Sensor Data Fusion for Intelligent Systems Monitoring and Decision Making
ABSTRACT: The DF & NN baseline Dual Node Network (DNN) Data Fusion & Resource Management (DF & RM) technical architecture will be compared and extended. The multi-source DF & RM problem functional decomposition provided by DNN will be the framework for rigorous problem-to-solution space technology trades. To address the AF Air Logistics Centers (ALC) machine maintenance problems, the DF & NN team will extend its Enterprise Satellite as a Sensor (E-SAS) Condition-Based Health Management (CBHM) software that is currently an operational prototype at Schriever AFB used to detect, characterize, and track abnormal satellite events. The technical objective for the IBM InfoSphere Streams integration tasks is to demonstrate the ability to leverage the computational capability, agility, performance, and scalability of Streams as a fusion system architecture platform. The DF & NN team will extend and test its large volume data mining, fusion, and CBHM tools on enormous ALC data sets. Then we will design, develop, and test its multi-source fusion system architecture implementations based upon the ALC identified operational needs and the DNN DF & RM technical architecture. The DF & NN team will provide affordable data-driven tools for abnormality detection and characterization which will be tracked and used to provide rigorous higher level DF & RM capabilities to include relationship, impact, performance, and context assessments. BENEFIT: The competitive advantage of this DF & RM CBHM technology is in its affordability derived from the reusability provided by the DNN technical architecture and its extendibility derived from the data driven pattern learning software. Due to the substantial data driven nature of the DF & NN team software, the Condition-Based Health Management (CBHM) products from our current SBIR Phase II will be easily extendable to detect, recognize, and track abnormalities in any commercial or government system that provides state of health or normal operations data to include manufacturing and cyber system applications. Commercial and military satellite operators have already shown significant interest in having these products applied for Blue Force Status (BFS) capabilities. We have already delivered operational prototypes of this capability to 2 satellite operations sites and applied the tools off-line to over 100 different large to enormous real data sets. The next generation fusion systems will need to automatically find or accept from the user relevant context for situation awareness (SA), assess the concurrency of the SA picture with the context, and then incorporate this context into the current SA picture as a basis for the fusion of the next set of SA inputs. This capability will require automated data fusion process assessment which drives the SSA system process concurrency management and visualization software such as being developed under this SBIR program. These advanced capabilities will be developed for DF & RM web service orchestration in net-centric, service oriented architectures (SOA) to maximize the utility to the broad user community to include the Joint Space Operations Center (JSpOC) and the National Air and Space Center (NASIC) all-source intelligence analysts. A recent example of this is the net-centric site provided by DF & NN for space catalog Abnormal Catalog Update (ACU) detections.
Small Business Information at Submission:
Data Fusion & Neural Networks, LLC
1643 Hemlock Wy Broomfield, CO -
Number of Employees: