Information fusion for natural and man-made disasters

The Center for Multisource Information Fusion (CMIF) at the University at Buffalo is in the early stages of a multi-year research program addressing information fusion (IF) to support crisis-center decision-makers dealing with post-event situations for both earthquakes (a natural disaster) and chemical attacks (a man-made disaster). This paper describes the overall strategic approach to this research program, to include the need for: careful domain analysis, developing representative data sets, an effectiveness-oriented approach to the evaluation of the derived IF technology, and a strategy for assessing IF technique robustness. Examples are given, of the research in: techniques for data generation, data association techniques for highly-disparate and latent data, dynamic vehicle routing, modeling of hospital decision-making operations, and initial IF-based concepts for inferring the state of lifeline-infrastructure systems as "situational estimates". Plans for out-year research will also be discussed, as well as interactions with industrial partners.