Activity prediction for in-home activity monitoring

In this paper, we propose an extension to a generic model developed for elaborating person´s lifestyle patterns and changes in those patterns which could be related to health decline. The baseline model aims at a better understanding of the behaviours engaged in various housing contexts such as conventional housing or Assisted Living (AL) facilities and exploring irregularities within the life rhythm of the resident´s everyday life activities. The model, embedded into a customized software program, has been extended to predict the progression of a particular activity on the basis of a circadian period (24- hour period) in order to detect unexpected events such as slowdowns or disruptions which could indicate changes in health status, cognitive or motor impairments. The objective is to help caregivers provide better assessment of functional health status and preventative assistance. A case study, elaborated through software simulations, is presented to show the potential of the software application to determine life patterns in conjunction with activity trend indicators that may indicate health deterioration.