Feasibility of unobstrusive ambient sensors for fall detections in home environment

Falls are the leading threats of death, injury and hospital admissions of seniors. About one third of seniors fall every year. Real time fall detection is thus critical to support independent livings of seniors by getting timely interventions from carers/family members. Recent years along with the IoT booming, many wearable sensor based fall detection systems have been developed. However privacy concerns or simply forgot-to-wear make wearable sensors not very accepted by seniors in a home environment seeking long term monitoring solutions. This motivates the development of unobtrusive ambient fall detection system. In this paper, we reviewed studies in this area and categorised them into two types of approaches, namely active and passive. We also evaluated their feasibilities within five domains: obtrusiveness, power connectivity, affordability, complexity of installation and being tested in field trials. The evaluation results could be used as guidance in designing new unobtrusive ambient fall detection systems.