| Sumario: | Purpose In recent years, an extensive bulk of technology was developed for helping PwD and its caregivers. Low-cost sensors, wearable devices, electronic health records and artificial intelligence (AI) spread in the healthcare system to gather and analyze data collected from PwD to get insights into his/her physical condition, behavior at home and lifestyle (Bevilacqua et al., 2020). Assistive technologies can improve the quality of life of seniors and support caregivers to detect and monitor the onset and progression of diseases. Future developments should apply a holistic approach that analyses multiple aspects of the quality of life of PwD using devices such as GPS trackers, tablet/mobile applications, lifestyle monitoring systems, and sleep monitoring sensors (Wójcik et al., 2021). Currently, there are many devices that work alone and focus on one or two needs of the PwD (Casaccia et al., 2019; Aloulou et al., 2013).This is the premise of the European HAAL project, (AAL-2020-7-229-CP), which aims at integrating a set of devices developed in past projects, into a system able to target and evolve with PwD through the whole course of the disease. A milestone of HAAL is that the combination of two or more devices, specifically chosen to monitor the most important aspects of the PwD's life, will provide added value to the platform and moreover, the combination of multiple heterogeneous collected data from the devices can give detailed insights into PwD's life. This work describes the preliminary sensor network integrated into the HAAL platform, to support the different stages of dementia. Method The software and hardware architecture was developed by integrating the following devices: a smart mattress (Whizpad), a social tablet (Compaan), a lifestyle monitoring system (Sensara) and a GPS sensor (Kompy Pico). Each device collects different data that are sent to a database developed in Amazon Web Service (AWS). Data from each sensor are retrieved by using dedicated HTTP requests that communicate with the API of each device. Data collected are stored in the database and are accessible for filtering and analysis. The resulting heterogeneous dataset is processed with the aim of choosing the most suitable combination of data that can be given as input to AI algorithms, to measure one or more aspects of PwD well-being. Result and Discussion The results of the proposed platform show that the sensor network chosen has multiple strengths: it is adaptable to different living environments, regardless of the layout of the user's home. The sensor network is minimally invasive, as the ADL monitoring system is installed at home and the GPS is a wearable device and can support the caregiver during the overall day, both in indoor and outdoor activities. The result is a heterogeneous dataset that together with the input provided by the users will be adopted in combination with AI algorithms to predict users' well-being and evaluate the progression of the dementia disease.
|
|---|