameters and body condition while user is sleeping.…”