Validation

Testing in AAL home laboratories 

As a first evaluation step – before moving to actual end users’ homes, we will evaluate the acceptability, usability and learnability of the proposed system and services, and the multi-modal human-robot interaction with end users and their family in two AAL test homes, provided by project partners SMH in the Netherlands and FDG in Italy. These ‘AAL Home Labs’ are actual real homes where people will be living for a couple of hours up to a couple of days, living their normal life but now experiencing the ENRICHME services.

Once the systems are fully functional, sufficiently reliable, and have undergone technical tests by our researchers, it will be tested with the involved elderly. The purpose of these tests is to find overall usability and user-experience flaws, and to improve the actual ENRICHME services, before going to the long-term user studies in peoples own homes.

In Italy, users will be recruited within the different hospitals and daily centres of FDG: in particular the FDG residential care facility for elderly people “RSA Beato Palazzolo” of Milan will be exploited for the recruitment. In the Netherlands, SMH will recruit users from their large existing pool of test users, from their partner home care organisations, and from existing contacts in Alzheimer patient organisations.

The smart home of FDG is an apartment of approx. 130 square meters integrated in the Occupational Therapy service of one of the FDG hospitals. The smart home of SMH is a 160-square meters standalone home, fully equipped with smart home technology, but still resembling a normal home. Both smart homes are equipped with a KNX™-based home automation system including a set of sensors (TV cameras, temperature, water flow meters, motion detection, brightness sensors, etc.) and actuators (doors, windows, curtains, blinds, light, etc.) that will be connected to the robotic system.

Users and their caregivers, family members (at least one for each user) and multiple social reference group members will be asked to perform the identified scenarios and provide their feedback on the experience. The test sessions of each user will last approximately 2 hours plus the time needed for collecting user feedback. In the Netherlands, an extra group of end users will stay in the smart home for a couple of days, as an intermediary step between the short-term user tests and the actual long-term pilots in people’s own homes.

This testing phase will provide further input and feedback for the improvement of the system that will be carried on in WP7.

 

Long term validation in Elderly Housing facilities

As consequence of system improvement, 6 ENRICHME systems (deliverable D7.1) will be delivered for the 7 months testing in elderly house facilities. Each testing site will be equipped with 2 systems.  An additional robot will be produced and kept at PAL premises for continuous integration, validation and remote support tasks.
Every one of the two systems will be assigned for 3,5 months to a certain elderly person who lives independently within the elderly facilities: LACE in Lincolnshire UK, AKTIOS in Agia Paraskevi, Greece or at their home supervised by a social worker (PUMS in Poznan Poland).  Therefore, there will be 2 monitoring runs for a total of 4 monitored users in each site.  However, as 3,5 months is still a short time for functional alterations to occur in subjects with mild cognitive impairment each run in PUMS will be prolonged for 2 extra months of regular care (1 month before and after ENRICHME system exposure). Longer introductory phase will be organised in PUMS, due to higher risk of drop-out.  Moreover, a control arm is added, expected to help to find the differences between participants receiving regular care and ENRICHME-based care. Each participant of the active arm would be matched with one of control arm.

ENRICHME system will to monitor the elderly with MCI permanently (24/7).The system will check both the mental as well as physical status of the person. This will allow for noticing possible behaviour alterations and changes in daily activities, the intensification of which leads to increased dependency on the help of others. Medical personnel will be enable to follow users status through the platform for networked care. The monitoring should also allow observing a possible intensification of problems which with high probability indicates a progression of MCI. At present these persons are supervised occasionally by a social worker who can contact them via intercom. A deployment of the system should minimise the workload of the worker but also improve the quality of monitoring due to its continuity. It should also improve the sense of security of the elderly and thus their quality of life.

The main aim of the system will be assistance in daily activities, e.g. reminding of meals and medications. Any suggested action will have to be confirmed on the display of the system. Another important system’s task will be an activation of the elderly person – physical (exercises according to programmed guidelines) and mental (exercises for cognitive functions, e.g. music therapy or photo art therapy). It has been proven that such activations delays cognitive changes. Also learning new activities (learning how to cope with the robot) has similar effect.

In the case of persons with hearing impairment the system, reminding of activities, will come close and signal the necessity of an activity (e.g. taking drugs) on its display. If failing to obey, the robot will accompany the elderly person, in order to turn their attention, until the activity gets confirmed on the display. For semi-blind the robot will generate audible messages.

At least one family member, one care giver and multiple social reference group members for each patient will be asked to provide their feedback on the system.