IoT systems for unobtrusive monitoring of physical activity and vital signs

This thesis presents the realization of two Internet of Things (IoT) systems designed for unobtrusive and continuous monitoring of activity and vital signs. The first one is a ballistocardiogram (BCG) and seismocardiogram (SCG) analysis system, which aims to help users and their caregivers to identify anomalous trends in the long term. An acquisition device and an analysis algorithm are the system’s main elements. The acquisition device senses the signals through an accelerometer. The analysis algorithm can identify heartbeats with high sensitivity and precision and provides low Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) for heartbeats annotation. Experimental measures on various databases assessed such abilities. The second system is an eight square meters smart floor that offers interactive games to the users to stimulate an active and healthy lifestyle, designed for the project PLEINAIR. The device is created by joining multiple smart tiles, which have a simple mechanical structure, are controlled by a dedicated electronic circuit, and can provide interaction through sensors and luminous feedback. A web application allows the users to play with the device and shows the user’s scores. Experimental measures assessed the system’s ability to sense weight little as 300 grams. Both systems well fit inside the AAL framework that searches for solutions for healthy ageing.