Ambient activity monitoring for medical applications in multi-person households

This paper outlines doctoral research towards automated activity monitoring in the home, including past results and future plans. The aim of this research is to understand the possibilities and limitations of automated, or "unsupervised", approaches to ambient activity monitoring. Numerous applications of ambient activity monitoring, from care assessments to rehabilitation monitoring, have been developed in the past, but most approaches require complex calibration routines or are limited to single-person households. Furthermore, the requirements and theoretical limits of these approaches, such as sensor density and inhabitant-sensor-ratio, have not been studied.