Occupancy detection via thermal sensors for energy consumption reduction

With the emergence of Internet of Things (IoT), the usage of sensors for controlling and monitoring remote devices to achieve sustainability has gained researchers’ interest. It has been observed that buildings are one of the largest consumers of energy hence, effective measures are required to achieve sustainability. In this sector, substantial amount of energy is used by HVAC (Heating, Ventilation and Air Conditioning) systems to offer ease for occupants. In most cases, HVAC systems of these buildings run on fixed schedules and do not provide any satisfactory control, based on detailed occupancy information. In this paper, a new solution is presented for estimating occupancy using network of thermal sensor arrays. The system provides near real time actionable information for controlling HVAC system and conditioning the rooms based on usage. The proposed system is a network of wired sensors, wireless sensors and gateway nodes, working together. Energy readings estimate the battery life of over two years, while working accurately. The system shows potential energy savings of 10% to 15%. Recurrent neural network are also used to train the model and compared with the proposed method. We conclude this new approach remarkably improves the results of occupancy detection using network of thermal sensor arrays.