Remote Health Monitoring System for the Elderly

Abstract

Population ageing is increasing the need for home-based care while the available care workforce remains limited, creating a need for unobtrusive systems that support routine observation without adding tasks for elderly residents. This thesis presents the design, implementation, and validation of a concept demonstrator for a remote health monitoring system built around a smart mirror and modular sensor printed circuit board. The system measures presence, heart rate, respiratory rate, environmental conditions, and ambient light using radar and supporting sensors. A resident-facing mirror display presents the most relevant information with large visual elements, limited text, and a traffic-light interpretation scheme, while requiring no active user input. Sensor readings are published over an encrypted, certificate-based local communication link to a broker hosted on the tablet behind the mirror. A gateway subscribes to these readings, buffers them during outages, converts vital-sign data into standard healthcare observation resources, and forwards them to a reference healthcare server. Validation showed that the access-control design confined devices to their own data paths and rejected unauthenticated connections, while the custom circuit board and sensor interfaces were brought up successfully. The vital-sign radar followed expected physiological trends, although heart-rate readings showed a systematic overestimation and breathing-rate readings were closer to the reference. The project demonstrates a secure path from ambient home sensing to a healthcare information interface, but further work is needed on clinical-grade validation, production endpoint authentication, deployment hardening, and resident identification.