SI4SI aims at creating a smart living environment based on validated innovative technology ( i.e. sensors like https://vayyar.com/home or https://www.portabiles-hct.de/en/) feeding an AI-based system to System to identify different levels of isolation symptoms.

Thanks to an AI Advisor, SI4SI improve decision making support while suggesting personalised activities to both Seniors and Caregivers. Furthermore, the AI Advisor feeds a Collaborative Application for Caregivers and a Smart Tablet (able to support both Video and Vocal interaction) provided by ERT which allows the caregivers and health practitioners can also interact with the seniors, sending input to the Tablet.

SI4SI has the challenge to:
- identify the social isolation of the elderly trough a less invasive and innovative Smart Living Environment, detecting both Physical, Social and effective behaviours.
- intervene promptly through AI Based Advisor.
- improve the levels of home care by Professional Caregivers
- foster emotional relationships between family members and the elderly
- promote the social involvement of the elderly.

The SI4SI Smart Living Environment is composed of:
- A set of non-invasive sensing devices (IoT sensors, wearables) to track Functional activities and SocialActivities, detecting Emotional Patterns and Behaviors that can be associated to Social Isolation and Social Exclusion.
-a Multimodal Fusion Analysis System will exploit data collected in order to
understand the evolution of the senior’s kinetic, emotional status, and his/her level of social participation.
- A Conversational Assistant
- an AI Advisor to identify recommendations and help the user (senior, caregiver, social carer, professional), by analysing her/his knowledge on her/his condition and the treatment or processes that need to be followed.
- A Social Platform
- A Collaborative Application for Professional, Social and Families Caregivers
- A Smart Home Tablet that will be the Vocally Interface of the Virtual Assistant and the Video Interfaces of the Social Platform

This work was supported by a grant of the Romanian Ministry of Research and Innovation, CCCDI - UEFISCDI and of the AAL Programme with co-funding from the European Union’s Horizon 2020 research and innovation programme.