Artículo: Dy-MIoT-Health: A Dynamic Mobile Internet of Thing Health Platform Based on Discovery Services and Edge Computing
We have previously developed an IoT platform for monitoring elderly people. Now, we propose extending it with dynamicity through two new functionalities: 1) A dynamic discovery service that discovers nearby sensors and selects them according to the specific application, context information, and data provenance; and 2) A functionality that dynamically decides whether to deploy a task/service at the edge or the cloud according to context information (e.g., computing capacity, device storage, gateway, battery shortage, redundancy in sensors used, and data collected). To validate our proposal, we propose a use case to detect stress and anxiety during activities of daily living (ADLs) in elderly people, analyzing differences between people’s age and gender, chronic stress, and coping styles (emotional reactions to crises), among others. We will collect data from physiological sensors, cortisol tests, and questionnaires. Then we will analyze these data with machine learning techniques. The analysis results will help psychologists and health professionals to personalize the detection of outbreaks and interventions based on our results (differences between ages, gender, etc.).