Multiagent System Helps Train for the Unthinkable

October 1, 2007

IEEE Intelligent Soceity

Article by Mark Ingebretsen

The UK terrorist attacks this June demonstrated once again the extremists’ preference for hitting multiple targets. This strategy aims to confuse authorities and increase casualties and damages. If a large city becomes the target of multiple attacks, responders will need to scramble to allocate fire trucks, ambulances, and other resources to any number of locations. Meanwhile, they’ll need to balance factors such as the intensity of the resulting fires or bomb-related damage at each location, as well as traffic and wind patterns throughout the city. And they’ll need to consider the possibility that more attacks might be imminent. If ever a problem was in need of an AIbased simulation, this is it. And that’s just what DEFACTO (Demonstrating Effective Flexible Agent Coordination of Teams through Omnipresence) aims to provide.

news_20071001.pdf391 KB
See also: 2007