Crowdsensing at Lee Wee Nam Library : a smartphone approach

Abstract

This project is a development of an extension to the popular U-Wave app which tracks the number of people at the NTU Lee Wee Nam Library. It uses common sensors in modern smartphones to implement an opportunistic Mobile Crowdsensing (MCS) solution that would benefit both users of the LWN Library and to provide utilisation insights to library staff. To create a holistic picture of the real-time crowd at the LWN Library, the app tracks the user’s current state – their presence at the library, their activity state (stationary or moving), and their current indoor level position within the library. The app deduces this state by using Machine Learning-powered classifiers to detect the user’s location and activity from GPS and sensors data. The sensors used include the accelerometer, barometer and geomagnetic field sensor. This app module creates a novel way to gather crowd data from a multi-level indoor complex without the need to install additional hardware infrastructure.