Enabling Disaster Early Warning via a Configurable Data Collection Framework and Real-time Analytics

The detection and prediction of natural catastrophes or man-made disasters before they occur has recently shone the light on several relatively new technologies. Due to the significant development of mobile hardware and software technologies, a smartphone has become an important device for detecting and warning about such disasters. Specifically, disaster-related data can be collected from diverse sources including smartphones' sensors and social networks, and then the collected data are further analyzed to detect disasters and alert people about them. These collective data enable a user to have access to a variety of essential information related to disaster events. Using the example of a communicable disease outbreak, such information helps to identify and detect the ground zero of a disaster, as well as make sense of the means of transmission, progress, and patterns of the disaster. In this paper, we discuss a novel approach for analyzing and interacting with collective sensor data in a visual, real-time, and scalable fashion, offering diverse perspectives and data management components.