Showing 1–1 of 1 results for author: Rai, H V
-
Automatically detecting activities of daily living from in-home sensors as indicators of routine behaviour in an older population
Authors:
Claire M. Timon,
Pamela Hussey,
Hyowon Lee,
Catriona Murphy,
Harsh Vardan Rai,
and Alan F. Smeaton
Abstract:
Objective: The NEX project has developed an integrated Internet of Things (IoT) system coupled with data analytics to offer unobtrusive health and wellness monitoring supporting older adults living independently at home. Monitoring {currently} involves visualising a set of automatically detected activities of daily living (ADLs) for each participant. The detection of ADLs is achieved {} to allow t…
▽ More
Objective: The NEX project has developed an integrated Internet of Things (IoT) system coupled with data analytics to offer unobtrusive health and wellness monitoring supporting older adults living independently at home. Monitoring {currently} involves visualising a set of automatically detected activities of daily living (ADLs) for each participant. The detection of ADLs is achieved {} to allow the incorporation of additional participants whose ADLs are detected without re-training the system.
Methods: Following an extensive User Needs and Requirements study involving 426 participants, a pilot trial and a friendly trial of the deployment, an Action Research Cycle (ARC) trial was completed. This involved 23 participants over a 10-week period each with c.20 IoT sensors in their homes. During the ARC trial, participants each took part in two data-informed briefings which presented visualisations of their own in-home activities. The briefings also gathered training data on the accuracy of detected activities. Association rule mining was then used on the combination of data from sensors and participant feedback to improve the automatic detection of ADLs.
Results: Association rule mining was used to detect a range of ADLs for each participant independently of others and was then used to detect ADLs across participants using a single set of rules {for each ADL}. This allows additional participants to be added without the necessity of them providing training data.
Conclusions: Additional participants can be added to the NEX system without the necessity to re-train the system for automatic detection of the set of their activities of daily living.
△ Less
Submitted 10 July, 2023;
originally announced July 2023.
