Authors:
Nicolas Bioul
1
;
2
;
3
;
4
;
Arthur Pisvin
1
;
2
;
3
;
4
;
Maxim Lamirande
5
;
Jérôme Leclère
1
;
2
;
3
;
4
;
Lucas El Raghibi
1
;
2
;
3
;
Adrien Denis
1
;
2
;
3
and
Benoît Macq
1
;
2
;
3
;
4
Affiliations:
1
OpenHub Place du Levant 3, Louvain-la-Neuve, Belgium
;
2
Pixels and Interactions Lab, UCLouvain, Place du Levant 3, Louvain-la-Neuve, Belgium
;
3
ICTEAM, UCLouvain, Place du Levant 3, Louvain-la-Neuve, Belgium
;
4
TRAIL Institute, Belgium
;
5
Inter'Act, ULiège, Allée de la découverte 9, BuildingB52/3, Liège, Belgium
Keyword(s):
AI, Digital Twin, Silver Economy, Innovative Testing Methods, Fall Detection, Assistive Technology, Geriatrics.
Abstract:
Addressing the issues of age and disability, our study presents a systematic technique for evaluating smart home technology designed to improve independent living. While acknowledging companies’ efforts in this field, we created a framework to assess potential solutions using a rigorous demographic study that defined various user profiles - or personae - as the foundation for our comparison research. Our methodology is based on a dual-focused analytical approach: analysing installation processes and operating performance, with a particular emphasis on fall detection and behaviour analysis. To evaluate fall detection, we developed a test protocol, which resulted in the compilation of a large database. We pioneered the use of virtual personae in a game engine for behavioural analysis, which are simulated in living contexts via probabilistic activity generation. This novel approach allowed the creation of virtual sensor data, which was then analysed by AI algorithms thus generating aler
ts. This study emphasises the possibility for combining IoT and AI to reduce the need for institutional care by offering real-time help and monitoring. Our methodology takes a thorough approach to assessing the efficacy of smart home devices, ensuring that they are adaptable to the real-world demands of the ageing population and people with disabilities.
(More)