Grid-Based Real-Time Enterprise Intelligence System for Logistics Data Mining and Integration

To address the U.S. Armys need for an enterprise data mining expert system with real-time decision support to fleet managers, operators, and commanders, Physical Optics Corporation (POC) proposes to develop a novel Grid Real-time Enterprise Intelligence Service (GREIS) incorporating grid-based data mining, advanced statistical classification, and a high-performance net-centric rule engine, all integrated within an enterprise system architecture. The innovative use of multidimensional hierarchical online analytical processing cubes, a net-centric rule engine, and advanced statistical classification methods enables the GREIS to perform multidimensional data mining on both structured and unstructured logistics data, quickly capture diagnostic knowledge and data mining results for diagnostic reasoning and trend prediction, and discover hidden fault patterns and trends with high consistency, reliability, and accuracy. The GREIS is implemented via a map-reduce programming paradigm with autonomous task splitting and parallel processing, saving an order of magnitude of time in real-time execution. In Phase I POC will demonstrate the feasibility of GREIS through analysis and simulation, including DoDAF OV-1 development, data gap analysis, and development and integration of key components into a Phase I prototype. In Phase II POC will develop a simulated prototype to demonstrate realistic operation of the GREIS in global fleet support environments.