Multi-Sensor Information Fusion and Information Visualization
Small Business Information
1235 Jefferson Davis Highway, Suite 400, Arlington, VA, 22202
Manager, Business Ops
Manager, Business Ops
AbstractAchieving system-level optimization across all sensor/weapon system functions with subsystems that are designed to perform optimally on a specific function is a difficult challenge. In Phase I of this SBIR we demonstrated the utility of applying advancedBayesian network techniques to perform multi-source information fusion for the purpose of providing optimal decision support for a specific mission function. The objective of this Phase II research is to expand this capability to provide a morecomprehensive, system-level optimization in the context of the OODA (Observe - Orient - Decide - Act) concept for the Joint Strike Fighter program. This prototype system will utilize advanced Bayesian network techniques to provide high level inference aswell as a unique Value of Information (VOI) query capability. The inference data provides fused information to support mission planning functions. The VOI query capability provides the means for determining optimal sensor resource allocations that willbest satisfy the near-future information needs of the mission planning functions. Furthermore, we will apply advanced algorithms for performing the sensor optimization between multiple distributed platforms. The fundamental benefit will be to relievepilots of the burden of detailed sensor management in both own-ship and multi-ship operations. Representations of the same software will be incorporated in mission planning applications so that mission plans will maximize the use of the Mission OptimizedSensor Tasking (MOST) system.Commercial applications apply to all dynamic decision problems in military systems where time-critical resource allocation must be made in uncertain environments. The development of a real-time situation-specific decision support system will havewidespread commercial application in both DoD and the private sector. Any problem that has a disparate set of data sources (each with its own uncertainty) that require fusion and real-time decision support is a candidate for this technology.
* information listed above is at the time of submission.