Detecting and assessing older adults' stressful interactions with the built environment : an elderly-centric and wearable sensing-based approach

Abstract

Globally, one in six people is expected to age 65 years by 2050. Not only is the global population ageing, but also the built environment infrastructure in many cities and communities are approaching their design life. This phenomenon is referred to as "double ageing". Ageing built environment infrastructure with defects are likely to result in environmental barriers with excessive demands; humans experience stress and/or their mobility is limited when the environmental demands exceed their functional capability. Given that human's functional capability declines with ageing, there is more likelihood for older adults to experience stressful environmental interactions that could limit their mobility than the average person. Current approaches to detect environmental barriers are inefficient, time-consuming, and costly, which may limit the frequency and scope of the built environment assessment. In order to promote active ageing in cities and communities, urban planners and municipal decision-makers need a more efficient approach to assess and detect excessively demanding environmental conditions. The aim of this research is to promote older adults' mobility by reducing environmental demands. The overall goal of this research is in two folds: (1) to enable practitioners to detect stressful older adults-environment interactions in near real-time and (2) to bring to the limelight the influence of urban environment configurations on older adults' stress response. To achieve this goal, this research harnessed the current advances in wearable sensing technologies to collect older adults' bodily responses (i.e., physiological, behavioural, and cognitive responses) to their interaction with the environment as a means of assessing and detecting environmental barriers. Specifically, a methodological framework was developed for researchers and practitioners to determine the relevance and informativeness of people's bodily responses in the context of their study. Based on this framework, it was identified that older adults' physiological response is more informative than the cognitive and behavioural responses. The informativeness of the cognitive response was affected by the walking activity, and the gait abnormality among older adults affected their behavioural responses. A statistical, spatial and space-time pattern mining was conducted to understand the relationships in older adults' physiological responses to the built environment. The results demonstrate that the relationships between older adults' physiological response and the environmental condition are less apparent at the individual level. However, using collective sensing (i.e., aggregating multiple participants' physiological responses) can accommodate the individual variability and capture any normality in the data, which is indicative of an environment's condition. An optimised environmental stress hot spot detection framework was developed using an Ensemble bagged tree algorithm that achieved 98% accuracy. A simulation-based approach was used to examine areas within the study area that are sufficiently powered to detect stress hot spots that pose high risk to older adults. The results demonstrate that urban planners and municipal decision-makers can use this approach to detect and alleviate stressful environmental conditions more efficiently; as a result, improving older adult's mobility in the built environment. An integrated methodology based on machine learning and an evolutionary rule-based system was developed to further understand the influence of visuospatial configurations (specifically, isovist indicators) of urban space on older adults' physiological stress. The results demonstrate that isovist minimum visibility, occlusivity and isovist area are the most influential determinants of older adults' physiological stress and non-stress response. Older adults prefer urban configurations where they can be seen. The generated visuospatial configurations can be used to inform urban design and planning.