NASA SBIR 2020-I Solicitation

Proposal Summary


PROPOSAL NUMBER:
 20-1- S5.03-6750
SUBTOPIC TITLE:
 Accelerating NASA Science and Engineering through the Application of Artificial Intelligence
PROPOSAL TITLE:
 Neuro-Cognitive Radar-Imagery Segmentation and Identification (Neuro-RSI)
SMALL BUSINESS CONCERN (Firm Name, Mail Address, City/State/Zip, Phone)
American GNC Corporation
888 Easy Street
Simi Valley, CA 93065
(805) 582-0582

Principal Investigator (Name, E-mail, Mail Address, City/State/Zip, Phone)

Name:
Dr. Francisco Maldonado
E-mail:
fmald@americangnc.com
Address:
888 Easy Street Simi Valley, CA 93065 - 1812
Phone:
(805) 582-0582

Business Official (Name, E-mail, Mail Address, City/State/Zip, Phone)

Name:
Emily Melgarejo
E-mail:
emelgarejo@americangnc.com
Address:
888 Easy Street Simi Valley, CA 93065 - 1812
Phone:
(805) 582-0582
Estimated Technology Readiness Level (TRL) :
Begin: 3
End: 4
Technical Abstract (Limit 2000 characters, approximately 200 words)

 

To provide NASA with innovative machine learning segmentation, classification, and analysis technology for radar imagery, specifically Synthetic Aperture Radar, American GNC Corporation (AGNC) and The Pennsylvania State University’s Radar and Communication Laboratory (PSU-RCL) propose the Neuro-Cognitive Radar-Imagery Segmentation and Identification (Neuro-RSI) system. Looking to ensure an adaptable and flexible system, Artificial Intelligence techniques enable a cognitive approach that emulates the human analysis process of SAR imagery data. While there have been advances related to development of new technologies for SAR Systems as well as methods for the analysis of their imagery, a generalized SAR imagery software analysis toolbox does not exist. The Neuro-RSI aims to provide a system that can be applied to diverse SAR technologies by a system architecture with cutting-edge low-level processing and segmentation as well as state-of-the-art pattern recognition, self-learning, and high-level cognition. This new AI-based SAR imagery analysis system is intended to then be deployed for enhancing NASA’s automatic image analysis needs such in the Earth Science Data Systems Program. Innovations include: (1) flexible and reconfigurable software architecture that dynamically tunes the processing flow according to analysis goals; (2) cognitive analysis building on incremental system knowledge, adaptive SAR imagery analysis, and logic inference mechanisms; (3) integral approach to optimize the processing framework to characteristics of SAR systems; (4) pattern recognition schemes based on a deep learning framework; and (5) goal oriented agent that enables expansion of analysis capability working with a Query Set.

Potential NASA Applications (Limit 1500 characters, approximately 150 words)

The Neuro-RSI System can be used in a wide variety of NASA Earth Science Data Programs and remote sensing applications for automated and intelligent analysis of SAR imagery. Examples include analysis of wildfires, vegetation, urban sprawl, weather patterns, hurricanes, soil, water/ice levels, biomass, droughts, deforestation, agricultural/land use, volcano effects, etc. NASA’s Carbon Cycle Ecosystems Office, Earth Observing System Data and Information Systems, and Earth Science Technology Office can all benefit from the Neuro-RSI’s technology.

Potential Non-NASA Applications (Limit 1500 characters, approximately 150 words)

The Neuro-RSI System is related to many growing global markets such as Synthetic Aperture Radar, Artificial Intelligence, Security and Surveillance Radar, and others. This AI-based SAR imagery system can facilitate the automated analysis for: (a) disasters (e.g. floods, wildfires, oil spills); (b) traffic patterns; (c) agriculture; (d) urban/construction planning; (e) facility surveillance, etc.

Duration: 6

Form Generated on 06/29/2020 21:12:48