Intelligent Human Presence Detection and Identification

In the current Internet of Things (IoT) era, user presence or absence detection inside home has become a predominant requirement for handling various scenarios, such as user preferences, Edge computing, IF This Then That (IFTTT) for a smart home environment in home assistance devices such as Amazon Alexa, Google home, Samsung Galaxy home etc. Currently such home assistance devices are lacking the intelligence to detect the human presence or absence detection along with exact user identification inside the home. In this work, we propose a novel unsupervised sophisticated classification idea to deal for the situations, where multiple users can query simultaneously on a single platform for presence detection using various connectivity technologies. The multiple technologies such as Bluetooth, Bluetooth LE, Wi-Fi, Speaker ID, Sonar, Voice, ZigBee, Z-Wave etc., will be queueing inside the system and will be aided sequentially or in parallel fashion to find the presence detection of a particular user. In this paper, we encompass three technologies presence detection, 1) using Bluetooth, 2) using Wi-Fi and 3) using Speaker ID (voice). Along with the presence detection using various methods, here we are focusing equally on user identification and association rules-based inferences generation. Also discussed various use cases and the performance of the system, which is optimized for the parallelizing the methodology for different presence detection methods. The experimentation results of our proposed algorithm's presence detection shows overall accuracy of 99.2% using Bluetooth, 98.6% using Wi-Fi and 92.7% using Speaker ID.