Human Centric IoT Lighting Control based on Personalized Biological Clock Estimations

Smart Buildings with connected lighting and sensors have become one of the first large-scale applications of the Internet of Things. However, existing efforts to make buildings smarter focus mainly on energy conservation and cutting costs. In this paper, we further address the beneficial effects of light upon humans. People nowadays spend more than 90% indoors and as such the indoor environment becomes paramount for people's health and wellbeing. We present a Human Centric Lighting solution that supports human health through the estimation of the biological effects of light. It exploits the possibility that IoT offers to monitor humans and their experience and to control internet-connected lights. Despite the existence of well-established and extensively tested models of the circadian mechanism, benefits of those models are not yet harvested in practical applications. We envision an application that controls the lighting system based on a suitable model that predicts the human response to light. Our work confirms that the statistical signal processing approach of a Particle Filter can account for sensor and model uncertainties. We show how the response of the biological clock to light depends largely on individual characteristics, such as the intrinsic characteristic duration of the day, and therefore we include such person characterization into our system design.