Higher level fusion for post-disaster casualty mitigation operations

The overall goal of the research presented in this paper is to develop a methodology for higher level fusion processes (situation assessment and impact prediction, STA) in a dynamic post-disaster environment. In particular, this paper addresses the problem of situation assessment to support casualty mitigation operations during the initial response phase after an earthquake. The purpose of the STA module is to infer and approximate the critical characteristics of the disaster scene in relation to the ultimate goal of the decision makers, which is to service the maximum number of the highest priority casualties with minimum service times, and further to reduce the risk of additional casualties. STA is performed by analyzing spatial and temporal relations of the casualties and casualty aggregations at different levels of granularity, and their dynamics relative to care-providers, transportation, communication and other critical resources within the overall situational context. This assessment process comprises preprocessing in which lower level fused casualty results are evaluated to see if they satisfy a certain quality criterion, dynamic hierarchical "any time" clustering, and reasoning about uncertain dynamic characteristics of aggregates, their relations, discovery of situational events and their causes, and prediction of consequence of the current situation.