Handwashing and Showering Recognition Utilizing Wearable Humidity Sensor

With the rising age of the population in the United States, there is a great demand for elderly care at long-term care facilities (LTCF). To address this shortage in care, CarePredict has introduced sensor-laden wearables that utilize machine learning to analyze residents’ behavior and activity within LTCFs. The wearable detects changes in activity and behavior to predict the risk of a fall, depression, or urinary tract infection (UTI), allowing caregivers to give proactive interventions and improve patient health. This study aims to improve the augmented activity monitoring of users by more precisely classifying activities in the bathroom. We demonstrated a proof-of-concept study to differentiate showering versus hand washing behavior by utilizing humidity data from the CarePredict wearable tempo with a machine learning and signal processing algorithm. The ability to precisely classify such events plays a key role in early detection of behavioral changes which enable predictive medicine, especially with regard to UTIs.