Information Fusion and Response Guidance

Chemical and biological (CB) detection technologies have become more mature, and some are now available for operational use. As the sensor availability has increased, so too has the realization that the interpretation of the sensor output is nontrivial. The sensor data can be coupled with non-sensor information about the operational and environmental conditions prevailing at the time of the sensor's output. We refer to this combined informa-tion—from sources such as intelligence, adversary or agent dissemination surveillance, weather conditions, and situational vulnerabilities—as contextual information. For example, a battlefield chemical sensor alert may be interpreted differently if additional contextual information indicates that there was an artillery shell that landed upwind of the sensor position in the recent past. What is needed is information fusion—that is, the intelligent combination of multiple information sources to enhance the decision maker's understanding of the data and its implications. The mere combination of this information through networking, communications, databases, and displays does not constitute true information fusion. One category of fusion problems requires automated pattern recognition. Pattern recognition methods produce decisions about which one out of multiple possible hypotheses is true with respect to a given object or event. One such hypothesis space is the question of whether a biological or chemical agent is present in the environment. Decision makers lack experience responding to CB contamination events, and so could use automated assistance in choosing the appropriate course of action. We define this assistance as response guidance. By the definitions above there are few if any information fusion or response-guidance solutions available