Indoor occupancy detection and visualization system

Abstract

The penetration of smartphones and the prevalence of Wireless Local Area Network have facilitated evolutionary applications in Indoor Positioning Systems. This project pertains to an indoor occupancy detection and visualization system that leverages the existing deployment of Wireless Local Area Network, to predict the distribution of indoor occupants through clustering algorithm and display both real-time and historical distribution on a web-based platform. Compared to other Location Based Services estimating individual location, this system transforms the large data set of received signal strength of mobile devices into refined and practical information about the distribution of indoor occupants, which can be further exploited for a diverse range of applications, such as human resource management systems and energy conservation systems in residential and business buildings. Heating, Ventilating, And Air Conditioning control is the primally targeted application of this project. The following research and development are discussed in this report: the system built up to collect the Wi-Fi received signal strengths of mobile devices of indoor occupants from multiple Wi-Fi access points, estimate the locations of the mobile devices by fingerprinting-based approach and detect the occupancy through cluster analysis; the website designed in line with Software Development Life Cycle to visualize the real-time and historical distribution of indoor occupants by using PHP, JavaScript, HTML5 and MySQL. The system exhibits high accuracy in repeated experiments and the website demonstrates its efficiency and preciseness in analytically presenting the results.