---------- OSD ----------

65 Phase I Selections from the 11.3 Solicitation

(In Topic Number Order)
Galois, Inc.
421 SW Sixth Suite 300
Portland, OR 97204
Phone:
PI:
Topic#:
(503) 808-7159
Mark Tullsen
OSD11-CP1      Awarded: 5/7/2012
Title:PICT, A Tool For Interactively Capturing and Managing Programmer's Intentions
Abstract:The so-called "software crisis" is continuing unabated. It is all too common for a software project to create software that is hard to maintain, inefficient, bug-ridden, late, and over- budget. Software flaws resulting from this current state of affairs are no longer an inconvenience but are the largest source of software security vulnerabilities. Motivated by a desire to increase our ability to reduce software flaws, we propose to add to the software engineer's toolbox a new tool---one that is complementary to existing type checking, testing, and verification approaches---which can measurably increase programmer productivity and decrease software defects. This new tool will interactively capture code properties and allow for the managing, analysis, and display of these properties during software development. The code properties that can be captured by our tool can range from low level code assertions ('p' isn't a null pointer) to high level design decisions (class 'C' is a singleton class). The tool is designed to provide feedback about global analyses of the program, work in real-time, be unobtrusive to the programmer, distinguish inferred properties from intended properties, and minimize programmer input.

GrammaTech, Inc
531 Esty Street
Ithaca, NY 14850
Phone:
PI:
Topic#:
(607) 273-7340
Michael McDougall
OSD11-CP1      Awarded: 5/7/2012
Title:Enhancing Code Awareness in Software Development Environment
Abstract:Most large pieces of software are sufficiently complex that no individual completely understands all of its subparts. In order to make progress, programmers must operate in this environment of incomplete information. In many cases, programmers make poor design and implementation decisions because of a lack of awareness of how pieces of code interact with each other. We propose developing the Code Awareness Assistant (CA2), a virtual coding partner that looks over an engineer’s shoulder and gives extra context while pointing out problems as they are introduced. CA2 will be integrated with the engineer’s development environment, and give the engineer a broader view of the code than is available in contemporary IDEs. CA2 will visualize dependence information, automatically extract key facts about code, and flag common software flaws. This immediate feedback will allow developers to quickly converge on high quality code; security, maintainability and knock-on effects will no longer be afterthoughts. The CA2 will leverage GrammaTech’s proven technology for static analysis, which captures the semantics and structure of software in enough detail to infer critical design elements of the code.

Maya Software, Inc.
11425 charsan Lane
Cupertino, CA 95014
Phone:
PI:
Topic#:
(408) 252-1808
Doron Drusinsky
OSD11-CP1      Awarded: 5/4/2012
Title:Enhancing Code Awareness in Software Development Environment
Abstract:Unit testing is the prevailing approach used by programmers worldwide to verify their decisions and assumptions on a unit, or module level. Even with the advent of automated unit- test frameworks, a programmer’s knowhow can seldom be shared with other programmers; in addition, no tool alerts to programmer about the concerns associated with his or her on- the-fly programming decisions. Our proposed innovation will enhance unit test frameworks with a sharable learning and pattern-matching tool. The tool will automatically learn from the experience gained by programmers in a group (as manifested by the concerns they addressed in their unit-test suites, e.g. in a buffer overflow related test-suite), and will later direct other programmers to be aware of similar concerns when the tool detects, using pattern matching techniques, that they are manipulating data-structures with similar concerns (e.g., when their code manipulates a buffer it is related to the buffer-overflow concern). Once alerting the programmer to concerns related to his or her emerging program, the tool will also suggest a set of tests developed by other programmers that covers this concern (e.g., buffer-overflow tests). Discovering new related tests will help the programmer to better understand address these concerns.

Aptima, Inc.
12 Gill Street Suite 1400
Woburn, MA 01801
Phone:
PI:
Topic#:
(781) 496-2444
Alan Carlin
OSD11-CR1      Awarded: 6/5/2012
Title:Higher Automated Learning (HAL)
Abstract:The preparation of training materials is a labor-intensive process. It typically starts with manual task analysis or knowledge elicitation sessions requiring significant time commitments from both training professionals and SMEs. Those products must then be manually transformed into the building blocks for training content development and learning assessment. Clearly it would be of great benefit to automate the development of training materials, and the proposed solution Higher Automated Learning (HAL), represents a first step in that direction. HAL uses a Partially Observable Markov Decision Processes (POMDP) model based on the domain ontology and student performance data to model the learning of the target domain. HAL’s model will feature nodal representations of students’ actual learning states as they progress (or struggle) through the curriculum. Additionally, HAL will automate the definition of metrics and assessments to provide trainees with a sense of where they stand, what their future performance potential is, and what experiences might support more rapid improvement. When fully developed at the end of Phase II, HAL will accelerate and deepen the development of student/trainee models and enable automatic discovery of performance metrics, which identify critical relationships among key constructs and support diagnostic assessments of student states.

Design Interactive, Inc.
1221 E. Broadway, Suite 110
Oviedo, FL 32765
Phone:
PI:
Topic#:
(407) 706-0977
Meredith Carroll
OSD11-CR1      Awarded: 6/5/2012
Title:Development of an Approach for Multi-axis Assessment for Diagnosticity Enhancement in Individualized Training (MADE-IT)
Abstract:Design Interactive (DI) proposes the Multi-axis Assessment for Diagnosticity Enhancement in Individualized Training (MADE-IT). MADE-IT will be a trainee model development approach and software tool that will support a user in developing a domain-specific individualized trainee model for integration into an ITS. MADE-IT will enhance current trainee modeling approaches in the following by 1) incorporating a domain-specific performance ontology that will: represent domain knowledge, reason about the elements and relations of that domain, support evaluation of performance and interpretation of data and provide scoring models and criteria for performance, 2) layering on affective state metrics of the learner to understand how emotions, moods, and attitudes are playing into performance and learning, 3) layering on cognitive state metrics of the learner to understand how cognition and limitations are affecting learning and performance, 4) factoring in individual traits that affect learning, such as learning styles and self-esteem, and 5) utilizing a machine learning agent make inferences and predictions about trainee knowledge, skills and learning based on available (and potentially limited) performance and physiological data.

Discovery Machine Incorporated
454 Pine Street Suite 1C
Williamsport, PA 17701
Phone:
PI:
Topic#:
(570) 329-0251
Todd W. Griffith
OSD11-CR1      Awarded: 6/5/2012
Title:Virtual Observer Controllers for Adaptive Training (VOCAT)
Abstract:This proposal describes how adaptive training for tactical decision making can be obtained through the use of Virtual Observer/Controllers (VO/C), which shall be developed to mimic human observer/controllers (O/Cs). Our objective in this proposal is to use our patented reflection technology to model VO/Cs that can perform the jobs of a human O/C for an Anti- Submarine Warfare Evaluator (ASWE), thus developing an adaptive training solution for ASW that effectively and efficiently individualizes training for trainees by identifying non- optimal instances of state and/or performance, identifying the root cause(s) of issues, mitigating root cause(s) of issues in an appropriate manner, and summarizing trainee status post-scenario via a comprehensive, targeted AAR. The end result is a training system that helps to enable realistic training. Although our approach shall be general, the proposed effort shall focus on the needs of our HIFAST transition partner, with initial focus on the ASWE. The ASWE is critical to operations and targeted for early transition of training. The proposed Virtual Observer Controllers for Adaptive Training (VOCAT) shall leverage work from past efforts in creating domain specific consoles to support instructors in training. We focus on the capture of expertise from instructors and the design of the VOCAT console.

Soar Technology, Inc.
3600 Green Court Suite 600
Ann Arbor, MI 48105
Phone:
PI:
Topic#:
(919) 967-5079
Robert Wray
OSD11-CR1      Awarded: 6/5/2012
Title:Adaptive, individualized training assessment capability (AITAC)
Abstract:The revolution in simulation and virtual technologies for training is enabling learners to practice skills and learn in realistic environments. Researchers are also seeking to develop algorithms that can tailor learner practice to the estimated abilities and needs of individual students, offering the promise of individual adaptation to each learner’s dynamically evolving zone of proximal development. Such systems will likely improve learning outcomes, reduce learner frustration, and mitigate the need for artificial (out of domain) scaffolding. However, such adaptive training requires that a learning environment estimate student proficiency and need as learning progresses and most approaches assume that the primary assessment dimensions are known in advance, which limits the ultimate effectiveness of adaptive training systems. What would be preferable is the ability of the system to assess progress and stumbling blocks within the environment and diagnosis and measure these deficits in terms of dimensions “outside” of the a priori trainee representation. We propose a direct questioning approach, building on domain ontologies and prior work in machine learning of student proficiency modeling, along with dynamic assessment and tailoring, to build a more accurate and more extensible representation of individual trainees.

Stottler Henke Associates, Inc.
951 Mariner's Island Blvd., STE 360
San Mateo, CA 94404
Phone:
PI:
Topic#:
(650) 931-2700
Sowmya Ramachandran
OSD11-CR1      Awarded: 6/5/2012
Title:Multidimensional Learner Assessment to Support Intelligent, Adaptive Digital Tutors
Abstract:One-on-one tutoring has long been the gold-standard for individualized, adaptive instruction. Intelligent Tutoring Systems are Artificial Intelligence systems that have as their goal realizing the benefits of one-on-one teaching using technology. In order to deliver on this promise, these tutors will need a multidimensional assessment capability that is the mark of the good human tutor. Such a capability is however still under development. We propose to develop and implement a technique for building individualized representations of trainee performance that can be used for a multi-faceted assessment of current performance and to forecast future training needs. Drawing upon research in diverse fields, we will in Phase I both develop an initial model of performance assessment and design techniques for automating this analysis. The model will integrate physiological measurements, psychometric evaluations, and simulations to support assessment and training. When situated within a specific training context, the model will (1) operationalize system performance and (2) define normative behaviors within that context. The assessment system will use a variety of ontological representation and probabilistic reasoning techniques to analyze performance and sensor data into an actionable student model. Phase I will also result in a draft model, Phase II system design and a proof-of-concept prototype.

361 Interactive, LLC
408 Sharts Drive Suite 7
Springboro, OH 45066
Phone:
PI:
Topic#:
(937) 743-0361
Michael J McCloskey
OSD11-CR2      Awarded: 5/7/2012
Title:C-DAT: The Cognitive Desktop Analysis Trainer
Abstract:The emergence of new technologies in the Air Force GEOINT analyst's arsenal of resources represents an evolutionary advance in potential real-time and forensic analysis capabilities, including more accurate and effective pattern of life development, more timely identification of insurgent networks and cells, and more accurate and targeted products for customers. But with these new capabilities come challenges. The ability to effectively exploit these datasets will be based on an analyst's abilities to understand important contextual factors of the operating environment and sift through massive datasets. Cognitively-based tools are needed to train analysts in optimally exploiting ISR datasets. We will leverage innovative simulation interview and cognitive task analysis methodologies to identify the cognitive challenges associated with conducting effective analyses in complex, uncertain environments. An analyst-centric focus will drive the development of an adaptive desktop trainer to support imagery analysts in better managing their ambiguous and time-critical tasking that flows from irregular warfare. Training modules will support the following objectives: increasing awareness of the importance of understanding contextual factors for intelligence analysis, identifying significant patterns of life, and developing effective, expert- based strategies for making sense of ambiguous analysis tasks. Such concepts will be generalizable across domains where imagery exploitation is critical.

Aptima, Inc.
12 Gill Street Suite 1400
Woburn, MA 01801
Phone:
PI:
Topic#:
(781) 496-2408
Cullen Jackson
OSD11-CR2      Awarded: 5/9/2012
Title:CREATE: Contextually Relevant Exploitation & Analysis Training Environment
Abstract:Imagery analysts face challenges that complicate their ability to gain and maintain contextual awareness. On the job training provides some instruction but tends to focus too much on the specificities of the domain. When analysts view imagery that deviates from that domain, they tend to make mistakes such as mis-identifying equipment or failing to recognize patterns of enemy activity. Aptima, Inc., and its partner Rickard Consulting, propose to develop the Contextually Relevant Exploitation & Analysis Training Environment (CREATE) to train analysts to exploit static and full motion imagery across a variety of contexts. CREATE is an innovative adaptive training system that accelerates contextual learning by constructing individualized training programs. As analysts progress, the system identifies exercises which will accelerate the student’s acquisition of contextual expertise based on an understanding of prior performance. Training through the CREATE system will be grounded within the analyst’s typical workflow, so that the same system can be used for both training exercises and mission rehearsal. In addition, it will integrate with SCORM-conformant LMSs. To support the content authoring processes, CREATE will monitor how students respond to content and use advanced modeling techniques to classify and update content difficulty profiles based on observed student performance.

Scientific Systems Company, Inc
500 West Cummings Park - Ste 3000
Woburn, MA 01801
Phone:
PI:
Topic#:
(781) 933-5355
Ranga Narayanaswami
OSD11-CR2      Awarded: 5/7/2012
Title:ISR-IT: Adaptive Desktop Trainer for Imagery Analysis Based on Contextual Factors
Abstract:The analysis and interpretation of Intelligence, Surveillance and Reconnaissance (ISR) imagery data is a key driver of actionable intelligence. With the current operations, being primarily in Iraq and Afghanistan, it is important for us to understand the cultural, environmental, behavioral, economic factors in the interpretation of ISR imagery. A number of scenarios exist where the ISR imagery can be potentially misinterpreted, leading to unintended civilian casualties as well as mission failure. Scientific Systems Company Inc. (SSCI) proposes the development of a desktop trainer for ISR imagery analysis based on contextual factors. The proposed ISR Imagery Trainer (ISR-IT) will allow the incorporation of scenarios where context plays a crucial role in interpretation of imagery. ISR-IT will be able to ingest EO and IR video streams and incorporate a game and simulation Engine. We will use a learning management system to adapt to an individual in presenting an ISR scenario. We will use an interactive user interface to show EO and IR imagery (via video and simulation) and provide for a varying menu of contextual factors. We will design ISR-IT to incorporate into a larger system with capability to accept scenario inputs and simulation inputs.

Toyon Research Corp.
6800 Cortona Drive
Goleta, CA 93117
Phone:
PI:
Topic#:
(805) 968-6787
John A. Berger
OSD11-CR2      Awarded: 5/10/2012
Title:Adaptive Desktop Trainer
Abstract:Developing an adaptive desktop trainer to effectively train imagery analysts is crucial for executing precise intelligence, surveillance, and reconnaissance (ISR) missions that rely heavily on the exploitation of full motion video (FMV). Typical training environments for aerial platforms include using low fidelity simulated FMV data, archived FMV data taken from prior missions, or live, small scale experiments with aircraft flown in representative environments with threat actors. Often, such training measures are cost prohibitive and require heavy logistics for execution and assessing analyst performance. Toyon Research Corporation proposes a compact Desktop Trainer that incorporates video containing ground truth information, sophisticated ISR algorithms to aid analysts, an advanced learning management system (LMS), geographic information system (GIS) environment modeling for incorporating context evidence, and adaptive training material for an innovative approach to training Imagery Analysts. Toyon’s Desktop Trainer will be designed to incorporate evolving mission concepts of operation (CONOPS), as well as analyst collaboration and wide-scale ISR missions. Imagery Analyst performance will be validated using a Training Scorecard tool that incorporates lessons learned from the training regimen.

Aptima, Inc.
12 Gill Street Suite 1400
Woburn, MA 01801
Phone:
PI:
Topic#:
(781) 496-2451
Michael Garrity
OSD11-CR3      Awarded: 3/16/2012
Title:Simulated Operational Communications and Coordination Integration for Aircrew Learning (SOCIAL) Tool
Abstract:As the Air Force responds to high demands for Remotely Piloted Aircraft (RPA) operations, devices and strategies for effectively training Pilots and Sensor Operators are likely to increase in importance. While simulation-based training has evolved in recent years, the majority of attention has been directed towards training piloting skills. Current tools and devices for preparing RPA crews lack a robust ability to realistically simulate the complex communication and coordination interactions with the ISR community. Thus, a Simulated Operational Communications and Coordination Integration for Aircrew Learning (SOCIAL) Tool, an adaptive, data-driven, model-based communications simulation tool, with performance measurement capabilities, is required to provide RPA crews with the relevant training experiences necessary to develop the critical communications and coordination skills required for their evolving missions and continuous operations. The SOCIAL Tool will be designed in a way that supports integration with existing RPA crew training devices. When fully implemented, the SOCIAL Tool will address a critical gap in current RPA training.

Lumir Research Institute, Inc.
301 East Fairmont Drive
Tempe, AZ 85282
Phone:
PI:
Topic#:
(480) 773-5241
Lisa Holt
OSD11-CR3      Awarded: 3/20/2012
Title:RPA Simulated Operational Communications and Coordination Integration for Aircrew Learning (SOCIAL)
Abstract:Lumir proposes to build the RPA SOCIAL System, a training supplement providing realistic communications between RPA crewmembers and other mission-relevant entities in simulation-based training environments. The RPA SOCIAL System will enable RPA crewmembers to train as they will fight—in a task-saturated, high-tempo environment where they interact with various entities as they control the aircraft and its sensors to accomplish mission objectives. At the heart of the SOCIAL System are SOCIAL Beings, capable of conducting realistic and appropriate interactive communications with RPA crewmembers. Instructors use the SOCIAL Director to build customized SOCIAL Being behaviors (without expert domain knowledge) for insertion into existing training scenarios. The SOCIAL Director uses information about training objectives, trainee performance and scenario complexity to guide the instructor in the enhancement of future scenarios so that they are appropriately adapted to meet trainee needs. The architecture is simulation agnostic so that SOCIAL Beings and the SOCIAL Director can be used in any HLA/DIS-compatible simulation environment. Although initial development will be focused on the RPA domain, the SOCIAL System can easily be adapted to operate with simulators in any domain by adding new SOCIAL Beings or enhancing their behaviors to encompass a wider range of situations.

ORION International Technologies, Inc.
2201 Buena Vista Dr. SE Suite 309
Albuquerque, NM 87106
Phone:
PI:
Topic#:
(505) 998-4000
Paul Hamilton
OSD11-CR3      Awarded: 3/20/2012
Title:RPA Simulated Operational Communications and Coordination Integration for Aircrew Learning (SOCIAL)
Abstract:To provide effective communication and coordination training to Remotely Piloted Aircraft (RPA) crewmembers, we propose a prototype adaptive training communication environment using our T-BORG simulation and integration framework to incorporate new and pre- existing intelligent agents and synthetic teammates at various levels of fidelity. The problem addressed is that simultaneous monitoring of multiple mIRC chat windows creates a potential for high workload, and the resulting asynchronous communication impacts coordination. These coordination breakdowns can, in turn result in the loss of a high-valued target or a failure to detect an emerging threat. Our system will address this by allowing multiple entities to interact in a complex, time sensitive scenario. Phase I will deliver a prototype system that demonstrates an adaptable architecture that simulates multiple chat windows, monitors performance, and increases the chat workload in response to this performance. The system will initially draw upon CERI’s UAV-STE but will be designed to connect to a number of simulation environments. Phase II will incorporate higher fidelity analytic and behavior subprograms with provision for substitution or addition of other code.

Sonalysts, Inc.
215 Parkway North P.O. Box 280
Waterford, CT 06385
Phone:
PI:
Topic#:
(860) 326-3792
James McCarthy
OSD11-CR3      Awarded: 3/27/2012
Title:RPA Simulated Operational Communications and Coordination Integration for Aircrew Learning (SOCIAL)
Abstract:The DoD is facing significant challenges in training Remotely Piloted Aircraft (RPA) crews. At the same time that the popularity of RPAs is increasing and the need for qualified crews is skyrocketing, airspace restrictions and limitations in simulators are making it increasingly difficult to offer effective training opportunities. To help the DoD meet these challenges, Sonalysts and our teammates are proposing to develop the REmotely Piloted Aircraft Synthetic Tutor (REPAST). REPAST is based on three innovative pillars. First, the event- based approach to training provides a theoretically rigorous and empirically validated instructional foundation. Second, interactions with members of the broader Intelligence, Surveillance, and Reconnaissance (ISR) community will be simulated by synthetic agents that use natural language processing to communicate with RPA crew members. Third, intelligent tutoring techniques are used to assess performance, provide feedback, and control task saturation, scenario difficulty, and curriculum progression.

AEgis Technologies Group, Inc.
410 Jan Davis Drive
Huntsville, AL 35806
Phone:
PI:
Topic#:
(407) 380-5001
Jake Borah
OSD11-CR4      Awarded: 4/4/2012
Title:Integrated Adaptive SCORM, HLA and DIS Compliant Learning Content Management System
Abstract:AEgis Technologies proposes in this Phase 1 project to develop a functional description document to define and document a framework capable of incorporating meta-data element enhanced SCORM conformant learning objects into a LMS to enable the delivery of challenging and motivating training tailored to an individual’s needs. The Phase 1 architectural framework will ensure individuals receive relevant training and feedback based upon established mission essential and training task listings. The architectural framework will be developed to monitor the performance of an individual, to adaptively manage the presentation of training material, and to evaluate the training level of an individual. Open standards technologies such as XML, http, DIS, and HLA will be used to provide a scalable and extendable architecture which can be commercialized in a cost-efficient manner. Phase II work will build a proof-of-concept prototype that will demonstrate capabilities of the Phase 1 architectural framework.

Aptima, Inc.
12 Gill Street Suite 1400
Woburn, MA 01801
Phone:
PI:
Topic#:
(781) 496-2449
Emily Wiese
OSD11-CR4      Awarded: 4/27/2012
Title:Adaptive Training Manager (ATM) system
Abstract:Traditional Learning Management Systems (LMSs) provide many beneficial capabilities to instructors and training organizations, mostly focused on course management in computer based training (CBT) rather than content learning. Similarly, simulation-based training is increasingly implementing the use of adaptive training methods to support more efficient use of valuable trainee and simulator time, yet lags behind other types of training in the management of the resulting performance data. Because CBT and simulation based training are often considered separately, there is little consideration of how to maximize trainee learning across instructional methods. A holistic look at trainee performance data management and use is needed to capitalize on advances being made in all areas of training and to integrate best practices and technologies currently used in both CBT and simulation based training. Aptima proposes to develop the Adaptive Training Manager (ATM) system, a module that resides between an LMS and the training systems. The ATM system will integrate the state of the art in LMS technology, performance measurement methods and tools, and adaptive simulation based training methods and technologies in order to guide the selection of training environment, the selection of available course materials, and the optimal scheduling of team training exercises.

Knowledge Based Systems, Inc.
1408 University Drive East
College Station, TX 77840
Phone:
PI:
Topic#:
(979) 260-5274
Michael Graul
OSD11-CR4      Awarded: 4/4/2012
Title:Framework for Adaptive Learning CONtent Management Delivery (FALCON)
Abstract:Knowledge Based Systems, Inc. (KBSI) proposes to research, design, and demonstrate a Framework for Adaptive Learning CONtent Management Delivery (FALCON). FALCON will include the following innovative capabilities (1) training content management and content delivery using multiple training modes and at multiple levels of fidelity; (2) automated and adaptive training performance assessment and testing mechanisms; (3) adaptive and dynamic multi-modal training content refinement based on measured learner deficiencies; and (4) compliance with training system standards including SCORM, HLA, and DIS. The proposed Phase I effort will (i) establish requirements for hybrid learning content design, assembly, and delivery for Air and Space Operations Center intelligence officer training, (ii) establish methods for combining Computer Based Training with Simulation Based Training, (iii) establish architecting approaches for combining SCORM-compliant training with HLA- DIS-compliant training; and (iv) design and demonstrate a proof-of-concept of the FALCON technology. The Phase II project will design and demonstrate a focused and scalable FALCON learning management training application.

Tier 1 Performance Solutions, LLC
6 East 5th Street Suite 400
Covington, KY 41011
Phone:
PI:
Topic#:
(859) 663-2115
Stu Rodgers
OSD11-CR4      Awarded: 4/20/2012
Title:Competency-based Adaptive Training System
Abstract:TiER1 Performance Solutions proposes to develop CATS (Competency-based Adaptive Training System), a system to address common warfighter training needs in the Air and Space Operations Center. A major goal of this effort is to prepare warfighters for mission readiness by integrating a Learning Management System (LMS), which delivers personalized content and learning events as needed and provides continuous evaluation of performance; a Learning Content Management System (LCMS), which maintains a catalog of learning and assessment opportunities; and a Competency Management System (CMS), which manages various domain-specific competency models and learner proficiency ratings. At the core of the CATS solution is a competency-based framework that is leveraged into learning objectives. These learning objectives are continuously evaluated while a learner is engaged in the system, enabling the dynamic creation and adaptation of learning paths to accelerate or remediate based on the trainee’s measured performance and proficiency goals. The design of CATS is a flexible system that incorporates the learner's previous experience, training, and proficiency in an organized way to efficiently prepare them for new career or mission goals.

Charles River Analytics Inc.
625 Mount Auburn Street
Cambridge, MA 02138
Phone:
PI:
Topic#:
(617) 491-3474
Michael Farry
OSD11-DR1      Awarded: 6/12/2012
Title:A Toolkit for Understanding Social Signals in Local Environments (TUSSLE)
Abstract:Increasingly, tactical personnel such as CoISTs, are responsible for making qualitative judgments of local conditions, or “atmospherics,” to provide insight to decision-makers in their unit as well as in higher echelons, despite a high op tempo and limited training. Tactical personnel must identify, correlate, and fuse dynamic, sparse sociocultural signals (“weak” signals) hidden within large, disparate data sets to recognize potential social disruptions or threatening conflicts. To help CoIST staff in forecasting alerts from weak sociocultural signals from diverse data, we propose to design and demonstrate a Toolkit for Understanding Sociocultural Signals in Local Environments (TUSSLE). TUSSLE incorporates a wide diversity of data, uses a hybrid approach to modeling that can specifically consider each data type, and expresses its outputs to meet the unique needs of the CoIST. TUSSLE is composed of four main components: (1) an Interactive Data Collection Tool to aid the CoIST in querying the various available sources; (2) a Sociocultural Indicator Measurement Tool to link weak signal observations to higher-order sociocultural meaning; (3) a Sociocultural Visualization Tool to represent results; and, (4) a Forecast Assessment Tool that enables the CoIST to represent and communicate the uncertainty (and the sociocultural context causing that uncertainty) to decision-makers.

TerrametricsLLC
4102 Woodbine Street
Chevy Chase, MD 20815
Phone:
PI:
Topic#:
(240) 606-8104
Gregory van der Vink
OSD11-DR1      Awarded: 5/24/2012
Title:Contextual Sociocultural Reasoning in Weak Signal Environments
Abstract:Successful military actions are becoming increasingly dependent on the compatibility of the military actions with the sociocultural and environmental context in which they occur. The objective of this proposal is to demonstrate the scientific merit and feasibility of a decision- making aid that will reveal the sociocultural characteristics most relevant for evaluating the possible success or failure of potential military operations and provide early warnings of emerging threats. We propose a methodology based on principles of signal processing, to generate probabilistic assessments of the potential outcomes, vulnerabilities, and risks for various operations. Our integrated approach to multi-disciplinary data analysis will provide the ability to detect and amplify “weak signals” – sociocultural characteristics of a population that are not readily apparent in any single set of observations but that may be predictive of military outcomes or serve as early warnings of emerging threats. TerrametricsLLC will use its substantial database on Sub-Saharan Africa for this proof-of-concept analysis.

Aptima, Inc.
12 Gill Street Suite 1400
Woburn, MA 01801
Phone:
PI:
Topic#:
(202) 552-6116
Jared Freeman
OSD11-DR2      Awarded: 6/21/2012
Title:VICTR-E: Visualizations for Integrating, Communicating and Tracking Reasoning Electronically
Abstract:As recent military missions make clear, intelligence gathering at the company level and below is essential to obtaining critical information about the socio-cultural environment. To accomplish this goal, there is an urgent need for tools to support the collaboration among these lower echelon operators in the task of collecting and analyzing socio-cultural information. Effective collaboration requires the communication of not just the findings of an analysis, but also the chain of reasoning that produced them. To address these needs, Aptima proposes to develop VICTR-E: Visualizations for Integrating, Communicating and Tracking Reasoning Electronically. VICTR-E will enable Company Intelligence Support Teams (COISTs) to inspect the range of intelligence data available to them, group related reports together, and develop graphical hypotheses or arguments that can be shared with analysts at higher echelons. Collaboration between COISTs and data gathers at the small unit level is also facilitated by VICTR-E through interactive mechanisms that allow rapid dissemination of collection requirements and retrieval of responses. Additionally, VICTR-E will support collaboration among COISTs though a workspace where analysts can create custom representations of the entities and complex relationships found in cultural data.

Charles River Analytics Inc.
625 Mount Auburn Street
Cambridge, MA 02138
Phone:
PI:
Topic#:
(617) 491-3474
Alexander Gee
OSD11-DR2      Awarded: 4/11/2012
Title:Collaborative Visual Exploitation and Reasoning Tool (COVERT)
Abstract:Company intelligence support teams (CoISTs) must successfully collect, analyze, report, and disseminate soft information from numerous heterogeneous sources through a variety of software applications, all while collaborating with other CoISTs and Battalion intelligence officers. To support these analysts, we propose to design and demonstrate the feasibility of a Collaborative Visual Exploitation and Reasoning Tool (COVERT). Four core components characterize our approach. First, we will enable CoIST analysts to exploit soft information by building COVERT on an extensible knowledge representation framework that supports the presentation and fusion of heterogeneous information sources through a single point of access. Second, we will enable the presentation, manipulation, and reasoning over multiple sources of information by providing analysts with an integrated view of the data through a set of advanced, human-centered display tools brought together within a tailorable information environment. Third, we will support collaboration between analysts by providing a work environment that supports real-time information sharing and manipulating via a collaborative visual analytics system. Fourth, we will assess the feasibility of our approach by designing and implementing a demonstration prototype in Phase I and defining an evaluation plan to more thoroughly assess our approach in Phase II.

Aptima, Inc.
12 Gill Street Suite 1400
Woburn, MA 01801
Phone:
PI:
Topic#:
(781) 496-2408
Cullen Jackson
OSD11-DR3      Awarded: 4/19/2012
Title:Real-time Analysis of Motivation and Behaviors for Operations (RAMBO)
Abstract:With the sharp rise in asymmetric warfare over the last decade, our military’s focus has shifted from a standard, kinetic military approach to much more complex missions, stressing counter-insurgency (COIN) and Stability, Security, Transition, and Reconstruction (SSTR) operations. These new types of missions have pervasive effects at strategic, operational, and tactical levels, and affect decisions at every echelon of command. In addition to the ongoing operations in Iraq and Afghanistan that have been responsible for this shift, the recent uprisings in the Middle East – referred to collectively as the Arab Spring – further illustrate the likelihood of military operations in countries, such as Libya, where pre-conflict intelligence gathering is minimal and rapid collection and integration of information across disparate social-cultural data sources is imperative. Operators require tools that will enable them to easily visualize and compare these different data sources and draw conclusions that are most appropriate to their specific information requirements. Aptima proposes to develop Real-time Analysis of Motivation and Behaviors for Operations (RAMBO). RAMBO is a visual analytics environment that supports planning, decision making, and analysis. RAMBO is an intelligent dashboard to help analysts visualize multi-source information, collaborate and coordinate, and link high-level and low-level data.

Intelligent Models, Inc.
9710 Traville Gateway Drive
Rockville, MD 20850
Phone:
PI:
Topic#:
(240) 401-9746
Yuri Levchuk
OSD11-DR3      Awarded: 6/14/2012
Title:Visual Representation and Implementation of Culturally Significant Information for Enhanced Tactical Decision-making
Abstract:The success of U.S. Army’s military and SSTR missions often hinges on the ability to understand HSCB dynamics of the theater of operations. More broadly, understanding of different socio-cultural terrains is the key to US national security. The costs of cultural ignorance or miscommunication--be it lost lives, major political mistakes and embarrassment, or lost opportunities and profits--are too large and too frequent to keep ignoring cultural differences. To promote rapid acquisition and effective use of relevant HSCB knowledge, we propose VISTA (Visual analytics for Intuitive Socio-cultural Terrain Awareness), which provides: (1) tool sets for visualization and analyses of HSCB data; (2) intuitive, clear, interactive representations of HSCB data; (3) automated analytics and translation of HSCB insights to decision-support products for different operational planning levels; and (4) evaluation of metrics to explain effects in relevant operational context. VISTA integrates two unique frameworks: (A) spatial-temporal visualization and navigation of latent dimensions of HSCB terrains; and (B) 4D data model of HSCB systems, scenarios, and interventions. It utilizes converging research findings from brain sciences, human psychology, psycho-linguistics, and advanced computing to operationalize the socio-cultural system-of-systems theory into clear, compact models of information processing, cognition, behaviors, and societal and trans-national effects.

Perceptronics Solutions, Inc.
3527 Beverly Glen Ter
Sherman Oaks, CA 91423
Phone:
PI:
Topic#:
(818) 788-4830
Amos Freedy
OSD11-DR3      Awarded: 6/20/2012
Title:Visual Representation and Implementation of Culturally Significant Information for Enhanced Tactical Decision-making
Abstract:This proposal is to develop a Cultural Information and Value Visualization System (CIVVS) that will provide new and powerful tools for analyzing, selecting, and visualizing critical cultural factors and applying them immediately in the context of military tactical planning. By integrating multi-dimensional visualization of culturally significant information with decision analytical modeling that imbeds cultural factors, CIVVS will enhance the quality and speed of tactical decision-making in foreign, culturally unknown and untried, and potentially hostile environments. CIVVS will support the visualization of tactically significant cultural information and values. It will do so by innovatively integrating and enhancing two sets of support systems previously developed Perceptronics: The first for cultural data acquisition and the second for tactical planning and decision making. Furthermore, CIVVS will include three visualizations that show how these information types interact with one another and with planning and decision making: (4) Visualization of a course of action as a region within a cultural space; (5) Visualization of cultural schemata or narratives can be visualized as causal networks; and (6) Visualization of uncertainty regarding the cultural impact of a candidate COA by representations in cultural value space. All six displays will be designed for both perceptual and cognitive usability.

Charles River Analytics Inc.
625 Mount Auburn Street
Cambridge, MA 02138
Phone:
PI:
Topic#:
(617) 491-3474
Michael Farry
OSD11-DR4      Awarded: 6/29/2012
Title:A System for Human-Rendered, Evidentiary Data Discovery and Entity Resolution (SHREDDER)
Abstract:Counter-insurgency (COIN) and irregular warfare (IW) missions have placed an emphasis on the role of tactical intelligence. Increasingly, tactical personnel such as Company-Level Intel Support Team (CoIST) members are responsible for assessing situations, providing insightful intelligence reports to higher echelons, and providing decision-relevant intelligence to their commander. The understaffing and undertraining of the CoIST results in two primary deficiencies: ineffective data gathering, and subsequent ineffective analysis based on the data sets that are available. To support the CoIST in performing more effective data gathering and analysis, we propose to design and demonstrate a System for Human- Rendered, Evidentiary Data Discovery and Entity Resolution (SHREDDER). Our solution is composed of two main elements. The first component is a Data Discovery Tool that employs collaboration support concepts from social media to enable CoIST personnel to identify the decision-relevant data scattered among databases and the mental models of other personnel. The second component is an Entity and Event Resolution Tool that enables personnel to collaborate visually to associate heterogenous and uncertain collected data in the intelligence analysis process. The structure provided by the Entity and Event Resolution Tool matches the unique needs of the CoIST, wherein a variety of information producers and consumers can each contribute dynamically.

Intelligent Models, Inc.
9710 Traville Gateway Drive
Rockville, MD 20850
Phone:
PI:
Topic#:
(240) 401-9746
Yuri Levchuk
OSD11-DR4      Awarded: 5/29/2012
Title:Building Semantic Knowledge of Large Data Sets through Collaborative Visual Approaches
Abstract:Human Social and Cultural and Behavior (HSCB) models are increasingly used to provide critical support to US military decision making. HSCB models are highly reliant on data; they need data from many sources, types, and areas of human behavior. The concomitant data streams have variable data quality and are constantly changing. Despite these challenges, HSCB applications may need near real-time access to relevant knowledge to make rapid decisions. To promote rapid acquisition and effective use of relevant HSCB knowledge, we propose SELMA (Semantic Exploration of Large Multi-modal Archives) which automates the processes of: (1) semantic cross-modal exploration of HSCB archives (to link people, events, objects, knowledge, and actions); (2) data-mining to fill gaps in information, resolve uncertainty, and classify behaviors and events; (3) mission-critical HSCB knowledge discovery; and (4) binding and visualizing the captured HSCB insights and semantic knowledge. Together, these components allow SELMA to offer a solution to HSCB data needs that: (1) collects, stores, and analyses mission-critical HSCB insights, (2) works autonomously for extended periods of time, and (3) actively reasons over the regional Human Terrains. SELMA dynamically creates an HSCB knowledge meta-network, explores concepts, discovers relationships with certain properties, and carries out versatile on-demand analyses.

Knowledge Based Systems, Inc.
1408 University Drive East
College Station, TX 77840
Phone:
PI:
Topic#:
(979) 260-5274
Perakath Benjamin
OSD11-DR4      Awarded: 4/3/2012
Title:Semantic Framework for Collaborative Analysis and Knowledge Exploration (CAKE)
Abstract:Knowledge Based Systems, Inc. (KBSI) proposes to design and demonstrate a semantic framework for Collaborative Analysis and Knowledge Exploration (CAKE). CAKE will provide the following innovative capabilities (1) Automation support for semantic tagging from multi-source text, image, and geospatial data; (2) Ontology-driven approach for semantic tagging and knowledge integration from text and image-derived tags; (3) Automation support for semantic knowledge discovery using sense making strategies, (4) Interactive visualization to support collaborative sense making and shared situational awareness . The research addresses critical technical voids that (a) limit the ability to exploit knowledge embedded in multi-modal data and (b) limit the ability of end users to make sense out of tagged text and image data in a manner that supports effective decision making. The Phase I effort will (i) establish requirements for collaborative sense making from tagged multi-modal data, (ii) formulate techniques for fusing tagged text data with tagged image data, (iii) design a method for collaborative sense making for enhanced situational awareness, and (iv) design and demonstrate CAKE technology. Phase II will build a focused military application of the innovation and rapidly transition the technology for large scale use.

21st Century Systems, Incorporated
6825 Pine Street Suite 141
Omaha, NE 68106
Phone:
PI:
Topic#:
(573) 458-5963
Robert Woodley
OSD11-DR5      Awarded: 5/17/2012
Title:Fusion with Uncertainty Reasoning using Nested Assessment Characterizer Elements (FURNACE)
Abstract:Situation/threat modeling and threat prediction require higher levels of data fusion to provide actionable information to the warfighter. Beyond sensor data and sources the analyst has control over, the use of out-sourced and re-sourced data are becoming more common. However, the current data fusion techniques do not account for the potential bias, ambiguity, and uncertainty in these other sources. Furthermore, despite the perception that the fusion process is sequential, work in human situational awareness has shown otherwise; a more nested approach is required. 21st Century Systems, Inc. proposes Fusion with Uncertainty Reasoning using Nested Assessment Characterizer Elements (FURNACE). FURNACE utilizes nested fusion loops building higher levels of information fusion without losing sight of the potential weaknesses of the underlying data. FURNACE uses advanced technologies in information filtering and reasoning to provide the levels of fusion. These reduce bias, disambiguate, and fill gaps in the data. FURNACE handles uncertainty through an innovative evidential reasoning technology that provides the necessary data to the analyst, such that they can account for the pedigree of the information supplied as it is aggregated and fused. And, with our excellent commercialization record, we are the company most likely to field this capability to the warfighter.

Aptima, Inc.
12 Gill Street Suite 1400
Woburn, MA 01801
Phone:
PI:
Topic#:
(781) 496-2404
Jennifer Roberts
OSD11-DR5      Awarded: 4/18/2012
Title:Towards Integration of Data for unBiased Intelligence and Trust (TID-BIT)
Abstract:To expedite intelligence collection process, analysts reuse previously collected data. This poses the risk of analysis failure, because these data are biased in ways that the analyst may not know. Thus, these data may be incomplete, inconsistent or incorrect, have structural gaps and limitations, or simply be too old to accurately represent the current state of the world. We propose the Towards Integration of Data for unBiased Intelligence and Trust (TID-BIT) system for characterizing the sources of error and bias specific to human-generated intelligence. TID-BIT will implement a novel Hierarchical Bayesian Model for high-level situation modeling, threat modeling, and threat prediction that allows the analyst to accurately reuse existing data collected for different intelligence requirements. By quantifying the reliability and credibility of human sources, TID-BIT will be able to estimate and account for uncertainty and bias that impact the high-level fusion process, resulting in improved situational awareness

Charles River Analytics Inc.
625 Mount Auburn Street
Cambridge, MA 02138
Phone:
PI:
Topic#:
(617) 491-3474
Avi Pfeffer
OSD11-DR5      Awarded: 4/18/2012
Title:Bias Formulation, Characterization, and Learning (BIFOCAL)
Abstract:A key challenge to higher level information fusion is fusing data originally collected for disparate tasks. Although much data may be available, data collected for a particular task may not be well suited to subsequent tasks because it embodies biases inherent in the original data collection process. A major challenge to applying data to a new purpose is understanding and correcting for biases in the data. To address this challenge, we propose a tool for Bias Formulation, Characterization and Learning (BIFOCAL). BIFOCAL uses deep learning to detect and characterize the biases in a dataset, thereby better enabling the data to be fused with other data to accomplish different purposes. Our approach is based on the idea that even biased data contains useful relationships, but we must have a deep understanding of the data. To accomplish this understanding, we will perform a systematic study of biases in complex and machine-generated data and use deep machine learning to build models of data sets and data generating processes as well as their target applications.

Data Fusion & Neural Networks, LLC
1643 Hemlock Wy
Broomfield, CO 80020
Phone:
PI:
Topic#:
(303) 469-9828
Christopher Bowman
OSD11-DR5      Awarded: 4/18/2012
Title:Innovative approaches to Situation Modeling, Threat Modeling and Threat Prediction
Abstract:The DF&NN team is composed of Christopher Bowman, Charles Morefield, Alan Steinberg, and Ed Waltz. We will develop methods to model and characterize the quality of data that has been ‘re-purposed’ for fusion applications. We will develop algorithms useful to High-Level Information Fusion (HLF), primarily the areas of situation modeling, threat modeling, and threat prediction. Our algorithms will specifically address bias and uncertainty when data sources include non-numeric qualitative measurements. Our focus is on methods that automatically learn to characterize such re-used/re-purposed data, thereby avoiding expensive off-line manually constructed data and model transformations. The HLF design will apply the Dual Node Network Data Fusion & Resource Management (DF&RM) technical architecture. We propose to cost effectively construct models of bias and error which will over time provide estimates of these errors. Technical objectives are: 1. Scenarios that expose appropriate HLF design and data uncertainty issues to fusion system development. Ontologies will include red/blue force strategies/tactics representing military/political Courses of Action 2. Multi-model approach to L2/3 HLF providing situation modeling, threat modeling, and threat prediction. 3. Develop algorithmic approaches for modeling re-purposed data, especially estimating biases and errors. 4. Deliver a software architectural design for HLF, using the DNN DF&RM technical architecture.

Charles River Analytics Inc.
625 Mount Auburn Street
Cambridge, MA 02138
Phone:
PI:
Topic#:
(617) 491-3474
Terry Patten
OSD11-DR6      Awarded: 4/30/2012
Title:Cloud Infrastructure for Retrieving and Refining Information from Sensors (CIRRIS)
Abstract:A wealth of sensor data is available to address the information needs of tactical Warfighters. However, this data resides in diverse, large, remote data stores. Delivering actionable information to the Warfighter requires discovering the relevant sources, retrieving relevant information, integrating the data so that it can be fused, and performing level one fusion. Since the preoccupied Warfighter cannot coordinate this workflow manually, it must be automatic. We propose a novel cloud infrastructure that orchestrates semantic discovery, retrieval, and fusion of information from diverse distributed sources. This effort addresses the technical challenge of applying map reduce processing to data that is external to the cloud and that is discovered (so the workflow required to integrate it cannot be known in advance). An innovative integration of semantic web service composition and map reduce processing is proposed. This combination of semantic technology and distributed processing will provide tactical Warfighters with the information they need, and provide it quickly.

DECISIVE ANALYTICS Corporation
1235 South Clark Street Suite 400
Arlington, VA 22202
Phone:
PI:
Topic#:
(703) 414-5015
Jonathan Day
OSD11-DR6      Awarded: 4/30/2012
Title:Discovering Valued Information in a Cloud Environment (DVICE)
Abstract:In combating terrorism, Warfighters must monitor at risk individuals and groups. The data sources needed to monitor such entities can consist of military sensors as well as open source literature. Key data types include unstructured text, audio, imagery, and biometric data. Currently, there does not exist a way to run specific searches in response to a tactical information need against large distributed data stores. DAC proposes a system called Discovering Valued Information in a Cloud Environment (DVICE). DVICE provides a framework to correlate, fuse, and exploit key observations which have been extracted from distributed data sources (text, imagery, and biometrics) in order to provide a consolidated and correlated knowledge product to the Warfighter. DVICE is based on the Hadoop cloud computing platform.

Intelligent Automation, Inc.
15400 Calhoun Drive Suite 400
Rockville, MD 20855
Phone:
PI:
Topic#:
(301) 294-4241
Onur Savas
OSD11-DR6      Awarded: 4/30/2012
Title:Discovery and Inform¬ation Retrieval from Distributed Multi-INT Data Sources in a Cloud Environment
Abstract:In recent years data intensive processing is beyond the capability of any individual machine and requires clusters, which means that massive data problems are fundamentally about organizing computations on dozens, hundreds, or even thousands of machines. This is exactly what MapReduce does, and the challenge is really how we can exploit MapReduce to bring valuable information to the Warfighter, for example, in combating terrorism to monitor at risk individuals and groups. In this proposal Intelligent Automation Inc. proposes a multi- layer workflow, initiated by Warfighter questions (plus metadata), which invokes other services, implemented as MapReduce processes in a Hadoop-based ecosystem. This workflow ends back at the Warfighter’s PDA, where ranked meaningful information is presented as a result of a data fusion process, with the system’s end goals being accuracy and speed. We also propose many improvements to the Hadoop kernel including dealing with bottleneck Map jobs, improvements for iterative (as opposed to parallel) tasks, and tandem parallel databases-Hadoop operations.

UtopiaCompression, Corporation
11150 W. Olympic Blvd. Suite 820
Los Angeles, CA 90064
Phone:
PI:
Topic#:
(310) 473-1500
Yatin Pandit Patil
OSD11-DR6      Awarded: 4/26/2012
Title:Cloud Based Critical Knowledge Retriever (CKR) for a PDA
Abstract:Effective information sharing services across the networked assets can provide a systemic benefit to DoD. The scale and complexity of DoD information enterprise and unavailability of user friendly access mechanisms makes optimal use of intelligence data difficult. UtopiaCompression proposes a cloud based Critical Knowledge Retriever (CKR) for PDA's. The system will use the capabilities of cloud computing to aggregate relevant information and provide it to the warfighter to help understand the surroundings and the people in it. Our proposed approach allows soldiers with portable devices to efficiently consult remote information services for critical information. Location based summarization will provide warfighters with multiple types of data, e.g., images, text, biometric information, and other unstructured data. UC's proposed architecture leverages well-known technologies such as Hadoop and FrameNet for data redundancy and improved Quality of Service (QoS). The Phase-I effort will include design and evaluation of the proposed CKR system for realistic scenarios.

ObjectVideo
11600 Sunrise Valley Drive Suite # 290
Reston, VA 20191
Phone:
PI:
Topic#:
(703) 654-9371
Zeeshan Rasheed
OSD11-DR7      Awarded: 4/30/2012
Title:Video Data to DDMS Cards
Abstract:In this SBIR proposal, ObjectVideo proposes an automated video content analysis system that ingests video and sensor data from a multitude of video sensors and produces meta information representing their contents in the form of modified DoD Discovery Metadata Specification (DDMS) cards. The human and machine-readable DDMS cards provide a compact representation of targets, their attributes, trajectories, and scene elements which are combined with the context of mission goals and specific needs of a warfighter in order to identify and retrieve video clips of most relevance. A GoogleEarth like map-based user interface allows interactive query input and result visualization to operators on desktop computers and handheld devices. The entire system is built upon a Service Oriented Architecture to ensure cross platform functionality, interoperability and scalability. To achieve these goals, ObjectVideo leverages its decade long experience of developing real-time intelligent video surveillance software and situational awareness systems. While the capabilities developed under this project will greatly enhance the automated dissemination of ISR video data, ObjectVideo envisions that they will also benefit its commercial software for image analysis and video event retrieval.

SEMANDEX NETWORKS, Inc
5 Independence Way Suite 309
Princeton, NJ 08540
Phone:
PI:
Topic#:
(609) 454-0448
Daniel Reininger
OSD11-DR7      Awarded: 4/30/2012
Title:Video Data to DDMS Cards
Abstract:The advances in Unmanned Vehicles (UAV) technology as well as in video technology have made it possible to record a large volume of video images from regions of interest. Cameras placed in UAVs enable to obtain more information for use by counterinsurgency operations. However, for this information to be integrated into the analysis system, the data that comes from these videos has to be translated into expressions that can semantically connect to priority and specific intelligence requirements. Currently, only a few expressions from the video populate the DoD Discovery Metadata Specification (DDMS) cards used to bring the information to the search and analysis system. Furthermore, creating these expressions requires in many cases that a human annotates frames on the video since there is no automatic process by which data from a video image automatically generates the metadata on the DDMS card. During Phase I, our team will investigate the feasibility of automatically generating the DDMS-structured metadata from video analytics and from additional contextual information for representative video files.

UtopiaCompression, Corporation
11150 W. Olympic Blvd. Suite 820
Los Angeles, CA 90064
Phone:
PI:
Topic#:
(310) 473-1500
Hieu Nguyen
OSD11-DR7      Awarded: 4/30/2012
Title:A Real-time Semantic Video Parsing and Tagging System
Abstract:Video sensors have undoubtedly become important sources of intelligence in the modern warfare to provide battlespace situational awareness. However, burgeoning numbers of battlespace sensors and the sheer volume of collected video data make it very difficult for battlefield decision makers to grasp a global picture of developing events as well as to detect specific events in real time. Video data need to be made available to users and systems across DoD through the Discovery Metadata Specification (DDMS) set as the common format for a wide-range of DoD data assets. UtopiaCompression (UC) proposes a novel real-time semantic video parsing and tagging (SAVANT) software system to help decision makers effectively process information from live streaming videos so as to accurately and timely respond to the development of battlespace events as well as conduct forensic analysis on archived video to satisfy warfighter priority and specific intelligent requirements. Innovative intelligent video analytics algorithms are employed to extract in real-time video metadata through automated object detection, recognition, context-based scene parsing and event recognition, and to generate DDMS compliant video content description representation. The proposed software will help to improve the successes rate of intelligence gathering missions as well as to reduce manpower requirements.

Advanced Cooling Technologies, Inc.
1046 New Holland Avenue
Lancaster, PA 17601
Phone:
PI:
Topic#:
(717) 295-6069
Anqi (Angie) Fan
OSD11-EP1      Awarded: 4/5/2012
Title:Electrical and Thermal Coupled Fast SiC Device Model for Circuit Simulations
Abstract:Silicon carbide (SiC) based devices are currently under investigation for their improved efficiency and high thermal stability. Compared to silicon, SiC has an exceptionally high thermal conductivity and also exhibits better mechanical properties. However, limited research has been performed on SiC device-level model development that can be used for developing circuit level designs. In many cases, thermal effects and other physics are neglected during circuit design because the current methods are too computationally expensive to be practical. In high power/high frequency devices, significant heat generation is caused by interactions between the energetic electrons and the lattice. The electrical behavior of the devices is therefore altered due to the coupling of electrical and thermal characteristics. Unrealistic simulation results are generated when the temperature change is not considered simultaneously. The proposed program will develop a hybrid of physical and tabular models for silicon carbide (SiC) high-powered devices for fast and accurate circuit simulation. The primary Phase I objective will be to demonstrate the feasibility of the proposed SiC-based device model and the ability of integrating the device model into fast circuit models.

CoolCAD Electronics
5000 College Avenue Suite 2103
College Park, MD 20740
Phone:
PI:
Topic#:
(301) 405-3363
Siddharth Potbhare
OSD11-EP1      Awarded: 6/6/2012
Title:Computer Aided Design Platform for Silicon Carbide Power Electronics
Abstract:Silicon Carbide (SiC) electronics has the potential for revolutionizing the high temperature high power electronics industry. There is a strong need for tools and models for circuit design using the new SiC power devices that are coming to market. Our work in this project will focus on developing analytical models for the newly commercially available SiC power MOSFETs that will then be used for design of efficient power converter circuits. We will extend our strong background work on developing complex device models for SiC power MOSFETs to analytical SPICE-type models that capture the unique physics of SiC devices, while at the same time can be used to simulate the electrical and thermal performance of a complex power converter circuit. The first phase of this project will focus on detailed measurements of the SiC power devices, developing SPICE models for the DC and transient behavior, and testing of key circuits for model verification and calibration. This effort will lead in to a more comprehensive Phase II work in which we will focus on coupled electro-thermal modeling and development of a SiC power system computer aided design (CAD) tool.

United Silicon Carbide, Inc
7 Deer Park Drive, Suite E
Monmouth Junction, NJ 08852
Phone:
PI:
Topic#:
(732) 355-0550
Petre Alexandrov
OSD11-EP1      Awarded: 4/18/2012
Title:Silicon Carbide Device Model Development for Circuit Simulations
Abstract:During this program, United Silicon Carbide, Inc. (USCi) will develop compact SiC power device models, prepare them for industry acceptance, and gain support of the models from commercial simulator suppliers. In Phase I, we propose to develop an initial compact circuit model for a 4H-SiC power MOSFET that will accurately describe device performance in the temperature range from -40°C to 250°C. The model will be based on a standard SPICE MOSFET model that will be expanded to include parameters reflecting the SiC device physics, thus allowing more accurate and robust electrical representation of the device in circuit simulation. In Phase II, we will fully develop the Phase I model to include electrical, thermal, mechanical, and material parameters that will allow 4H-SiC power MOSFETs to be used in system level modeling applications. We will also develop similar models for JFETs and Schottky diodes to be able to model different power system topologies. In Phase III, we will provide models commercially and achieve industry adoption through industry standards organizations (Compact Model Council and Global Semiconductor Alliance) and collaboration with commercial system vendors.

Luna Innovations Incorporated
1 Riverside Circle Suite 400
Roanoke, VA 24016
Phone:
PI:
Topic#:
(540) 769-8423
Gustave Anderson
OSD11-IA1      Awarded: 4/19/2012
Title:Enterprise Anti-exploitation for Mission Assurance
Abstract:Supply chain threats have invalidated the assumption that the adversary has no reasonable avenue of attack even if the software protections are structured properly and augmented with secure hardware. In today’s cyber-attack environment, one must assume that a subset of systems are or will be eventually compromised. Luna proposes a fault tolerance, federated cloud computing architecture supporting design separation for high reliability and information assurance. By leveraging a hybrid fault model with multiple, parallel execution paths and resultant execution trace comparison, Luna has created an architecture that identifies suspect nodes and assures trusted execution. Further, the model architecture can be scaled through proactive thread diversity for additional assurance during threat escalation. The solution provides dynamic protection through distributing critical information across federated cloud resources that adopt a metamorphic topology, redundant execution, and the ability to break command and control of malicious agents.

Siege Technologies, LLC
33 South Commercial Street
Manchester, NH 03101
Phone:
PI:
Topic#:
(315) 215-1018
Joseph Sharkey
OSD11-IA1      Awarded: 4/25/2012
Title:Anti-Exploitation Software Protection Systems
Abstract:State-of-the-art software protection and anti-tamper systems move critical software and data “out-of-band” to the adversary, by using a hypervisor or on “secure” hardware. Unfortunately, the systems running this software are built using untrusted commercial-off-the-shelf (COTS) parts. Supply chain threats to critical components, such as hardware or firmware Trojans, have invalidated the assumption that we can move our critical software and data completely “out-of-band” to the adversary, since the hardware components on which the software ultimately executes is untrusted. As a result, one must re-think the fundamental approach to building software protection and anti-tamper systems. Siege Technologies proposes a methodology and system to address these concerns on a COTS system, or a system composed of COTS parts. Our solution is specifically designed to augment existing software protection mechanisms, shielding them from attack by malicious hardware alterations. The research and systems design conducted in Phase I will shed new light on the problem space and establish new protection capabilities that guard against this threat and enable the system to run protected code in the presence of potentially compromised hardware.

Charles River Analytics Inc.
625 Mount Auburn Street
Cambridge, MA 02138
Phone:
PI:
Topic#:
(617) 491-3474
Terry Patten
OSD11-IA2      Awarded: 5/8/2012
Title:Adaptive Deception using Linguistic Inference of Behaviors (AD-LIB)
Abstract:As cyber attacks become increasingly and more effectively targeted, so must our cyber defenses become less “one size fits all.” We need to develop active defenses that can adapt themselves to an attack in progress. This development will require not only improvements in our detection capabilities but also techniques to engage and out-game attackers. If we can develop robust models of attacks and attacker behaviors, we can then infer an attacker’s intentions and likely next moves. Most importantly, we can use this knowledge to our own advantage though use of targeted deceptions—the more we can lead an attacker down paths of our own making, the more we can delay and deter him, and the more we can gain valuable insight into his mission—what he believes he is accomplishing and why. We propose to research, design, and demonstrate a strategy and architecture for Adaptive Deception using Linguistic Inference of Behaviors (AD-LIB). AD-LIB uses techniques from natural language processing (NLP) and specifically systemic functional grammar (SFG) to generate targeted software deceptions to coax intent information from cyber attacks and attackers based on real-time understanding of their goals and most-probable next moves.

Galois, Inc.
421 SW Sixth Suite 300
Portland, OR 97204
Phone:
PI:
Topic#:
(503) 808-7216
Adam Wick
OSD11-IA2      Awarded: 7/3/2012
Title:Deceiving the Deceivers: Active Counterdeception for Software Protection
Abstract:DoD operations and infrastructure increasingly depends on software, which makes it an attractive target for our adversaries. Not surprisingly, deception plays a central role in most cyberattacks. To better protect these critical systems, we propose to design and build an “active counterdeception” software protection system which we call CYCHAIR that both incorporates the right sensors, and enables the right mind-set for its operators. CYCHAIR consists of two complimentary technologies. The first piece the ability to easily generate large numbers of reusable, extensible and highly reconfigurable decoys. These decoys serve multiple purposes: first of all, they serve to increase the adversary’s workload while confusing them as to the manner and location of the real targets. Secondly, they serve as intelligence gatherers, recording all the adversarial interactions. These records are fed to the second piece of the system, an inference engine we call LAIR (Logic for Adversarial Inference and Response). These inferences can be used to automatically trigger dynamic reconfiguration of the decoys (to further frustrate and slow down the adversary), and used as recommendations to the human-in-the-loop for additional active responses to the attack.

Sentar, Inc.
315 Wynn Drive Suite 1
Huntsville, AL 35805
Phone:
PI:
Topic#:
(256) 430-0860
Al Underbrink
OSD11-IA2      Awarded: 6/4/2012
Title:Software Deception as a Countermeasure to Attacks on Software Protection Systems
Abstract:An advanced strategy and architecture for software deception is proposed. The proposed research and development leverages a DoD SBIR Phase I contract awarded in 2009. The original project performed a feasibility assessment and early design of an architecture for providing software deception. Leveraging the prior work provides an opportunity to reach deeper into the concepts and challenges in using deception for software protection. Extensions of the previous work focus on the control of a distributed software protection system on an end node. Further development of the software deception concept employs a game theory approach for experimenting with, and implementing, deception strategies. The extensions to the architecture and infrastructure enable more sophisticated deceptions which can be used to promote a greater understanding of an adversary’s strategies and tactics. This understanding of adversaries in turn enables the development of more adaptive software protection systems.

Smart Information Flow Technologies, d/b/a SIFT
211 N 1st St. Suite 300
Minneapolis, MN 55401
Phone:
PI:
Topic#:
(612) 789-0559
Dan Thomsen
OSD11-IA2      Awarded: 4/26/2012
Title:ADEPT: Advanced Deception Enhancing Protection Technology
Abstract:The proposed ADEPT technology combines the science of cognitive psychology with centuries of magicians' practical performance experience to create a new science of deception for computer security. The ADEPT approach learns about the motives and real-world attributes of people by analyzing their cognitive processes based on their reactions to specifically designed stimuli. The resulting technology forms an additional protection layer that defenders can quickly integrate with any protection system. The ADEPT technology identifies adversarial intent and thought processes, finds real-world attributes to identify the adversary, delays the adversary, and gives defenders information that they can use to increase security. ADEPT represents a new direction in proactive software protection that learns about the adversary executing an attack and adapts its defenses accordingly.

21st Century Technologies Inc.
6011 West Courtyard Drive Bldg 5, Suite 300
Austin, TX 78730
Phone:
PI:
Topic#:
(512) 682-4719
Igor Frolow
OSD11-IA3      Awarded: 4/12/2012
Title:Identification of Critical Resources and Cyber Threats in the Physical Domain
Abstract:Military organizations, civilian government agencies, and private companies are all undergoing an unprecedented growth in cyber attacks targeting their information infrastructures and causing serious disruptions to their operations. To date, much attention has been paid to determining ways to defend the information networks and preserve information flow. Clearly, there is an insufficient amount of resources to provide this protection and so it becomes necessary to determine the criticality of the network assets and activities to preserving the mission, even in the face of cyber and/or physical attacks. To address this problem, 21CT is proposing the Asset Threat Analysis Capability (ATAC), a system dynamics modeling approach that will represent the cyber and physical networks used to accomplish one or more military missions. The model will consist of the key assets and activities and the interactions needed to satisfy military objectives. It will give military planners a capability to analyze these assets and activities using the ATAC model as a “what if” tool to determine the effects of disrupting a particular activity on one or more military missions.

EDAptive Computing, Inc.
1245 Lyons Road Building G
Dayton, OH 45458
Phone:
PI:
Topic#:
(937) 281-0792
Adam Langdon
OSD11-IA3      Awarded: 5/14/2012
Title:Identification of Critical Resources and Cyber Threats in the Physical Domain
Abstract:With the increase in net-enabled warfare, network security will continue to be of critical importance to the DoD. Efforts to protect against intrusions have resulted in effective detection systems that can aid in network protection. However, another type of serious threat is the attack on the physical infrastructures that provide these critical network capabilities. One of the key challenges in addressing this problem is identifying the physical infrastructure on which a virtual network depends. Characterizing these elements and mapping them to cyberspace assets cannot be done solely through existing network tools. OSD is seeking techniques to analyze this data and automatically identify vulnerabilities, points of failure, or other critical risk areas. In response to this need, Edaptive Computing, Inc. (ECI), proposes DECYPHER, a research and development effort that will result in a unique and commercially viable innovation to assessing cyber vulnerabilities from physical infrastructure analysis. Our proposal specifically addresses the stated requirements of the solicitation; we will develop and deploy tools, methods, and models for detecting threats to cyber systems through the physical domain. The resulting capabilities will result in new techniques for analyzing and visualizing physical risks and their impacts on critical systems.

Knowledge Based Systems, Inc.
1408 University Drive East
College Station, TX 77840
Phone:
PI:
Topic#:
(979) 260-5274
Perakath Benjamin
OSD11-IA3      Awarded: 4/24/2012
Title:Cyber-to-Physical Domain Mapping Toolkit for Vulnerability Analysis and Critical Resource Identification Enablement (CEPHEID VARIABLE)
Abstract:Knowledge Based Systems, Inc. (KBSI) proposes to develop a Cyber-to-Physical Domain Mapping Toolkit for Vulnerability Analysis and Critical Resource Identification Enablement (CEPHEID VARIABLE) that enables military analysts and enterprise infrastructure management stakeholders to perform vulnerability assessment of cyber-physical systems. CEPHEID VARIABLE is an application framework that enables the acquisition, representation, storage, mapping, vulnerability and dependency analysis of information linking cyber and physical resources, and supports both static and dynamic vulnerability analysis.

The Design Knowledge Company
3100 Presidential Dr Suite 103
Fairborn, OH 45324
Phone:
PI:
Topic#:
(937) 427-4276
Jeff Walrath
OSD11-IA3      Awarded: 5/7/2012
Title:INCITE: Innovative Cyber/Infrastructure Threat assessment Environment
Abstract:TDKC proposes the INnovative Cyber/Infrastructure Threat Assessment Environment (INCITE) as part of this SBIR call. Our Phase I concept includes: (1) the collection and processing of publically available infrastructure data to include roadways, gas pipelines, electrical grids, water lines, telecommunication lines, and other data (as available); (2) the prototyping of automated clustering, filtering data attribution, and ontology tagging tools to translate raw infrastructure data into a common malleable format; (3) the prototyping of analyst data tools that process the attributed infrastructure data to highlight potential areas of concern; and (4) the integration of all capabilities into an innovative mashup-like decision support tool. Our approach ensures an operationally focused Phase I concept that is effective, demonstrable, and most of all operationally practical. Our goal with INCITE is to automate critical time-consuming tasks that are well-defined elements of the overall infrastructure analysis workflow. INCITE is an open-source, standards-based framework that integrates with existing OSD/AF resources and environments.

3 Sigma Research, Inc.
503 S. River Oaks Dr.
Indialantic, FL 32903
Phone:
PI:
Topic#:
(321) 674-9267
Michael Tulloch
OSD11-IA4      Awarded: 5/18/2012
Title:Cyber Security High Abstraction Contextual Visualization and Decision Support System
Abstract:3 Sigma Research investigates and develops an advanced innovative Tangible User Interface (TUI) that provides a novel operator interface for real-time operator manipulation of the cyber environment. This TUI provides a highly abstract physical embodiment of the complex cyber environment, that commanders and controllers can touch, visualize, and recognize patterns. This kind of manipulation of objects within the TUI enables a new paradigm of on-the-fly control of autonomic systems. One of the key advantages of our envisioned system is that by mapping the cyber domain into a tactile and physical representation, humans can effectively and efficiently apply human pattern recognition to the analysis and response to cyber attacks never before possible. The use of pattern recognition to solve problems has sometimes been called “intuitive” or reactive decision making. Intuitive decision making draws heavily on pattern recognition derived from previous experience. Thus the mapping of the domain into the representational space is critical to both aid comprehension and support application of prior experience. Reading is like telling; graphical displays are like showing; our approach is involving. Our approach allows operators to make more accurate, faster, and more trustworthy decisions.

Architecture Technology Corporation
9977 Valley View Road
Eden Prairie, MN 55344
Phone:
PI:
Topic#:
(952) 829-5864
Kenneth J. Mosher
OSD11-IA4      Awarded: 4/19/2012
Title:Decision Assistance, Visualization, and Intuitive Natural Control Interface (DAVINCI)
Abstract:Architecture Technology Corporation (ATC) proposes the Decision Assistance, Visualization, and Intuitive Natural Control Interface (DAVINCI), a system that provides decision assistance through a multi-level topic-mapped knowledge base, abstract visualization of situational and metric data, and high-level abstract control of manually initiated and autonomic response protocols through a tangible user interface that utilizes physical control tokens in combination with multi-touch gestural input. In situations where information rapidly shifts context (such as during net-enabled spectrum warfare, or cyber security analytic operations), the metrics used to evaluate the situation are highly dynamic, to the point that existing technologies do not provide an optimum solution for presenting a visualization of the relevant metrics and situation evaluation data to the user. DAVINCI uses a topic-mapped knowledge base to process sensor inputs, determine the current situation, select only the metrics and response protocols relevant to that situation, and present that data to the user in a clear, concise manner. DAVINCI then allows the user to manage autonomic and manually-initiated response protocols using a tangible user interface wherein the user manipulates physical tokens in conjunction with multi-touch gestures on a touch-sensitive display and control device. These visualization and control capabilities can be customized and expanded via plug-in modules that implement the DAVINCI API. The DAVINCI solution is particularly suited to the cyberspace security and net-enabled spectrum warfare domains, but as a customizable software framework solution, DAVINCI can be applied to any problem domain where reduction of data unnecessary to the given situation and abstracted visualization and control capabilities are desired.

Charles River Analytics Inc.
625 Mount Auburn Street
Cambridge, MA 02138
Phone:
PI:
Topic#:
(617) 491-3474
Ryan Kilgore
OSD11-IA4      Awarded: 5/22/2012
Title:Tangible trustworthiness for mixed-initiative network defense (T2-MIND)
Abstract:Cyber defenders must interact with a variety of diverse feeds and explore network information in many abstract contexts. Unfortunately, existing digital tools provide poor affordances for the perception and manipulation of critical metrics of network trustworthiness, unnecessarily complicating the network defense task and increasing cognitive workload. To address these challenges, we propose to design, demonstrate, and evaluate a system for Tangible Trustworthiness for Mixed-Initiative Network Defense (T2-MIND). T2-MIND will improve the efficiency and efficacy of network security by providing novel, tangible controls and interfaces that merge the intuitive affordances of physical control objects with the power of modern digital systems. Three main components characterize our approach. First, we will use demonstrated cognitive systems engineering methods to identify metrics of trustworthiness that are useful to cyber defenders in a dynamic, adversarial network security environment. Second, we will capture these metrics through tangible user interface controls that extend the physical affordances of security tools and provide high-visual-momentum displays that reduce the cognitive workload of cyber defenders moving between abstract representations of network status. Third, we will use an iterative design approach to develop prototypes that capture the underlying cognitive work, analytical tasks, and workflows specific to cyber defenders.

Knowledge Based Systems, Inc.
1408 University Drive East
College Station, TX 77840
Phone:
PI:
Topic#:
(979) 260-5274
Ronald Fernandes
OSD11-IA4      Awarded: 5/18/2012
Title:Advanced Visualizations and User Interfaces for Cyber Situational Awareness (ADVICE)
Abstract:Knowledge Based Systems, Inc. (KBSI), along with their research partner SA Technologies, Inc. proposes to design, assemble and demonstrate an innovative experimental framework to develop capability for Advanced Visualizations and User Interfaces for Cyber Situational Awareness (ADVICE). The ADVICE framework is a suite of tools in an extensible framework that support the use of newer computer-human interaction technologies and newer visualization paradigms to maximize the situational awareness (SA) of the cyber system that is being monitored. ADVICE will contain interface abstractions that allow for new input paradigms such as Tangible User Interfaces (TUIs) and Tactile Output Interfaces (TOIs) that will be powered by relevant SA-oriented design (SAOD) principles that include SA requirements analysis such as Goal-Directed Task Analysis (GDTA). Additionally, ADVICE will allow a cyber SA designer to derive measurements that yield metrics relating to the computer-human experience. As a result, the cyber SA designer can ascertain which user interfaces are more effective under different situations.

Broadata Communications, Inc.
2545 W. 237th Street, Suite K
Torrance, CA 90505
Phone:
PI:
Topic#:
(310) 530-1416
Dean Mumme
OSD11-IA5      Awarded: 5/7/2012
Title:Program Instruction Sequence Monitor for Hijack Detection and Proactive Zero-day Defense
Abstract:Numerous malicious means have been developed for the hijacking software program execution to gain unauthorized access and functionality on computer systems and network nodes. The techniques are many and varied, but generally fall into the framework of placing a pointer to an execution location into a program’s normative execution path, in order to cause a jump to the targeted location for subsequent execution. To address this critical need, Broadata Communications, Inc. (BCI) proposes the Program Instruction Sequence Monitor (PRISM); a robust, accurate, and efficient mechanism of deterministically detecting program hijacking. PRISM provides hijack detection that is out- of-band to attack. It does not require virtualization technology but can (optionally) leverage virtualization it to enhance robustness against attack. The key advantages of PRISM include: (1) Insensitive to replay attacks, (2) Efficient performance in comparison with dynamic tainting, (3) No need for source code, emulation, or instrumentation of the executable, (4) Works even in the presence of an executable stack, or writable code area, (5) Monitoring mechanisms are robust and out-of-band to attackers in user space. PRISM can respond to zero-day attacks, and can greatly enhance existing military, governmental, and industrial cyber-security systems.

EDAptive Computing, Inc.
1245 Lyons Road Building G
Dayton, OH 45458
Phone:
PI:
Topic#:
(937) 281-0782
Nicholas Kovach
OSD11-IA5      Awarded: 5/6/2012
Title:Deterministic Detection for Hijacked Program Execution
Abstract:Our proposal specifically addresses the stated requirements of the solicitation; we will develop software and system protection tools for monitoring the execution of untrusted software in large complex networks. This includes attacks which may be present in software applications used within the DoD (e.g. buffer overflow, code injection, rootkit patching, etc.). The proposed EDAptive® DeforcE solution builds on previous Edaptive Computing, Inc. (ECI) knowledge and technology – itself innovative – to provide detection and reporting of software based attacks. The resulting capabilities will result in innovative compiler technology, along with a software monitor, to determine if anomalous behavior (i.e. an attack) has occurred in the execution of the application.

GrammaTech, Inc
531 Esty Street
Ithaca, NY 14850
Phone:
PI:
Topic#:
(607) 273-7340
Thomas Johnson
OSD11-IA5      Awarded: 5/14/2012
Title:Deterministic Detection for Hijacked Program Execution
Abstract:We propose technology to automatically detect software exploits that hijack the execution of a running Linux kernel or one of its hosted applications. The key enabling technology for the proposed work is provided by KATE, the Kernel Analysis and Translation Engine. KATE uses a combination of software dynamic translation and hardware virtualization to provide fine- grained, efficient monitoring of an entire system. We propose to use KATE to enforce execution invariants on a running system. Violations of the invariants will indicate that the system is under attack.

Zephyr Software LLC
2040 Tremont Rd
Charlottesville, VA 22911
Phone:
PI:
Topic#:
(434) 284-3002
Clark L. Coleman
OSD11-IA5      Awarded: 5/6/2012
Title:Preventing Program Hijacking via Static and Dynamic Analyses
Abstract:Control flow hijacking occurs when an attacker overwrites a control-flow data item (e.g. return address or function pointer) to take control of the execution of a program. We propose to detect and prevent hijacking by using a low-overhead per-process dynamic run-time virtualization monitor, called an SDT (software dynamic translator) to make shadow copies of control-flow data items each time they are initialized or updated, and detect overwriting changes that occur between initialization and use. A static analyzer that operates on program binaries will help identify all control-data items, and reduce run-time overhead by identifying control-data items that are provably safe (not susceptible to overwriting between initialization and use). Remedial actions to be taken when attempted hijacking is detected will not be limited to program termination; program recovery techniques will be studied and designed.

21st Century Technologies Inc.
6011 West Courtyard Drive Bldg 5, Suite 300
Austin, TX 78730
Phone:
PI:
Topic#:
(512) 682-4730
Jonathan Mugan
OSD11-IA6      Awarded: 4/19/2012
Title:Active Software Defense to Reduce Threat Capability Effectiveness
Abstract:The protection of cyber assets is a critical need for the U.S. military. The theft of sensitive software and data threatens our technological and information superiority, putting lives and valuable assets at unnecessary risk. Attacks occur too quickly for human intervention and current automated defenses are designed to thwart only known means of attack. Active software defenses are required that can recognize a broad spectrum of threats and respond intelligently. To address this need, 21st Century Technologies proposes Cyber Security using Qualitative Learner of Action and Perception (Cy-QLAP), a software system that develops an understanding of the protected cyber assets using machine learning methods inspired by early mental development in humans. Past automated cyber defense systems have been plagued by availability issues, where disruptions to normal activity cannot be tolerated because system administrators are unable to analyze and correct the system’s behavior. Cy- QLAP solves this problem by learning human-understandable models that enable meaningful user interaction and feedback, allowing the benefits of quick action while maintaining the confidence of system administrators that availability problems can be managed. Our solution has the potential to significantly reduce the overall cost of protecting computer networks, and minimize the number of successful attacks on networks.

Power Fingerprinting, Inc.
2200 Kraft Drive, Suite 1200 R
Blacksburg, VA 24060
Phone:
PI:
Topic#:
(540) 200-8344
Carlos R. Aguayo Gonzalez
OSD11-IA6      Awarded: 5/1/2012
Title:Active Software Defense to Reduce Threat Capability Effectiveness
Abstract:Perimeter and passive cyber defenses must be complemented with an active defense mechanism to elevate the risks, or costs, a potential attacker must face. The effectiveness of an active defense mechanism is ultimately limited by its ability to detect threats fast, accurately, and reliably. For this Phase I project, we propose to determine the feasibility of creating an active defense solution based on a novel integrity assessment approach called power fingerprinting (PFP). PFP is capable of detecting integrity violations and malicious intrusions with extreme accuracy and speed, even at the lowest levels of the software stack. The proposed solution includes effective mechanisms to gather intelligence, deny access to peripherals, and even reset the processor to prevent an attacker from stealing information or install a back door. For Phase I, we will develop a PFP monitor for an embedded Android platform, validate the different active responses and countermeasures, and develop a general mechanism to interpret response policies. We will evaluate the performance of the complete system in terms of probability of threat detection, probability of false alarm, and response effectiveness in different attack scenarios and blind tests.

Private Machines Inc.
PO Box 4056
Stony Brook, NY 11790
Phone:
PI:
Topic#:
(888) 791-5147
Radu Sion
OSD11-IA6      Awarded: 4/30/2012
Title:ARMOR: Active-Defense Resilient Mission-Oriented Secure Cloud Platform
Abstract:The Active-Defense Resilient Mission-Oriented (ARMOR) Secure cloud platform provides significant competitive advantages in the defense against vulnerabilities and threats in distributed and cloud computing infrastructures. It achieves this by endowing a traditional infrastructure with a collaborative security layer of active transparent defense and control, a cloud “immune system” which transparently introspects, monitors and controls the traditional cloud servers at runtime and can strongly react and immediately present a united defense across the entire cloud. The immune system is comprised of isolated (i.e., out-of-band) hardware which has its own control network. This separate conduit enables it to instantly share critical information and react independently enabling meaningful operation and fast recovery even in the case of sustained attacks and severe compromise. Ultimately, ARMOR technology significantly reduces the effectiveness of security threats in infrastructures running critical applications, through active defense mechanisms such as real-time monitoring of host activity, transparent introspective control and recovery of compromised hosts, and immediate reactivity in cross-cloud collaborative network and host- level defense. The physical protection assurances of ARMOR allows for safe deployment with full privacy in untrusted environments such as battle-field and commercial hosting scenarios. DOD can now outsource the infrastructure itself, thereby maximizing operational efficiency and capabilities.