DomAssist technology

0 Comment

The DomAssist platform provides software orchestration of communicating objects placed in the person’s home. These hardware and software objects include:

wireless sensor-actuators available in large areas at low cost (motion detector, contact detector, electrical outlet controller, etc.);

software services (shared agenda, internet resources);

two digital touch tablets.

The first tablet, stationary, is placed at a central location of the house, to serve as a dashboard where the user will receive support services if needed detected by the system, or a digital photo sharing frame with the close entourage. The second tablet, mobile, serves the social and leisure activities of the person.

The analysis of the needs for independent home life collected from 525 elderly people living at home as well as the study of needs expressed in terms of assistive technologies desired by 100 elderly and caregivers, conducted during the first stage of the project , helped to define the scope to be covered by the DomAssist technology audience. DomAssist offers support services in three main areas:

the daily activities with monitoring the implementation of activities (eating meals, washing, dressing, up-down, etc.), a reminder of appointments and a personalized assessment of activities during the day;

the safety of the person and his home with, for example, the provision of a light path, a cook’s watch, and an alert to a carer in case of unusual or worrying situation;

the social participation through communication activities and entertainment, offering a simplified service emails in a video conference call application and customized entertainment applications based on user preferences (kitchen, multimedia library, solitary games or collaborative, etc.).

schéma3Support applications are all designed with a user-centric approach and take into account each person’s specificities, abilities, needs and expectations.

For example, some of the apps are based on the activities that the person performs throughout the day at home. In the context of extended home life assistance, it is important to be able to monitor these activities because they are skills that reflect the functional status of an individual. Verifying that these activities are actually being carried out is, in particular, a decisive factor in determining the type and level of assistance the person needs.

In this perspective, we have developed and experimentally validated an original approach integrating acquired knowledge about the daily functioning of the elderly person and ubiquitous technologies based on sensors aimed at the recognition of activities. Precisely, our approach is based on the fact that with senescence, the more a person suffers losses associated with aging, the more routine activities become routine in order to optimize daily functioning. Therefore, its activities do not need to be recognized accurately or inferred (as in conventional “data-driven” signal detection approaches), but can be approximated in terms of compliance with the self-declared routine (so-called “knowledge-driven approaches”). We were able to demonstrate that our automatic activity verification approach was as reliable as that performed by an ergonomist.

The figure above illustrates the detection of a participant’s daily activities over a five-day period. On this graph, we can see that the detection of breakfast and toilet activities is optimal, but the dressing activity is detected less accurately. This effect can be explained by the number of sensors used. Indeed, the recognition of breakfast and toilet activities is dependent on two or more sensors, while the dressing activity is linked to a single contact sensor placed on the dressing room door or door. the wardrobe.

The activity deviations can then be interpreted as functional deterioration warning signals of the person, which can be notified to the assisted person or sent to the caregiver.