Context aware ad hoc network for mitigation of crowd disasters

Our research works focuses on the design and implementation of a novel ubiquitous multi context-aware mobile phone sensing network for mitigation of crowd disasters using machine-to-machine (M2M) communications. A mobile sensor network system integrated with wireless multimedia sensor networks (WMSNs) was designed for effective prediction of a stampede during crowd disasters. This proposed sensor network consists of mobile devices that are used as crowd monitoring participant nodes that employ light sensors, accelerometers, as well as audio and video sensors to collect the relevant data. Real-time crowd dynamics modeling and real-time activity modeling have been achieved by implementing the algorithms developed for Context Acquisition and multi-context fusion. Dynamic crowd monitoring was achieved by implementing the context based region identification and grouping of participants, distributed crowd behavior estimation, and stampede prediction based on distributed consensus. The implementation of the proposed architecture in Android smartphone provides light-weight, easy to deploy, context aware wireless services for effective crowd disaster mitigation and generation of an in time alert to take measures to avoid the occurrence of a stampede. The system has been tested and illustrated within a group of people for stampede prediction by using empirically collected data.