Goals: The objective of this project is the research and development of crowd safety solutions in the domain of (urban) mass events. It was found that many crowd disaster - such as the 2012 Duisburg Loveparade disaster - could have been avoided given better management structures and warning systems. The project aims to develop a complete solution to support event organizers and rescue forces in detecting and counteracting potential dangerous situations in the domain of (urban) mass events. The approach that is followed in this project is predicated on crowd sourcing and participatory sensing utilizing mobile devices of spectators. Following a data-based approach, algorithms will be developed and optimized utilizing annotated context traces collected during real-world mass events and in turn tested in the real-world. The specific objectives that will be pursued in a tightly orchestrated way by the proposal include:
Creating a comprehensive dataset of annotated context traces representing human movement patterns and situations of high crowd density in various environments.
Analysis of existing methods and development of new methods to forecast crowd density based on collected datasets.
Experimentation with- and Evaluation of developed methods in real-world mass events.
Implementation of a crowd safety solution based on the developed crowd density forecasting methods.
The project will be carried on by the Institute of Pervasive Computing with complementary competences from partner institutions and will touch on - but not exclusively - the SAPERE, SOCIONICAL and DISPLAYS project.