SBIR-STTR Award

Layered Data to Areas of Interest
Award last edited on: 10/29/2018

Sponsored Program
SBIR
Awarding Agency
DOD : Navy
Total Award Amount
$708,165
Award Phase
2
Solicitation Topic Code
OSD13-LD3
Principal Investigator
Yuri Levchuck

Company Information

Intelligent Models Inc

9710 Traville Gateway Drive
Rockville, MD 20850
   (301) 236-5150
   info@intelligentmodels.com
   www.intelligentmodels.com
Location: Single
Congr. District: 08
County: Montgomery

Phase I

Contract Number: N00014-14-P-1122
Start Date: 3/13/2014    Completed: 9/13/2014
Phase I year
2014
Phase I Amount
$149,803
According to the U.S. Navy Information Dominance Roadmap, 2013?2028, prevailing in the 21st Century combat environments is contingent upon addressing the intensifying challenge of Battlespace Awareness and Information Dominance, i.e., upon promptly deriving enhanced information content from advanced means to rapidly sense, collect, process, analyze, evaluate and exploit intelligence about adversaries and operational environments. Unfortunately, the exponential growth in ISR data volume severely limits manual processing of the available regional text and imagery/video datasets, necessitating the automated exploitation tools to aid systematic extraction of mission-critical Intel. To exploit very large data streams over wide areas and autonomously highlight areas of interest for tactical decisions without a priori knowledge of the area and/or location of high value, we are developing INFLECT (Intelligent Filtering for Layered Exploration with Cognitive-aid Technologies) -- a novel data-mining, correlation, and reasoning engine equipped with scalable data association, pattern-matching, and value-based filtering algorithms that automate: (1) statistical tracking of situational dynamics to extract useful patterns and regularities; (2) heuristic appraisal of mission-specific value of information; (3) value-based filtering, triage, fusion, and summarization of large multi-modal datasets to prioritize and highlight relevant information; and (4) multi-layered inference and semantic exploration of large text/video datasets.

Keywords:
Automated Value-Of-Information Appraisals, Naturalistic Text-And-Video Fusion, Associative Semantic Knowledge Representations, Layered Exploration, Pattern Recognition Algorit

Phase II

Contract Number: N00014-16-C-1052
Start Date: 8/17/2016    Completed: 1/16/2018
Phase II year
2016
Phase II Amount
$558,362
According to the U.S. Navy Information Dominance Roadmap, 20132028, prevailing in the 21st Century combat environments is contingent upon addressing the intensifying challenge of Battlespace Awareness and Information Dominance, i.e., upon promptly deriving enhanced information content from advanced means to rapidly sense, collect, process, analyze, evaluate and exploit intelligence about adversaries and operational environments. Currently, the exponential growth in ISR data volume severely limits manual processing of the available regional text and imagery/video datasets, necessitating the automated exploitation tools to aid systematic extraction of mission-critical Intel. To rapidly exploit very large data streams over wide areas and autonomously highlight areas of interest for tactical decisions without a priori knowledge of the area and/or location of high value, we are developing INFLECT (Intelligent Filtering for Layered Exploration with Cognitive-aid Technologies) -- a novel data-mining, correlation, and reasoning engine equipped with scalable data association, pattern-matching, and value-based filtering algorithms that automate: (1) statistical tracking of situational dynamics to extract useful knowledge, patterns, and regularities; (2) heuristic appraisal of mission-specific relevance/value of information; (3) value-based filtering, triage, fusion, and summarization of large multi-modal datasets to prioritize and highlight relevant information; and (4) multi-layered inference and semantic exploration of large text/video datasets.

Benefit:
The INFLECT brings a variety of operational, tactical, and strategic benefits to the US Navy and US Marine Corps, including: (1) drastic acceleration of the Intel Preparation of the Battlespace by the Carrier Strike Group (CSG) to determine Indications and Warnings of threats in the area where the Expeditionary Strike Group (ESG) is going to land (e.g., when moving the USMC force from sea to shore); (2) rapid acquisition of consolidated, contextualized multi-INT reports compiled by the CSG, e.g., in order to determine when the current maritime domain environment is safe for the ESG to enter the area; and (3) the accelerated and more accurate CSGs Strike or Battle Damage Assessment (BDA) missions while generating summary multi-INT reports transmitted to the commanding officers and tailored to the Information Requirements (IRs). INFLECT provides the rapid (real-time) automated multimedia and ISR processing capabilities critical to successful tactical operations of the Navy and USMC. INFLECT system will imbue the future Naval Forces with a capability for immediate processing of ISR feeds, rapid acquisition of high value Intel, and its dissemination in order to generate speed and accuracy in decision making. This capability will be critical for achieving the vision outlined in the U.S. Navy Information Dominance Roadmap, 20132028, for successfully addressing the intensifying challenge of Battlespace Awareness and Information Dominance -- i.e., for promptly deriving enhanced Intel about adversaries and operational environments from multi-modal sensor feeds and multimedia content -- in order to prevail in the 21st Century combat environments. INFLECT enables the extraction and incorporation of space-time and geo-cultural context into the Intel collected over expansive areas of mission interest (e.g. Pacific Rim), in order to provide essential information to the tactical Navy/USMC users. INFLECT provides the warfighters with an automated capability for organizing the unstructured Intel (e.g., automating the content tagging while ingesting and prioritizing the flow of real-time sensor feeds), thus making the data discoverable, accessible, shareable, and ready for analysis and timely discovery of high-value targets from the all-source ISR feeds. INFLECT will benefit the ISR and Intel communities by helping teams of human operators rapidly analyze large amounts of information in complex information environments. INFLECT will benefit the organizational research community by extracting the maximum mileage 0x9D from both machine and uniquely human information processing capabilities. INFLECT will enable rapid adaptability and reusability of modern Intelligence analytics, extending the lifecycle and applicability of many technologies within the Intel analysts toolkit.

Keywords:
n/a