|
|
|
|
|
|
Advances in communication and sensor technologies have brought about huge increases in the types and amounts of information available for battle management. Shortcomings in the ability to integrate and arbitrate missing and conflicting information and the current inability to correlate and reason about vast amounts of information in real time are an impediment to providing a coherent overview of unfolding events. Integrating disparate information, such as voice, video images, and tactical displays, that has varying degrees of reliability is a first step towards battle management. Such integration can be accomplished through the application and further development of Bayesian network and intelligent agent methods. Bayesian networks provide a sound basis for a robust and potentially very efficient solution to the problems posed by incomplete/unreliable data and have proven suitable to the problem of integrating disparate types of data. Other aspects of managing different types of data can be addressed through intelligent agents.
Specific features of the decision support system that we will develop include:
The overarching conceptual framework for the proposed work in data fusion and battle management is Bayesian decision theory. The ability of commanders and warriors to perform risk assessment and make effective tactical decisions will be improved by the dynamic normative system that supports real-time integration and augmentation of disparate data
[Univ. of South Carolina] [Dept of Computer Science] [College of Science and Math]