Human centric sensing by Android phones

Abstract

The main objective of this project is to perform human centric sensing on Android smartphones. Nowadays, Android smartphones provide several built-in sensors that can monitor the user’s location as well as motion. This project aims to create an Android application which makes use of these sensors to track users’ location. This application will start queue detection when user is located in a queue potential area such as canteen or supermarket. A location detector service is implemented for controlling start and end of queue detector. This service uses two layers of detection; location sensing and Wi-Fi sensing to monitor the user’s visits to the canteen. Location sensing makes use of GPS and network provider to check if user is near to the location while the Wi-Fi sensing uses BSSID of access points to verify if user is in the location. Results show that the application has a detection rate of 100%. Location detector service took a maximum of 18 seconds to start the queue detector after the user step into the canteen and took less than a minute to detect user leaving the canteen. Current application implementation requires manual learning of nearby WAP BSSIDS. Recommendation include automated learning of BSSID to include other locations, more energy efficient location detector and exploration in in areas other than queuing for food.