A Deep Learning Approach for Enhanced Discrimination of Nuclear Weapons Testing

Award Information
Agency: Department of Defense
Branch: Defense Threat Reduction Agency
Contract: HDTRA119P0026
Agency Tracking Number: T182-005-0049
Amount: $150,000.00
Phase: Phase I
Program: SBIR
Solicitation Topic Code: DTRA182-005
Solicitation Number: 2018.2
Timeline
Solicitation Year: 2018
Award Year: 2019
Award Start Date (Proposal Award Date): 2019-04-03
Award End Date (Contract End Date): 2019-11-07
Small Business Information
7474 Greenway Center Drive, Greenbelt, MD, 20715
DUNS: 112935437
HUBZone Owned: N
Woman Owned: N
Socially and Economically Disadvantaged: N
Principal Investigator
 Dr. Hafidh Ghalib
 (321) 752-4985
 hafidh.ghalib@arrayinfotech.com
Business Contact
 Robert Deegan
Phone: (301) 345-8188
Email: bob.deegan@arrayinfotech.com
Research Institution
N/A
Abstract
Seismic, hydro acoustic, and infrasonic (SHI) analyses possess many desirable properties, such as long-distance wave propagation and attributable spectral characteristics. They have been, by large, relegated to capturing massive explosions, earthquakes, bolides, and space-bound rocket launches. Although machine learning techniques have been employed on various military and commercial projects, rarely has there been an attempt to marry the adaptive capabilities of the machine learning algorithms with all three sources that surround events of interest. Our revolutionary approach, proposes to develop augmented machine learning techniques that inherently leverage the existence of all three sensor data types. The inclusion of each data type is of utmost importance, where event characterization and development of propagation models have clearly demonstrated the impact of source types, atmospheric conditions, and terrain geology on acquired SHI waveforms, respectively. This is contrast to most contemporary methods that look at one or at most a few features or discriminants to make a decision on source type, based on physical modeling or observational. Our approach utilizes both, the sensor data type when available or perform without it for a strictly automated solution. The innovative technical merits we propose have real value that is direct applicable to support DTRAs mission to

* Information listed above is at the time of submission. *

Agency Micro-sites

SBA logo
Department of Agriculture logo
Department of Commerce logo
Department of Defense logo
Department of Education logo
Department of Energy logo
Department of Health and Human Services logo
Department of Homeland Security logo
Department of Transportation logo
Environmental Protection Agency logo
National Aeronautics and Space Administration logo
National Science Foundation logo
US Flag An Official Website of the United States Government