Contextual occupancy detection for smart office by pattern recognition of electricity consumption data

The advent of IoT has resulted in a trend towards more innovative and automated applications. In this regard, occupancy detection plays an important role in many smart building applications such as controlling heating, ventilation and air conditioning (HVAC) systems, monitoring systems and managing lighting systems. Most of the current techniques for detecting occupancy require multiple sensors fusion for attaining acceptable performance. These techniques come with an increased cost and incur extra expenses of installation and maintenance as well. All of these methods are intended to deal with only two states; when a user is present or absent and control the system accordingly. In this paper, we have proposed a non-intrusive approach to detect an occupancy state in a smart office using electricity consumption data and introduced a novel concept of third state as standby for dealing with situations when the user lefts his seat for small breaks. We demonstrated our approach using electricity data collected within our research centre and detected occupancy state with efficiency up to 94%. Furthermore, our solution does not require extra equipment or sensors to deploy for occupancy detection as smart energy meters are already being deployed in most of the smart buildings.