Effect of the Information Support Robot on the Daily Activity of Older People Living Alone in Actual Living Environment

Abstract

Information support robots (ISRs) have the potential to assist older people living alone to have an independent life. However, the effects of ISRs on the daily activity, especially the sleep patterns, of older people have not been clarified; moreover, it is unclear whether the effects of ISRs depend on the levels of cognitive function. To investigate these effects, we introduced an ISR into the actual living environment and then quantified induced changes according to the levels of cognitive function. Older people who maintained their cognitive function demonstrated the following behavioral changes after using the ISR: faster wake-up times, reduced sleep duration, and increased amount of activity in the daytime (p < 0.05, r = 0.77; p < 0.05, r = 0.89, and p < 0.1, r = 0.70, respectively). The results suggest that the ISR is beneficial in supporting the independence of older people living alone since living alone is associated with disturbed sleep patterns and low physical activity. The impact of the ISR on daily activity was more remarkable in the subjects with high cognitive function than in those with low cognitive function. These findings suggest that cognitive function is useful information in the ISR adaptation process. The present study has more solid external validity than that of a controlled environment study since it was done in a personal residential space.

The effects of a sleep robot intervention on sleep, depression and anxiety in adults with insomnia – A randomized waitlist-controlled trial

“…A second strength is the prospectively published study protocol (Støre et al, 2020), including a statistical analysis plan, which was followed, to reduce the risk of bias even further. Thirdly, the use of standardized assessment is a strength, as this is a common limitation in other robot studies (Mizuno et al, 2021;Støre et al, 2022). A fourth strength is the fact that we studied a robot that has been created to target sleep, as opposed to previous studies conducted on the effects of robots on sleep, where the companion robot Paro has been used (Mizuno et al, 2021;Støre et al, 2022).…”

“…Thirdly, the use of standardized assessment is a strength, as this is a common limitation in other robot studies (Mizuno et al, 2021;Støre et al, 2022). A fourth strength is the fact that we studied a robot that has been created to target sleep, as opposed to previous studies conducted on the effects of robots on sleep, where the companion robot Paro has been used (Mizuno et al, 2021;Støre et al, 2022). The relatively long intervention phase is yet another strength, although it is possible that a 3-week intervention was too brief to induce an effect.…”

“…The relatively long intervention phase is yet another strength, although it is possible that a 3-week intervention was too brief to induce an effect. Previous studies on the effects of robots on sleep have been conducted in care facilities or in lab settings (Mizuno et al, 2021), which is why a major strength of our study is that it is conducted in real-life settings, that is, it has high ecological validity.…”

Activity discovery using Dirichlet multinomial mixture models from discrete sensor data in smart homes

“…The Kasteren dataset is a benchmark that has been cited in previous studies. However, because reed switch sensors requires significant time and effort to install, PIR motion sensors, which are battery-powered and easy to install, are often used to monitor and analyze the daily activities of the elderly [6,7]. For this study, we used the Aruba testbed data from the WSU CASAS dataset [20].…”

“…Because the cost of an annotation is generally high, preparing annotated data for each resident is cost-prohibitive, and the cost is warranted only if a model trained for a particular resident is applicable to other residents. However, sharing activity models between residents is difficult, particularly when only unobtrusive binary sensors, e.g., passive infrared ray (PIR) motion sensors, are allowed [6,7]. More informative sensors such as video, audio, or wearable sensors are unacceptable in monitoring activities conducted over extended periods owing to intrusiveness, and potential to interfere with daily life [2,5].…”