Output details
11 - Computer Science and Informatics
Manchester Metropolitan University
Mutual information for the detection of crush
<25>Core work from Ph.D. thesis of Harding (successfully defended November 2011). Describes a computationally “cheap” method for detecting crush in in-silico simulations of evacuations, and is one of the first applications of mutual information to the study of crowd behaviour. Highlighted at time of publication on MIT Technology Review blog. Could offer significant benefits in terms of future crowd safety, as our method allows for rapid detection of the onset of crush conditions in crowd simulations, without the need for complex physical force calculations. Describes a comprehensive computational platform, extensively validated against well-documented historical events.