Knowledge Based AI and IU Strategies for Situation Awareness and Assessment

In the face of the information overload that will occur with the routine use of UAVs, distributed ground sensors, and prior and updated maps, situation awareness and assessment involving force monitoring is a task that will not be feasible without the application of a wide range of artificial intelligence (Ax) and image understanding (IU) technologies, to assist human operators and command officers. The integration of these knowledge-based tools into an automated system requires careful attention to issues in data fusion, control strategies for composition of component tools into meaningful tasks, and a knowledge representation that will support spatio-temporal reasoning about terrain and force disposition. The key to solving these issues lies in the ability to select the right algorithms, with the right set of parameters at the right time, and apply them to the appropriate subset of data. We propose a novel approach to knowledge-directed control and construction of context-sensitive IU strategies for force monitoring. The system will apply Al and IC technology intelligently by choosing a sequence of algorithms that is appropriate to the environmental context, as well as a partial interpretation of the current state of the battlefield, and any relevant stored knowledge. Quite diverse forms of knowledge must be represented, including knowledge necessary to support the following activities: recognition of general classes of natural terrain, cultural objects, and functional areas; tracking of single objects or groups over the terrain; the disposition of forces (such as location and possible intention of forces) ; aspects of the dynamic battlefield (such as the estimated time of arrival for forces on the move) analysis of visibility for force concealment; terrain navigability for determining maximum speed of vehicle movement; and control knowledge for applying IC modules effectively to achieve these goals within highly varying contexts. The technologies proposed for Artificial Intelligence Enhanced Information Processing, strongly correspond with a number of commercial or dual-use applications, such as aircraft tracking and control for commercial airfields, and intelligent highway system applica