Publish-subscribe smartphone sensing platform for the acute phase of a disaster: A framework for emergency management support

The advanced sensors embedded in modern smartphones opens up for novel research opportunities, as for instance manifested in the field of mobile phone sensing. Most notable is perhaps research activities within human activity recognition and context-aware applications. Along a similar vein, the SmartRescue project targets monitoring of both hazard developments as well as tracking of people in a disaster, taking advantage of smartphone sensing, processing and communication capabilities. The goal is to help crisis managers and the public in early hazard detection, hazard prediction, and in the forming of risk minimizing evacuation plans when disaster strikes. In this paper we propose a novel smartphone based communication framework for disaster specific machine learning techniques that intelligently process sensor readings into useful information for the crisis responders. Core to the framework is a robust content-based publish-subscribe mechanism that allows flexible sharing of sensor data and computation results. The proposed communication platform has been tested at the proof of concept level, with several detailed features providing promising results. We also provide the initial results from the development of this platform and discuss how to enhance the platform to become a disaster monitoring system for practical use.