Data Analysis and Mining for Penetration Environment Dynamics (DAMPED)

ABSTRACT: We propose to develop an analysis/tool integration architecture to enable the synergistic exploration and capture of the critical features of the penetration fuze environment found within combined experimental and computational datasets. We will develop a design of the architecture and investigate the required capabilities of the architecture (software and hardware). We will implement a subset of the complete architecture as a prototype and perform preliminary modal data mining to demonstrate the efficacy of the approach to hard target fuze penetration environment characterization. As this effort proceeds into Phase II, we envision the development of a systematic design of experiments methodology that can perform data mining on the complex underlying database, ultimately serving as a fuze design tool for survivability and performance. BENEFIT: The tool developed in this SBIR will enable the government to integrate the tools required to perform penetration data generation, analysis, and mining to characterize the complexity of the penetration environment dynamics. It will address a number of challenging features from the efficient access and management of the underlying database to the incorporation of tools that can analyze the data and identify metrics capable of mapping a survivability envelope of fuzes and predicting their failure. We focus on aspects of the canonical problem domain in which complex weapon target interaction drives the development, modeling, and testing of penetrating weapons and their associated fuzes. By using the tools from this SBIR, contractors and government agencies will be able to create integrated analysis plans that span the test, simulation, and analysis domains, encouraging reuse of algorithms, tools, and analysis packages without the challenges of consideration for the ability to use data across the domain boundaries.