Full-coverage unobtrusive health monitoring of elders at homes

Abstract

Due to the fast increase of the elderly population, in-home eldercare has emerged as an important topic of research. Existing studies in intelligent health monitoring have yielded promising outcomes. Nevertheless, these approaches fall short of achieving round-the-clock, unobtrusive surveillance. Herein, we introduce a system offering full-dimensional and unobtrusive in-home health monitoring for elderly individuals living alone. We deploy smart sensors, including infrared, bed, sound, and physiological sensors, to provide all-around coverage of the elderly living space. The proposed system proficiently identifies activities such as cooking, coughing, snoring, talking, listening to music, and walking, while concurrently estimating the indoor location and providing intelligent sleep monitoring and assessment of physiological parameters. Our proposed algorithm has undergone evaluation using real-world data, encompassing a 30-day activity record of ten subjects aged 65 and older. The results demonstrate the remarkable effectiveness of our method, showcasing a high degree of accuracy in recognizing a diverse range of activities. This substantiates the practical implementation of our system in authentic homecare environments.