Non-Visual and Contactless Wellness Monitoring for Long Term Care Facilities Using mm-Wave Radar Sensors

We propose a radar-based system for wellness monitoring for long-term care (LTC) facilities. Three standalone systems are used to monitor a resident in the washroom, living room and bed. A novel resident detection algorithm is proposed to find the occupied room. Based on the outcome of the algorithm, the resident's washroom frequency, washroom usage time, and location can be recorded. For the resident in the living room area, gait analysis, activity recognition, and vital sign monitoring are performed using sequential deep learning models. Additionally, the sequential deep learning model identifies fall incidents and fall recovery. In the case of a non-recovered fall, an alert is sent to a caregiver or supervisor. The experimental results obtained from a local LTC are highly accurate, demonstrating the effectiveness of radar-based sensors for LTC facilities.