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

88 Phase I Selections from the 03.2 Solicitation

(In Topic Number Order)
EDAPTIVE COMPUTING, INC.
1245 Lyons Road, Building G
Dayton, OH 45458
Phone:
PI:
Topic#:
(937) 433-0477
Mr. David L. Barton
OSD 03-021       Selected for Award
Title:Specification and Test Based Reverse Engineering of Low Level Code System (EDAMarksman)
Abstract:There is an urgent need in the military to upgrade and re-implement low-level software embedded in operational weapon systems, despite the millions of lines of often undocumented code. EDAptive Computing, Inc. (ECI) presents an innovative solution to the problem of creating specifications from low level code, such as assembly and CMS-2. Our Specification and Test Based Reverse Engineering of Low Level Code System (EDAMarksman) program will apply a combination of formal specification, reverse engineering (compilation), and test generation to provide a specification-based system for testing. Specifically, we propose to utilize an iterative approach to 1) derive specifications from existing code using formal reverse engineering techniques, 2) derive tests from the specifications, 3) apply the tests to the original code and receive feedback, and 4) update the specifications based on test results. Tests produced from specifications will be used to validate the specifications, indicating where tests fail on the actual system, and guide the production of new specifications. The result of this solution will be a specification creation process that will be verified by the actual execution of the system, as guided by the generated specification. We will use a real-life Navy/DoD system as a candidate Phase I demonstration vehicle.

GRAMMATECH, INC.
317 N. Aurora Street
Ithaca, NY 14850
Phone:
PI:
Topic#:
(607) 273-7340
Dr. David Melski
OSD 03-021       Selected for Award
Title:Modernization of Legacy Software: Re-engineering Low-Level Code
Abstract:As hardware platforms age, manufacturers are less willing to support them and related technologies. Legacy weapons systems must eventually be modernized. Re-implementation of the legacy software by hand is prohibitively expensive. A system is needed for automatically translating software written in obsolete languages like CMS-2 or assembly to a modern language like C. Previous attempts at translation have amounted to transliteration: the produced "high-level" code closely mimicks the low-level code, even using global variables for registers. Translators that do more than this often assume that the low-level code was generated by a compiler, and hence adheres to certain conventions; such assumptions do not hold for legacy code. GrammaTech is a leader in developing technology for the static analysis of software binaries. Furthermore, we are collaborating with leading researchers at the University of Wisconsin for in developing analyses for discovering malicious and/or buggy code in binaries. We will leverage this background to create breakthrough translation technology that: (1) produces high quality code; and (2) succeeds at translation where others fail. Our translator will be integrated with Northrop Grumman's state-of-the-art emulation technology, for those cases where translation is not possible.

INTERACTIVE SOFTWARE ENGINEERING, INC.
356 Storke Rd.
Goleta, CA 93117
Phone:
PI:
Topic#:
(805) 685-1006
Mr. Emmanuel Stapf
OSD 03-021       Selected for Award
Title:Extracting contracts for the safe reengineering of legacy software
Abstract:Work on reengineering existing software has largely consisted of parsing existing code (machine of assembly), reconstructing the control flow and data flow, and providing hopefully equivalent behavior in a modern language. We propose a development complementing current approaches by focusing on the high-level picture: the *contracts* behind software elements, expressing their specifications, according to the widely influential methodology of Design by Contract developed by our company. The approach, applied to legacy software units, will yield:  Key abstractions of the software, leading to a modular structure conforming to modern views of modularity, in particular data abstraction and object technology.  For each software element, contracts expressing the purpose behind the implementation. Such a form of reengineering is intended to yield a form of the software that has strong modularity and precise specifications, and so is ready for further evolution and reuse, on a par with more recently written program elements benefitting from the full power of modern programming techniques.

MODELLION SYSTEMS, INC.
1335 Woodcliffe Dr
Monroeville, PA 15146
Phone:
PI:
Topic#:
(412) 372-6933
Dr. Raj Rajkumar
OSD 03-021       Selected for Award
Title:Modernization of Legacy Software: Re-engineering Low-Level Code
Abstract:We will design and prototype a tool to translate legacy real-time programs to modern languages running on modern real-time operating systems. Our technology will convert multi-tasking languages such as Ada 83 and Ada 95, high-level programming languages such as Jovial, and low-level assembly languages such as CMS-2. Legacy software will be first translated into a common intermediate format and then be translated into desired target languages such as C, C++ or Java, and even into reusable component models. Our tool shall automatically extract and model the concurrency and timing characteristics of legacy software. Converted software can therefore be directly subject to schedulability analysis and extensibility. Companion tools will enable independent functional and timing validation of the translated software.

SEMANTIC DESIGNS
12636 Research Blvd #C214
Austin, TX 78759
Phone:
PI:
Topic#:
(512) 250-1018
Dr. Michael Mehlich
OSD 03-021       Selected for Award
Title:Modernization of Legacy Software: Re-engineering Low-Level Code
Abstract:The proposal describes a semi-automatic tool translating source code implemented in mixed low-level programming languages for obsolete hardware platforms into more easily understandable source code implemented in modern high-level programming languages for modern hardware platforms by (1) parsing the textual representation of the source code into a syntax tree, (2) translating the syntax tree into a control- and data-flow graph, (3) removing accidential implementation details from the control- and data-flow representation, (4) recognizing concepts, which are defined by software engineers using the notations of the involved programming languages, in the resulting control- and data-flow graph by performing graph matching, (5) translating the source syntax tree into a target syntax tree by abstracting the recognized concepts, reimplementing them in the target language, and translating the remainder of the source code using straightforward syntactic transformation rules, and (6) formatting the target syntax tree into a textural reprentation. The proposed phase I work consists of initial research laying the foundations to develop such an automated translation tool. It will build upon an existing program analysis and transformation system infrastructure.

CODESOURCERY, LLC
9978 Granite Point Ct.
Granite Bay, CA 95746
Phone:
PI:
Topic#:
(916) 791-8304
Mr. Mark Mitchell
OSD 03-022       Awarded: 04DEC03
Title:High Performance Object Oriented Software for Parallel Embedded Systems
Abstract:The hardware used to perform vector, signal, and image-processing computations is becoming increasing parallel. This parallelism is present both at the low level (in the form of Single Instruction Multiple Data instructions in the microprocessors) and at the high level (in the form of multiple processors in a single machine.) We suggest making modifications and enhancements to VSIPL++ (a highly abstract C++ library being developed by the DoD) to take advantage of both forms of parallelism. Our objective is to demonstrate that VSIPL++ can run as fast as VSIPL, despite its heavy use of high-level abstractions and object-oriented technologies.

COHERENT LOGIX, INC.
101 West 6th Street, Suite 200
Austin, TX 78701
Phone:
PI:
Topic#:
(512) 479-7732
Mr. Jan Garmany
OSD 03-022       Awarded: 04DEC03
Title:Self-Optimizing Parallel Implementation of the VSIPL++ Standard (SOI)
Abstract:Coherent Logix, Incorporated (CLX) proposes to develop a self-optimizing parallel implementation of the VSIPL++ specification. This implementation will include a configuration tool that automatically extracts critical performance parameters from the target system and a template-metaprogrammed C++ implementation that automatically tunes the VSIPL++ software based upon the values of these parameters. This combination will enable both highly-efficient and highly-portable implementations of VSIPL++. During Phase I of the program, CLX will target a subset of the standard, provide a prototype implementation for this subset and will evaluate the appropriate parallel performance models and metrics to be used in a complete implementation. Phase II will include a full implementation and supplemental tools to further enhance the performance of Signal and Image Processing (SIP) applications.

MPI SOFTWARE TECHNOLOGY, INC.
101 S. Lafayette #33
Starkville, MS 39759
Phone:
PI:
Topic#:
(662) 320-4300
Mr. Brian Chase
OSD 03-022       Awarded: 10DEC03
Title:High Performance Object Oriented Software for Parallel Embedded Systems
Abstract:This Phase 1 SBIR Proposal entitled "Object Oriented Techniques for Parallel Embedded Computing" offers an examination of the problems associated with Object Oriented Technology and High Performance libraries or frameworks. An attempt is made to show that object oriented technologies need not be discounted when developers look for tools to solve problems requiring high levels of performance with low resource utilization overheads. Under this solicitation, Proposer will examine existing facilities such as POOMA, PETE and related technologies and the methods by which they work around performance problems that are inherent in C++. Proposer will identify techniques that yield the desired outcomes of both elegant expression and efficient performance of object oriented code. Proposer will obtain libraries such as PETE, POOMA, BLITZ, and FACT! for the purpose of evaluating their effectiveness in the distributed, embedded arena. Proposer will study and contrast the design of these tools in comparison to Proposer's own C++ VSIPL engine. Consultations with customers such as Lockheed Martin will yield test cases that can be used to make an appraisal of the various technologies. In conclusion, Proposer will make recommendations regarding which sets of tools and methodologies will best optimize VSIPL functionality and provide performance predictive capabilities for embedded systems.

PENTUM GROUP, INC.
846 Hollenbeck Ave.
Sunnyvale, CA 94087
Phone:
PI:
Topic#:
(603) 502-4712
Dr. Henk Spaanenburg
OSD 03-022       Awarded: 04DEC03
Title:SystemVSIPL: Optimized Object Oriented Software Technology for Heterogeneous Computing
Abstract:Pentum Group, Inc. proposes to implement SystemVSIPL by writing the VSIPL/VSIPL++ libraries and templates in the SystemC system development language. SystemC has been developed for complex electronic and real-time systems design, implementation, and verification. As an extension of C++, SystemC has all the benefits of Object-Oriented programming for efficient software development productivity, while its key features to support event synchronization and parallelism allow for the creation of very efficient and high performance parallel applications. SystemVSIPL will result in better optimizations and greater efficiency for modern multi-threaded COTS processors and complex heterogeneous systems. Since SystemC is a also used commercially for high-level firmware and hardware design, a SystemC approach will result in tighter integration across the whole design space from chip to software application. Most of all, SystemVSIPL should result in code that meets the needs for real-time response.

EDAPTIVE COMPUTING, INC.
1245 Lyons Road, Building G
Dayton, OH 45458
Phone:
PI:
Topic#:
(937) 433-0477
Mr. David L. Barton
OSD 03-023       Selected for Award
Title:Design for Error Management (ErrorX)
Abstract:EDAptive Computing, Inc. (ECI) and Lockheed Martin Missile and Fire Control Group (Lockheed) present a proposal to design and implement robust error handling techniques in aircraft, shipboard, and other complex mission-critical systems. Our Design for Error Management (ErrorX) program provides a new design paradigm allowing the user to unambiguously describe correct system behavior at a high level of abstraction; model error conditions; specify possible repair strategies; generate tests to detect when error conditions occur; and correct those error conditions whenever possible. Specifically, ErrorX leverages a standards-based System Level Design Language (SLDL) called Rosetta, ECI's system design/block diagram editor called SyscapeT, ECI's VectorGenT - automated test vector generation tool, a system-monitoring engine for test application and error detection, and our team's significant experience with military, aerospace, and commercial engineering system management and development. The ErrorX approach will provide significant time and cost savings to the DoD, and increase the safety and reliability of complex software systems. The resultant gains in the quality of software error handling will make a direct contribution to the robustness of mission-critical DoD applications. Our Phase I solution will provide an easy transition to full-scale functionality in Phase II and will generate a product with significant commercial potential.

INTELLIGENT AUTOMATION, INC.
7519 Standish Place, Suite 200
Rockville, MD 20855
Phone:
PI:
Topic#:
(301) 294-5250
Dr. Leonard Haynes
OSD 03-023       Selected for Award
Title:Error Handling Based on Software Agent Technology
Abstract:In order to build complex software systems that are highly reliable, it is essential that the software conform to a rigid structure that bounds the types of errors that can occur at each point in the software. IAI has spent nearly a decade developing complex software systems based on autonomous software agents. In these systems, each agent has its own thread of control, and interacts with other agents via message passing. This structure is particularly well suited to advanced error detecting and error handling methods because errors associated with the interaction between agents can be separated from the errors related to each individual agent. Additionally, any errors caused by side effects or by data intended to be used by more than one agent must occur through the content of explicit messages. We have developed a rigid structure for these agents and within that structure, have developed a tool that proves the correctness of the design against a set of important errors. However, at this point the tool has no services to assist in debugging, and no services to support run time error checking or error handling. The work herein proposed will develop these features for our agent-based software CASE tool.

KESTREL TECHNOLOGY LLC
3260 Hillview Ave.
Palo Alto, CA 94304
Phone:
PI:
Topic#:
(650) 320-8888
Dr. Douglas Smith
OSD 03-023       Selected for Award
Title:Error Handling Techniques for Robust Mission Critical Software
Abstract:We will study the feasibility to build a tool, called HandlErr, which automatically enhances readable but non-robust programs with code to handle unhandled errors, according to user-specified error handling policies, thus making the programs robust. HandlErr performs a static analysis of the given program via data flow analysis, complemented by automated reasoning tools, to detect potentially unhandled errors. HandlErr reads declarative error handling policies that are specified by the user using a domain-specific language to be developed as part of HandlErr. HandlErr automatically enhances the program with error handling code to handle unhandled errors according to the user-specified policies. If the input program includes annotations like loop invariants and pre/post-conditions (e.g. as resulting from a formal refinement process of the program from high-level specifications), HandlErr performs a more accurate analysis of unhandled errors; however, annotations are not required.

SOHAR, INC.
5731 W. Slauson Ave., , Suite 175
Culver City, CA 90230
Phone:
PI:
Topic#:
(310) 338-0990
Dr. Herb Hecht
OSD 03-023       Selected for Award
Title:Error Handling Techniques for Robust Mission Critical Software
Abstract:We propose to develop a method and tool that will formalize and streamline error handling already at the very earliest stages of software design. We propose to tackle this problem at the level of object oriented design in a commonly used modeling language, the Unified Modeling Language, UML. Our scheme includes the development of formal error handling notation as an extension to UML (Robust UML) and interpreting error handling models with the use of a Timed Petri Net simulation. The tool we will develop: Robust UML Simulation tool (RUSim) will formalize & streamline the approach to error handling on the one hand, while also improving coverage of exceptional condition identification on the other. The pillars include: 1. Error handling notations: notations will be defined as extensions to UML that will include error condition identification, error handling roles and error coverage. 2. Automated translation of a UML design, including the exception handling aspects, to a Timed Petri Net (TPN) model for full and complete analysis of "unsafe markings" in the net. 3. Online Petri Net Analysis resulting in translation of "unsafe markings" in the net to error conditions that can occur to the software as is implemented in the design model.

WW TECHNOLOGY GROUP
4519 Mustering Drum
Ellicott City, MD 21042
Phone:
PI:
Topic#:
(410) 418-4353
Dr. Chris J. Walter
OSD 03-023       Selected for Award
Title:Error Handling for Large-Scale Software Systems
Abstract:Our proposed Phase I effort is to characterize, in a formal framework, the enabling methods and technologies required for error handling operations. The theoretical base offered in our approach provides a rich framework and comprehensive design and analysis space encompassing the range of fault/error models. A set of fault detection and isolation principles will be established and notation with an associated group of approaches developed for extending, modifying, augmenting or interfacing to various system elements techniques in large-scale software architectures. The concept of error containment regions enables specification and analysis properties to be put in place which can be evaluated and measured. The approach develops a framework that can be communicated to designers and users to make the error handling process more efficient and improve policy decisions. With these concepts in place, it is possible to construct fault tolerant systems from intolerant components. The approach is consistent with the use of custom designs or integrated COTS products and software repositories.

0 DB CODING, INC.
3 Annette Drive
Athens, OH 45701
Phone:
PI:
Topic#:
(740) 594-2308
Dr. Jeffrey C. Dill
OSD 03-025       Awarded: 03DEC03
Title:Fast, Flexible, Adaptive Channel Coding with Near-Shannon-Limit Performance
Abstract:The work proposed here presents a novel approach to channel coding which will provide flexibility and robustness which are necessary to alleviate many significant limitations in current RF communication systems. The proposed approach is entitled Tail-Biting Circular-Trellis Block Coding (TBC2). This family of coding methods provides unique advantages in the tactical military communication environment and in fact achieves the required fast, flexible, adaptive channel coding with near-Shannon limit performance. In particular, the coding approach presented here provides the following key advantages over other approaches: 1. Near-Shannon Limit bit error performance 2. Low latency, achieved by using very short codewords 3. High adaptivity in the dynamic selection of codeword size and code rate 4. Efficient decoding, using a highly parallel decoding algorithm and few iterations Taken together, the features of this code will provide a new level of agility and robustness for future battlefield communication systems, allowing the system to adapt to the highly dynamic nature of channel characteristics, quality of service requirements, latency requirements, and available frequency bands.

EFFICIENT CHANNEL CODING
600 Safeguard Plaza, Suite 100
Brookyln Hts, OH 44131
Phone:
PI:
Topic#:
(216) 635-1610
Dr. Fan Mo
OSD 03-025       Awarded: 03DEC03
Title:Fast, Flexible, Adaptive Channel Coding with Near-Shannon-Limit Performance
Abstract:For Phase I, ECC proposes to design a new type of LDPC codes optimized for the unique requirements of adaptive systems. The new set of adaptive codes will be designed with a common, parent-type of code that will comprise the core of our unique, highly efficient and low complexity encoder and decoder. The design will make it possible for fast on-the-fly switching while providing a set of codes that offers the performance characteristics required for varying practical channel conditions. We will also research the code designs and select codes required to accommodate other advanced techniques used in advanced communication systems, such as high order modulation, Space Time Coding (STC) and Code Division Multiple Access (CDMA).

XENOTRAN LLC
898 Airport Park Road, Suite 205
Glen Burnie, MD 21061
Phone:
PI:
Topic#:
(410) 761-2445
Dr. Kevin J. Page
OSD 03-025       Awarded: 03DEC03
Title:Fast, Flexible, Adaptive Channel Coding with Near-Shannon-Limit Performance
Abstract:Xenotran proposes to design a Forward Error Correction (FEC) device that is flexible and adaptive to changing communications needs. The FEC is based upon High-Dimensional Circular Trellis Coded Modulation (HDCTCM) and takes advantage of recent developments in iterative coding techniques. The underlying HDCTCM allows for dynamic trade-offs in latency, bandwidth, and spreading codes. A highly flexible FEC device, as proposed herein, will be paramount for the development of future communication systems. The implications to RF spectrum management are significant. The RF spectrum, which is currently partitioned according to frequency in a one-dimensional manner, could be managed dynamically as a multidimensional object of frequency, time, spreading code, and space.

ANDRO COMPUTATIONAL SOLUTIONS, LLC
Beeches Technical Campus, Bldg. 3, Ste. 4, 7902 Tu
Rome, NY 13440
Phone:
PI:
Topic#:
(315) 334-1163
Mr. Andrew L. Drozd
OSD 03-026       Awarded: 05DEC03
Title:BEYOND SPECTRUM: Multiobjective Joint Optimization for Efficient Utilization of the Radio Frequency Transmission Hypercube
Abstract:Electromagnetic (EM) propagation at RF frequencies is currently governed by a one-dimensional "real estate" approach to allocation of frequency bands, where the licensee has specific legal right to transmit within a band. The entire spectrum from 3 kHz to 30 GHz is currently allocated in this fashion. Unfortunately, this "set-it-and-forget-it" management scheme is straining under the immense pressure of exponentially increasing demand by burgeoning numbers of various types of wireless devices, spanning commercial and military applications all the way from short range home networks and cordless phones to the global information grid. Given the finite nature of the RF frequency spectrum and the fact that it is fully allocated, it is desired that alternative approaches to management of the resource be developed. This proposed effort is to develop innovative concepts for spectrum management that enables the effective and efficient joint utilization of all orthogonal EM transmission resources, including, but not limited to, time, frequency, geographic space, modulation/code, and polarization. This multi-dimensional environment is hereafter referred to as the Transmission Hypercube (TH), a term intended to convey the notion of a multi-dimensional resource space (with n "degrees of freedom" expressed as an n-tuple) in which each dimension allows orthogonality amongst users. Since there are multiple "dimensions" available in the RF transmission "space" and since these dimensions (RF resources) number greater than three, the term Hypercube is invoked. Currently, there are no known technological approaches to RF transmission in spectrum management that consider all five of these dimensions jointly, and certainly none that consider them in the context of a system optimization problem. This research project is aimed at developing approaches that consider the multi-dimensional nature of the transmission space, the results of which are expected to garner several orders of magnitude improvement in RF resource utilization and therefore, aggregate information throughput. The research will investigate the exploitation of optimization and orthogonality schemes that allow for multiple users to operate without interference. These include considerations for time slicing, frequency division multiplexing, directional antenna arrays, spread spectrum codes, and polarization. Conceivable approaches to a system for joint optimization of the multiple orthogonalizing transmission parameters will show that no two users are transmitting at the same time, even though they may be using the same frequency in the same space with the same exact spread spectrum code. Similar illustrations can be given in the case of a spatially orthogonalized system in terms of transmit beam patterns that do not overlap and cross-polarized waves in ideal cases. The innovation research will center on the application of multiobjective joint optimization and control system algorithms assisted by knowledge base technologies, novel frequency/time-domain interference rejection models, and waveform diversity techniques to study dimensional "synergy" and "prioritize" the TH cell dimensions. Also, joint time-frequency transforms will be considered. These approaches will encompass existing policies and ideas being pursued in Government, academia, and industry, the subset of which includes the obvious solutions such as: a single centralized global broker of TH cells, multiple distributed and coordinated local brokers of TH cells for local users, d-hoc "mesh" networking, fixed assignment, and hybrid approaches. This research supports the advancement and application of Information Systems Technology for decision-making, modeling and simulation technology, information management assurance and distribution; seamless communication; and computing and software technology. The motivation for this research has been the FCC's thrusts to reform current national spectrum policy, spearheaded by the President's Spectrum Policy Initiative and the formation of the Spectrum Policy Task Force.

ENGENIUM TECHNOLOGIES, INC.
4220 Eagles Wing Ct.
Ellicott City, MD 21042
Phone:
PI:
Topic#:
(410) 908-8003
Dr. Mike Pascale
OSD 03-026       Awarded: 02DEC03
Title:BEYOND SPECTRUM: Multiobjective Joint Optimization for Efficient Utilization of the Radio Frequency Transmission Hypercube
Abstract:Current generation communications systems already support multiple channel operation, store and forward relay functionality, and variable data rate features. What is lacking is a robust ad-hoc networking protocol that efficiently utilizes the multiple channels over time and over spatially distributed nodes to effectively communicate. The protocol should embrace the ability of nodes to bridge from one to another so that nodes that are not within direct communication range of each other can communicate. It will be shown in the sections that follow that the physical, data link, and networking layers in inextricably joined if maximum network utilization is sought. The objective of this proposal is to initiate the development of an approach and a practical implementation to the efficient use of resources among a network of RF communication systems.

ROCKFISH RESEACH
10210 Pepperhill Lane
Richmond, VA 23233
Phone:
PI:
Topic#:
(804) 740-8287
Mr. Chris Argenta
OSD 03-026       Awarded: 05DEC03
Title:An Intelligent and Flexible Transmission Hypercube Information Model
Abstract:This Phase I project investigates the dual use potential of advanced information modeling technology in supporting distributed management of the Transmission Hypercube (TH) resources. Rockfish Research specializes in innovative technologies for the modeling, simulation, visualization, and automation of complex system dynamics. We intend to investigate and evaluate innovative methods of distributing, visualizing, and automating aspects of the TH. Our approach is based on developing a component-oriented TH Information Model (THIM) that can easily be supplemented, in a plug-n-play manner, as new model elements, rules, and capabilities are developed to manage the TH. By separating engine and model, we believe that this underlying modeling system can be commercialized to support a wide variety of modeled system domains.

THE RIGHT STUFF OF TAHOE, INC.
The Right Place, 3341 Adler Court
Reno, NV 89503
Phone:
PI:
Topic#:
(775) 322-5186
Dr. Laurence E. LaForge
OSD 03-026       Awarded: 04DEC03
Title:Voronoi-Hamming Algorithms for Optimizing Channel Selection From the Radio Frequency Transmission Hypercube
Abstract:Advancing and applying our recent NASA-sponsored research for deep space avionics, we propose to develop new algorithms which - provably, autonomously, and in the presence of faults i) optimize the selection of channels from the radio frequency Transmission Hypercube, as it is termed in the OSD03-026 solicitation, in order to ii) connect healthy nodes into a working network quorum, and iii) keep them connected. To this end, we crystallize objectives for channel capacity (throughput) and latency (packet delay) in the language of optimization. Maximizing aggregate Shannon capacity (Eqns (11), (12)), our scalable algorithms will adaptively activate and deactivate channels, subject to constraints on channel bandwidth, number of channels, and per-band limits on signal power and noise. Similarly, and in the presence of faulty nodes and channels, our distributed algorithms will minimize latency (network radius or diameter, Eqn (1)). Where throughput and latency can be viewed as constraints, our algorithms will also solve dual problems whose objective is to minimize consumption of channel resources, such as power (Eqn (13)). Our approach blends information theory with breakthroughs in the mathematics of connectivity, especially 1.2.6) optimal routing enabled by Hamming graphs and 1.2.7) Voronoi algorithms from computational geometry.

TOYON RESEARCH CORP.
Suite A, 75 Aero Camino
Goleta, CA 93117
Phone:
PI:
Topic#:
(805) 968-6787
Dr. Richard E. Cagley
OSD 03-026       Awarded: 05DEC03
Title:BEYOND SPECTRUM: Multiobjective Joint Optimization for Efficient Utilization of the Radio Frequency Transmission Hypercube
Abstract:The wireless spectrum is a limited natural resource that is extensively used by a large number of users with diverse interests. Currently, this spectrum is allocated on a "set and forget" methodology whereby users or institutions are given exclusive rights to specific slices of this resource. It is desired that alternative management schemes be put into place that more efficiently utilize the wireless spectrum whereby it is treated as a multi-dimensional resource. This hypercube strategy will build upon the current one dimensional frequency separation to include time, space, modulation/coding, and polarization. Toyon proposes to provide an overview of how the spectrum is currently accessed, general strategies for increasing the efficiency of spectrum use, technical means of achieving these gains, and measures for quantifying whether one strategy should be used over another. This work will focus on general strategies for how radar and communications systems jointly optimize their use of the different dimensions of the transmission hypercube. In many cases, such systems will be able to sense available resources and be robust against interference. A Phase II effort will involve indepth analysis and simulation, with possible prototype deployment, of a select set of systems proposed in Phase I.

21ST CENTURY SYSTEMS, INC.
12152 Windsor Hall Way
Herndon, VA 20170
Phone:
PI:
Topic#:
(402) 505-7883
Mr. Jeffrey D. Hicks
OSD 03-027       Awarded: 30DEC03
Title:Webster
Abstract:In response to SBIR solicitation OSD03-027, 21st Century Systems, Inc. (21CSI) is pleased to propose investigating and developing a software agent system organized in a hierarchically networked architecture that incorporates use of multiple existing systems into a high-level information fusion system, entitled "Webster." Webster combines multidisciplinary expertise for the integration and inference of information products with quantification of data certainty and source reliability. As an agent-based information service enabler, special attention will be given to functionalities that retrieve, filter, analysis, and fuse information that are time sensitive and critical to counter-terrorist operations from accessing and sharing intelligence-based secure sources, web-based open sources, and field sources. Webster will employ two major technologies initiated at 21CSI namely, a Hierarchical Collective Agent Network (HCAN) architecture and a Hybrid Intrinsic Cellular Inference Network (HICIN) paradigm. HCAN Agents are responsible for overall supervision, scheduling, coordination, and execution of information retrieving, filtering, analysis, and fusion tasks. HICIN is an information integration engine embedded with advanced reasoning and uncertainty handling mechanisms. It takes a strong, multidisciplinary team to cover such a diverse problem domain as homeland defense. 21CSI has brought together the best in computing, cultural modeling, operations, and algorithms to meet the challenge.

21ST CENTURY TECHNOLOGIES, INC.
11675 Jollyville Road, Suite 300
Austin, TX 78759
Phone:
PI:
Topic#:
(512) 342-0010
Dr. Sherry Marcus
OSD 03-027       Awarded: 08JAN04
Title:Information Fusion System for Counter-Terrorism Operations
Abstract:In this Phase I SBIR , we shall demonstrate, the Integrated Fusion Dashboard, a system that can fuse uncertainties from multi-source data sets using two distinct detection mechanisms. To ensure that the fused product makes sense to the counter-terrorism analyst, we are pleased to have Mr. Craig Robinson, a decorated and retired 30 year veteran of the DIA to evaluate the results 21st Century Technologies has developed two detection methods to detect possible terrorist activity. TMODS, Terrorist Modus Operandi detection system, contains graph pattern matching and social network analysis detection capabilities that has been developed on multiple projects for DARPA (EELD, Genoa II, TIA programs), and Rome Laboratories and inserted into operational facilities at the DIA, Terrorist Information Awareness Center (TIA), and NYPD Counter Terrorism Division. In this Phase I effort, we shall compute and fuse uncertainties from these detection mechanisms. In the Phase I effort, we shall focus on open source data as a proof of concept with later extensions to multi-source data sets.

ALPHATECH, INC.
6 New England Executive Park
Burlington, MA 01803
Phone:
PI:
Topic#:
(703) 284-8444
Dr. William H. Bennett
OSD 03-027       Awarded: 21JAN04
Title:Confident Declaration Fusion Methods for Intelligence Processing of Relational Evidence
Abstract:Developing confident fused estimates derived by combining multi-source data (e.g. sensor data, human intelligence reports and open-source data) is a challenge for intelligence analysts. Data that can provide evidence about terrorist network operations are sparse, uncertain, disparate, and can include both misinformation and disinformation. In particular, open source data is often uncorroborated and has incomplete or unknown pedigree. Intelligence analysts need improved information fusion capabilities that can 1) detect/recognize patterns that indicate operations of terrorist networks while 2) characterizing the uncertainty of pattern recognition consistent with the uncertainty of the evidence. This effort will develop a consistent confidence assessment capability based on analytical methods that are independent of the cognitive assessments of human intelligence analysts. Our Phase I effort will lead to improved models of terrorist networks that characterize uncertainty, enabling improved fusion of uncertain information for counter-terrorist operations. ALPHATECH proposes an innovative approach in developing a confidence assessment capability for information fusion of evidence on terrorist operations by applying the theory of complex networks to model the uncertainty in information sources and pedigree. To facilitate this effort, we will leverage an existing ALPHATECH prototype system for fusing relational evidence to establish link detection for counter-terrorist operations.

APTIMA, INC.
12 Gill Street, Suite 1400
Woburn, MA 01801
Phone:
PI:
Topic#:
(215) 763-1284
Dr. Kari Chopra
OSD 03-027       Awarded: 21JAN04
Title:Infusiun: A Decision Support System for Intelligence Fusion and Uncertainty Analysis
Abstract:We propose to develop the Infusiun decision support tool to provide quantitative support to counter-terrorism analysts performing Level 3 information fusion (Threat Assessment) from open-source and human intelligence. The focus of our effort is on providing intelligence analysts with methods to manage uncertainty of information and generate quantitative measures of the validity of their assessments. We represent each intelligence source as providing evidence from which the analyst will, in combination with evidence drawn from other sources, seek to draw conclusions about the status and intent of a terrorist organization. Our method will provide the analyst with a multi-dimensional measure of the confidence level of these conclusions based on the following three properties: Robustness: the extent to which the conclusions are based on multiple independent sources of information; Reliability: the extent to which the conclusions are based on intelligence determined to be accurate and reliable; Strength: the extent to which the conclusions are supported by the evidence reported in the intelligence sources. We propose an innovative approach that draws upon techniques from social network theory, reliability theory, fuzzy logic, and belief theory to integrate empirical measures of validity with the judgment and expertise of the analyst.

CHARLES RIVER ANALYTICS, INC.
625 Mount Auburn Street
Cambridge, MA 02138
Phone:
PI:
Topic#:
(617) 491-3474
Dr. Subrata K. Das
OSD 03-027       Awarded: 16JAN04
Title:Decision Aid for Counter-Terrorist Analysts (DACTA)
Abstract:We propose to develop and demonstrate a Decision Aid for Counter-Terrorist Analysts (DACTA) by fusing and aggregating information from distributed sources. The DACTA architecture integrates a key decision aiding module with two supporting modules for information visualization and scenario simulation. The decision-aiding module provides the key decision-support functionalities at multiple levels: 1) low-level fusion of data into task-relevant information, supporting skill-based processing; 2) aggregation of information providing high-level abstraction of task-relevant knowledge, supporting rule-based processing; and 3) response recommendations based on the current context and prior knowledge, supporting knowledge-based processing. Robust performance is assured through the use of complementary artificial intelligence (AI) techniques (e.g., fuzzy logic, belief networks, probabilistic argumentation), thus providing strong foundations for quantifying uncertainty associated with terrorist threat assessment results from DACTA. The visualization module displays the information produced by the decision-aiding module in a format best suited to the current task, and the analyst's decision-making style and training. DACTA will incorporate a simulation environment supporting speculative analysis and exploration of model-generated hypotheses, via realistic simulated terrorist attack events. We intend to rapidly develop a prototype based on our in-house fusion and belief network engines, and demonstrate its validity in the context of a bioterrorism scenario.

SCIENTIFIC SYSTEMS CO., INC.
500 West Cummings Park - Ste 3000
Woburn, MA 01801
Phone:
PI:
Topic#:
(781) 933-5355
Dr. Adel El-Fallah
OSD 03-027       Awarded: 08JAN04
Title:Unified Robust-Bayes Multisource Counter-Terrorism Fusion
Abstract:If the states and information sources for counter-terrorism problems were like those associated with conventional sensor/target problems, then they could be solved using a systematic and proven mathematical methodology---the Bayes filter. This filter provides an explicit methodology for modeling uncertainties, propagating these uncertainties through time, and extracting estimates of desired quantities (as well as measures of reliability of those estimates) that correctly reflect the influence of system uncertainties. The ambiguousness of human intelligence information sources and of A PRIORI human cultural context would seem to automatically preclude the feasibility of the Bayes filter in counter-terrorism applications. Scientific Systems Company, Inc. (SSCI) and its subcontractor Lockheed Martin Tactical Systems (LMTS) believe that this may not be the case. Certain more conventional DoD problems---force structure analysis and single-target filtering using unconventional information (natural language, inference rules)---can be addressed using Bayes filter methods, and these problems bear a family resemblance to counter-terrorism applications. Consequently, we propose the investigation of a novel Bayes-filter information-fusion approach to counter-terrorism applications that both hedges against, and accounts for, inherent uncertainties. Specific Phase I tasks are: (1) develop a theoretical/mathematical/algorithmic foundation; (2) design high-level techniques for modeling states and measurements; (3) develop high-level designs for mathematical algorithms, including uncertainty assessment; (4) design potential test simulations; (5) develop implementation, simulation, and test plan; (6) develop a detailed plan for further analysis and implementation in a Phase II effort. The project team includes Dr. Ronald Mahler of Lockheed Martin. Lockheed Martin will provide both technical and commercialization support in the application of counter-terrorism technologies.

STOTTLER HENKE ASSOC., INC.
951 Mariner's Island Blvd., STE 360
San Mateo, CA 94404
Phone:
PI:
Topic#:
(206) 545-3533
Mr. Matt Broadhead
OSD 03-027       Awarded: 30DEC03
Title:Improving Information Fusion Capabilities through the Explicit and Pervasive Consideration of Uncertainty
Abstract:A variety of forces have led to decreases in the average experience of intelligence analysts. This shift is an incredibly important development since a lack of experience has been strongly linked to: failures in locating key information sources; the incorporation of faulty information in reports; and the premature closing of investigations. To help mitigate these impacts and generally improve intelligence analysis capabilities we propose an approach to information fusion that employs an integrated probabilistic framework in order to explicitly account for disparate indicators of uncertainty and thereby improve human decision making. The use of this integrated framework, in which information extraction and fusion decisions are made with a common pool of evidence and inference procedures, will also allow much richer forms of inference than possible with the current state of the art technologies. To further extend the reach of the proposed system, Assure, to higher level fusion tasks, we will employ a mixed-initiative interaction paradigm that will enable a form of cooperative information fusion and uncertainty analysis. Phase I research and development of a limited prototype will provide a solid foundation for the complete implementation of Assure in Phase II and its eventual commercialization.

ALPHATECH, INC.
6 New England Executive Park
Burlington, MA 01803
Phone:
PI:
Topic#:
(781) 273-3388
Dr. Gregory S. Lauer
OSD 03-028       Awarded: 05DEC03
Title:Information Uncertainty Portrayal
Abstract:The information age has produced a flood of information to system operators and decision-makers in a multitude of application domains. Each piece of data that feeds a system can produce complex interactions with other data in the scenario. This project will examine the uncertainty associated with data that describes a battlefield awareness scenario and will determine techniques for visualizing these uncertainties with the goal of aiding in the decisions that a commander must make. This work will draw upon previous ALPHATECH work, including the outputs of tracking, automatic target recognition, and information fusion software algorithms, for the generation of data and uncertainty information. A scientific experimental procedure will be used to guide a group of mock decision-makers through a series of experiments. Results of the experiments will be analyzed, with the goal of extracting those visualization techniques that work across the information domains tested. We will also extrapolate of the results of the experiments to non-military domains. Universal concepts for visualizing uncertainty will complete domain specific concepts, resulting in a demonstration of the most promising visualization techniques identified.

ATC - NY
33 Thornwood Drive, Suite 500
Ithaca, NY 14850
Phone:
PI:
Topic#:
(607) 257-1975
Dr. Robert Joyce
OSD 03-028       Awarded: 05DEC03
Title:Interactive Display of Probabilistic Geo-Spatial Information
Abstract:Systems that integrate and present data from multiple sources often ignore the probabilistic uncertain nature of the information they are handling. Such systems are often designed with pre-determined thresholds or other algorithms to make binary go/no-go decisions, rather than presenting the entire picture and all data sources. Indeed, simply displaying all information sources and uncertainties used in making a calculation can be unwieldy and confusing to decision makers. ATC-NY, in collaboration with Architecture Technology Corporation, will design and build uVIZ, an intuitive, interactive visualization system to display uncertain information associated with both raw data sources and derived information. uVIZ will provide a layered view, where users can select the level of detail displayed about any particular piece of information: single values that are the result of estimation algorithms, modeling error distributions, and so forth, all the way down to raw sensor error information and human observations. As information uncertainty generally increases as observations age, existing probabilistic models that incorporate time can also be employed, allowing visualization of past or future scenarios as well as time-lapse animations.

BMA ENGINEERING, INC.
11429 Palatine Drive
Potomac, MD 20854
Phone:
PI:
Topic#:
(301) 299-9375
Dr. Bilal M. Ayyub
OSD 03-028       Awarded: 05DEC03
Title:Information Uncertainty Portrayal
Abstract:The objective of Phase I: identify and develop innovative methods for portraying information for decision makers so that data uncertainty and content are readily understood. The specific objectives of the proposed effort are: 1. Review the present practice of visualizing uncertainty in battlefield data; 2. Examine the use of various visualization methods, as discussed in Section 1, for various uncertainty and ignorance types as defined in Figure 2; 3. Develop methods for visualizing uncertainty based on our ignorance hierarchy; 4. Incorporate our risk management process in the development of these methods; 5. Define, format, and collect the data needed to support these methods; 6. Demonstrate the effectiveness of the methods, especially concerning improvement in the ability to make more appropriate (right or correct) decisions; 7. Develop a plan for Phase II of the SBIR effort in the form of prototype products with a high potential for successful commercialization. We will develop innovative methods for portraying information for decision makers, so that data uncertainty and content are readily understood, based on our unique hierarchy of ignorance. The visualization developed visualization methods will satisfy the following principles: (1) apprehension, (2) clarity, (3) consistency, (4) efficiency, (5) necessity, and (6) truthfulness. The following icon attributes can be used to communicate information about uncertainty and ignorance: (1) form, (2) orientation, (3) color, (4) texture, (5) value, (6) size, (7) position, (8) motion, (9) intensity, (10) shading, and (11) special effects, such blinking, animation, etc. We will further develop our unique Ignoriconsc, and perform the following tasks: (1) Define the types of ignorance and uncertainty; (2) Identify visualization requirements for battlefields; (3) Develop visualization methods; (4) Incorporate our risk mitigation tools in developing the methods; (5) Development a variety of Ignoriconsc; (6) Define data needed to select and format Ignoricons; (7) Demonstrate models; and (8) Develop a plan for Phase II of the SBIR.

CHARLES RIVER ANALYTICS, INC.
625 Mount Auburn Street
Cambridge, MA 02138
Phone:
PI:
Topic#:
(617) 491-3474
Dr. Jonathan Pfautz
OSD 03-028       Awarded: 05DEC03
Title:Acquiring representations of Meta-information to Enhance Battlespace Awareness (AMEBA)
Abstract:Modern battlespaces are characterized by large amounts of complex information that must be rapidly processed by the decision-maker. Part of this complexity is due to meta-information, or, characteristics of information such as uncertainty, staleness, etc. that add to the decision-making burden. Knowledge regarding this meta-information, and methods for effectively portraying it, have the potential of not only relieving the decision-maker's workload, but also encouraging more battlespace-aware decisions. To support meta-information visualization, we propose to design and prototype a system for Acquiring representations of Meta-information to Enhance Battlespace Awareness (AMEBA). Three core components characterize the proposed process. First, we will perform a cognitive task analysis of a selected set of case studies to develop a principled categorization of the key types and sources of information and meta-information, to identify the meta-information required by the decision-maker, and to identify display elements in existing military systems. Second, we will design and prototype a toolkit for augmenting incoming data with the identified meta-information types and rapidly creating new meta-information visualization methods and interfaces. This toolkit will support integration of prototype visualization methods with existing DoD display systems. Third, we will develop an evaluation methodology, including metrics of the utility of each meta-information visualization technique.

SA TECHNOLOGIES, INC.
4731 East Forest Peak
Marietta, GA 30066
Phone:
PI:
Topic#:
(601) 898-9053
Dr. Jennifer Riley
OSD 03-028       Awarded: 08DEC03
Title:Information Uncertainty Portrayal
Abstract:We propose to develop novel representations of uncertain command and control (C2) data that will be semantically rich and support quick, accurate mental visualizations of battle states. We will analyze current information visualization methods and human-centered design principles to support situation awareness and decision making of intelligence officers for providing the commander with a complete and timely understanding of the opposition force. Information visualization and graphical methods will be evaluated to assess their utility for application or adaptability to various types of C2 data. We will also describe approaches for heuristically and empirically investigating the usability and usefulness of information portrayals developed as part of the Phase I research. Metrics for assessing the quality of the portrayals will be developed. Phase II work will involve development of a prototype of an information visualization system that can provide the user (intelligence officer and commander) with a customizable operating picture of the battlespace and the ability to easily locate, integrate, and comprehend information from distributed sources of varying reliability. Work from Phase I will also provide initial support for the development of a domain and task specific ontology and user guide for the conceptual system.

SECURBORATION, INC.
695 Sanderling Dr
Indialantic, FL 32903
Phone:
PI:
Topic#:
(321) 591-9836
Mr. Lee Krause
OSD 03-028       Awarded: 05DEC03
Title:Information Uncertainty Portrayal
Abstract:Securboration Inc., teaming with Dr. Eugene Santos of the University of Connecticut, is pleased to propose our innovative approach for visualizing and assessing uncertainty entitled, Uncertainty Prediction System (UPSYS). During the phase one effort we plan to demonstrate how the innovative UPSYS technology can be applied to supports visualization and assessment of event uncertainty within the context of Predictive Battlespace Awareness (PBA) as shown in figure 1. UPSYS consist of four main components: 1) an ontology that specifies PBA classes and relationships 2) an Uncertainty Layer that uses the ontology to automatically generate Bayesian Networks for event uncertainty calculations, 3) a Casual Analysis Engine that reasons over the ontology to infer cause and effect, 4) visualizations that provide information of all aspects of an event and its uncertainty in relation to other events and PBA workproducts.

VISINTUIT
280 Shuman Blvd., Suite 115
Naperville, IL 60563
Phone:
PI:
Topic#:
(630) 428-0910
Dr. T. A. Keahey
OSD 03-028       Awarded: 05DEC03
Title:Uncertainty Awareness for Battle Space Visualization Systems
Abstract:A critical problem for military commanders and decision makers is to understand all of the relevant factors that must go into the assessment of a battle field situation. This problem is made exponentially more difficult by the fact that so many of those factors have an associated uncertainty level, which may or may not be known a priori. Creating effective visual tools for intuitively conveying this information along with its certainties to the battle field commander has a tremendous potential for increasing the accuracy and outcome of military operations. Over the past decade there has been a massive increase in computer graphics capability, driven largely by the economics of the computer game industry. This increase in graphics processing power has also been accompanied by the new generation of real-time graphics tools for generating Hollywood-style visual effects that can greatly enrich the display of information. Together, these have opened up new opportunities for another generation of visualization tools in general, and battle space visualization systems in particular. This research project proposes advanced visualization techniques that leverage modern graphics processing power; moving beyond simple polygons, texture mapping, and 3D to create a new class of visualization tools aimed at accurately conveying the certainty of information in a battle space overview tool. Our ideas exploit animation, shading, illumination, and new visual techniques to overcome inherent human perceptual limitations. Although relatively new to battle space visualization, these approaches have shown great progress in other domains. Taken together, our approach and the vastly more powerful graphics capabilities have great potential for dramatically improving our ability to understand information in the presence of uncertainty.

APTIMA, INC.
12 Gill Street, Suite 1400
Woburn, MA 01801
Phone:
PI:
Topic#:
(781) 496-2428
Dr. Gavan Lintern
OSD 03-029       Awarded: 25NOV03
Title:Portraying Uncertainty in Battlefield Weather Situational Awareness
Abstract:Military planning and logistics must account for weather effects on both a strategic and tactical level. There is, however, uncertainty associated with any forecast that confounds planning and decision-making in relation to anticipated weather effects and their impacts on equipment and personnel. The Aptima team proposes to develop visualizations of weather uncertainty from forecast data to present information about weather uncertainty to warfighters in a format that can support their decision processes. This decision aid will be designed to help warfighters anticipate the functional effects of weather on their operations so that they can exploit weather-related opportunities and avoid weather-related disruptions. Within this project we will employ methods of cognitive task analysis to identify decision processes and strategies and the information used in those decision processes as warfighters deal with weather uncertainty. We will also document the informational properties available from current weather forecasting systems, the nature and sources of uncertainty, and the nature of weather impacts on selected operational systems. We will integrate the products of these analyses into a representation of essential weather information in a manner that supports the typical decision processes and strategies of users. We will then develop a visualization prototype for portraying that representational concept.

CHARLES RIVER ANALYTICS, INC.
625 Mount Auburn Street
Cambridge, MA 02138
Phone:
PI:
Topic#:
(617) 491-3474
Dr. Jonathan Pfautz
OSD 03-029       Awarded: 02DEC03
Title:Weather Prediction Uncertainty Management And Representation (PUMAR)
Abstract:Strategic weather centers run numerical weather prediction (NWP) models, but military operators do not often have the time or capability to analyze the NWP information -- warfighters must rely on decision aids to succinctly translate the NWP information into battlefield intelligence. Since NWP models are not perfect, methods are needed to determine and display the meta-information (e.g., uncertainty, resolution, timeliness) about a forecast to assess mission impact. To help address this need, we propose a Phase I SBIR program to prototype and demonstrate a system for weather Prediction Uncertainty Management And Representation (PUMAR). Four core components characterize our approach. First, we will perform a cognitive task analysis of a weather-influenced decision-making scenario, followed by a principled categorization of the types and sources of uncertainty influencing the decision-maker. Second, we will identify and develop methods to manage the key types of uncertainty in NWP forecasts (e.g., ensemble forecasting, Bayesian hierarchical methods). Third, we will prototype several meta-information visualization techniques and user interfaces in the context of a decision aid that uses weather information to determine weapon capabilities. Fourth, we will integrate these concepts and methods into a computerized system designed to interface with existing DoD meteorological systems.

NEXT CENTURY CORP.
8101 Sandy Spring Road
Laurel, MD 20707
Phone:
PI:
Topic#:
(301) 939-2600
Mr. Clark Dorman
OSD 03-029       Awarded: 13NOV03
Title:Portraying Uncertainty in Battlefield Weather Situational Awareness
Abstract:The proposed initiative will yield a working prototype of our innovative Weather Risk Analysis and Portrayal (WRAP) system that helps quantify uncertainty in disparate weather forecasts and presents this information to users in highly intuitive ways that facilitate at-a-glance analysis and rapid decision making. Next Century Corporation has teamed with Marcus Weather Inc. to build the WRAP system based on our combined expertise in developing solutions for customers in weather-sensitive businesses as well as designing highly intuitive, easy-to-use graphical user interfaces and decision aids for a wide array of application domains. Some key innovations proposed are: 1) enabling the user to interactively adjust the minimum certainty threshold that must be satisfied for those weather conditions being forecast; 2) providing a temporal slider that enables users to interactively view battlespace conditions over time, either by going forward in time based on forecasts and predictions or by going backward in time to view historical data including actual, predictive, and certainty model overlays; and, 3) allowing users to perform an interactive review and comparison of predicted vs. actual weather conditions to aid in hotwash/training activities and to aid analysts in tuning of predictive models, rule sets, and uncertainty thresholds.

SONALYSTS, INC.
215 Parkway North, P.O. Box 280
Waterford, CT 06385
Phone:
PI:
Topic#:
(860) 326-3635
Mr. Cranston R. Coleman,
OSD 03-029       Awarded: 12DEC03
Title:Portraying Uncertainty in Battlefield Weather Situational Awareness
Abstract:Current automated weather forecast models and associated tactical decision aids often produce information that is insufficient or misleading because they do not adequately reveal the accuracy of the forecast. Users of these decision aids need a way to understand the reliability of the forecast. Tacticians make decisions semantically, reasoning in terms of qualitative values and the underlying meaning of data. This process can be reinforced through an effective visualization method that reflects the way a user thinks about weather data. This Phase I effort will investigate the feasibility of a method that provides a semantic mapping of data to quantifiable terms. Using this method, Sonalysts will develop and demonstrate an approach for more effective processing and visualization of weather forecast data. This effort will demonstrate the ability to use weather prediction terms in tactical decision-support applications, while the visualization effort will address the underlying uncertainty of weather forecast data.

WEATHER VENTURES LTD.
240 Mt. Pleasant Drive
Boalsburg, PA 16827
Phone:
PI:
Topic#:
(814) 466-2231
Dr. John A. Dutton
OSD 03-029       Awarded: 28NOV03
Title:Battlefield Weather Probabilities
Abstract:Effective management of battlefield weather risk requires a clear understanding of the likelihood of weather events critical to military operations. By combining contemporary capabilities in meteorological observations, atmospheric science, computer prediction, and communications, we will design a multiscale ensemble prediction system (MEPS) for predicting battlefield weather probabilities. This system will involve readily available large-scale ensemble forecasts, ensembles of MM5 forecasts at high resolution, and a forecast calibration process. Key design issues will be resolved by Weather Ventures in collaboration with the Penn State MM5 group using results from a remarkable high-resolution forecast experiment performed in the summer of 2003 by the National Oceanic and Atmospheric Administration (NOAA). This data set, including some 1500 mesoscale forecasts in a virtual ensemble of 32 members, provides a rich and unprecedented source of information for designing a MEPS for battlefield weather probabilities. The data will help us explore the balance between forecast skill, reliability, and cost. To best serve commanders and soldiers, battlefield probabilistic forecasts must be visualized and summarized in decision aids. Work in progress on visualizing probabilistic forecasts and constructing decision aids will be extended to the battlefield case, in preparation for a more concentrated effort in Phase II.

CYCORP, INC.
3721 Executive Center Drive, Suite 100
Austin, TX 78731
Phone:
PI:
Topic#:
(512) 514-2951
Dr. Michael Witbrock
OSD 03-030       Selected for Award
Title:Leadership Agent for Multi-source Information Fusion in Counter-Terrorism
Abstract:Current document search techniques are inadequate to handle queries whose answer requires establishing links among a number of (potentially obscure) facts. We propose to remedy this situation with a prototype system that applies statistical and knowledge-based methods to the problem of identifying chains of connections between terms and entities. We will extend existing predictive annotation methods for indexing entities by tagging them with Cyc concepts. In addition, our system will be capable of indexing entire facts, by applying Cyc's parsers to text sentences. Parsing a sentence into the CycL representation language results in a full semantic translation, so links among entities mentioned in a sentence will be transparent in the CycL representation. We propose also to demonstrate the discovery of "anonymous" links, by using technology we developed during the ARDA AQUAINT project. This module, the Holistic Query Expansion Algorithm (HQEA), is designed to process a textual query and produce a set of pairwise highly-correlated terms that connect all the terms within the original query into a coherent topic. Finally, a user-feedback mechanism will be developed to refine the performance of the Cyc-based semantic annotators.

INSIGHTFUL, INC.
1700 Westlake Ave N, Suite 500
Seattle, WA 98109
Phone:
PI:
Topic#:
(206) 283-8802
Dr. Giovanni Marchisio
OSD 03-030       Selected for Award
Title:Leadership Agent for Multi-source Information Fusion in Counter-Terrorism
Abstract:We have a proprietary algorithm for Latent Semantic Analysis (LSA) called Latent Semantic Regression (LSR). We will demonstrate our unique ability to establish semantic networks that follow latent links across large collections of documents. We will extend applications of LSR to more complex indices than the classical term-document matrix. In particular, we will demonstrate our ability to perform latent semantic retrieval not only on indices that measure keyword frequencies, but on more complex linguistic attributes or structures. These experiments will include demonstrations of latent semantic retrieval on entity-entity matrices, and action-entity or "event" matrices produced by our proprietary technology for deep parsing and information extraction.

IXMATCH, INC.
5555 West 78th. Street, Suite E, Dewey Hill Busine
Edina, MN 55439
Phone:
PI:
Topic#:
(952) 842-7500
Dr. Eui-Hong (. Han
OSD 03-030       Selected for Award
Title:Concept-based Information Retrieval and Fusion Engine
Abstract:Existing search technology is inadequate for addressing the information retrieval requirements of today's complex tasks that require the integration of information from multiple and heterogeneous sources, the discovery of arbitrary and non-obvious relations between the documents, and the uncovering of information that is intentionally trying to hide. The goal of this project is to advance the state-of-the-art in text-based search technology by developing an experimental prototype that combines best-of-bread technologies to address these problems. Our concept-based information retrieval system combines background knowledge with knowledge discovered dynamically by globally analyzing the target collection(s) and locally analyzing the retrieved documents to identify the appropriate concept space(s) for each query and progressively follow all the clues that are present in them. Our information fusion system takes advantage of alternate representations of the underlying collection and information retrieval approaches to identify the documents that are the most relevant to the desired concepts.

KNOWLEDGE ANALYSIS TECHNOLOGIES, LLC.
4940 Pearl East Circle, Suite 200
Boulder, CO 80301
Phone:
PI:
Topic#:
(303) 545-9092
Dr. Karen Lochbaum
OSD 03-030       Selected for Award
Title:Knowledge Triage for Military Leadership
Abstract:Intelligence for combating terrorism is made difficult by the hundreds of minor languages and dialects potentially spoken by terrorists. There is a pressing need for automated methods that greatly reduce reliance on scarce bilingual analysts and automatic translation systems. The project objective is to develop a computer-based tool that will allow military leaders and intelligence analysts to quickly and accurately select only important documents for translation and/or analysis. An important aspect of the problem is known as Knowledge Triage. For example, in Iraq, over one million pages of paper documents, the majority in poorly handwritten Arabic script, have been collected, Most of this is almost certainly of no intelligence value, while some may contains hints, and a very small number "smoking gun" information. An existing cross language retrieval technique based on Latent Semantic Analysis (LSA) is capable of performing the triage, but only if given text in electronic form. In the proposed system a target language speaker would read paper documents into a speech recognizer, and the system would assess their possible relevance and summarizes their contents in English keywords. The system is robust to speech recognition errors and can be adapted to a new language in a few weeks.

SEHDA, INC.
465 Fairchild Dr.,, Suite 123
Mountain View, CA 94043
Phone:
PI:
Topic#:
(650) 864-9900
Mr. Farzad Ehsani
OSD 03-030       Selected for Award
Title:Massive Scale Text Processing Using Phrases
Abstract:Sehda plans to separate relevant data from the irrelevant through a technique in which every document being processed would be marked automatically with contextual meta-data to characterize the genre, author, and information in the article. Application of Sehda's technology to the results of this categorization scheme should permit the automatic extraction of information responsive to any inquiry in a natural, flexible, manner far more effective than current key word searches at identifying and grading relevant responsive material. Sehda will also build a browser which will allow the user to navigate this data seamlessly, by examining different cuts through the data in order to narrow in on relevant associations the viewer wishes to focus on. Finally, we have devised potential ways to do "Rare Event Detection" using statistical analysis of the content of unusual responses, communications or word associations, although this idea is only at the formative stage, and would require additional research and testing.

HI-Z TECHNOLOGY, INC.
Suite 7400, 7606 Miramar Road
San Diego, CA 92126
Phone:
PI:
Topic#:
(858) 695-6660
Dr. Aleksandr S. Kushch
OSD 03-H01       Awarded: 31DEC03
Title:Optimized Off-the-shelf Self-Powered Portable Trap for Haematophagous Arthropods
Abstract:An optimized off-the-shelf self-powered trap for haematophagous arthropods with an enhanced mosquito capturing efficiency will be designed, fabricated and tested by Hi-Z Technology and the John W. Hock Company. The self-powered feature will be realized by using a fossil fuel (Diesel/JP-8/Propane) driven burner coupled with a thermoelectric power generating section. Electric power that is generating by the thermoelectric section will drive the existing Model 1012 mosquito trap components, such as fan, light bulb and circuit board. The burner exhaust will supply CO2, H2O vapor and heat to the trap entrance in order to attract mosquitoes. This proposed effort will make extensive use of Hi-Z's expertise in the field of thermoelectric power supplies and self-powered appliances development and the John W. Hock Company's experience in the insects sampling devices and techniqe development. The fossil fuel power supply will have dual-use applications. Simple, rugged and driven by a logistic or propane fuel the power supply can be employed in the both, military and civilian markets as a battery charger or direct power source for some electronic devices, such as radios, cell phones, lighting, etc.

LYNNTECH, INC.
7607 Eastmark Drive, Suite 102
College Station, TX 77840
Phone:
PI:
Topic#:
(979) 693-0017
Dr. Craig C. Andrews
OSD 03-H01       Awarded: 01JAN04
Title:Optimized Off-the-shelf Portable Trap for Haematophagous Arthropods
Abstract:There is a definite need for the development of an efficient portable device that can reliably be used to survey a broad range of disease carrying mosquito populations in the military field to assess health risk prior to service members' deployment in the region. Limited to requiring CO2 supplies from dry ice, CO2 gas cylinder, and propane gas, commercially available light traps can not be properly operated in regions where these supplies do not exist. Supply can be transported to the mission field at the expense of cost, time, and extra risk exposure to service member through handling these material (propane gas cylinders has a potential to explode when handled inappropriately). Lynntech's innovative CO2 generation technology overcomes these limitations by incorporating a miniature liquid fuel (JP-8, diesel, kerosene, etc.) CO2 generator to the existing miniature light trap. The add on module is compacted into smaller than a gallon jar package, including all control peripherals, attractant delivery devices, system monitoring equipment, and batteries, at the price of less than $400. In addition, this innovative technology can be applied to larger traps and any type of CO2 baited traps coupled with other chemical attractants. CO2 generator allows the use of CO2 baited trap in more remote regions, making them an attractive option for low cost trap add on module for disease carrying mosquito detection or population management in military operations, as well as civilian applications.

RADIATION MONITORING DEVICES, INC.
44 Hunt Street
Watertown, MA 02472
Phone:
PI:
Topic#:
(617) 926-1167
Mr. Stuart Miller
OSD 03-H01       Awarded: 30DEC03
Title:Portable and Environmentally Friendly Trap for Haematophagous Arthropods
Abstract:Haematophagous arthropods including ticks and mosquitoes spread many deadly diseases and pose serious threats to people throughout the world. Current measures only mediate the problem, and no solution exists to stop the spread of disease by these vectors. Trapping these arthropods can be a significant component in controlling and monitoring their populations, yet current trapping devices are either ineffective, or impractical for widespread use. To address this problem we propose to develop a trap that will be more effective in attracting a wide variety of these haematophagous arthropods. The new trap design will require little power, will be lightweight, and will be portable and environmentally friendly. The basic approach of our trap is to simulate the human host in terms of the important components that make humans attractive to these arthropods. We will use new methods to mimic both the heat and the carbon dioxide produced by humans. This new design is expected to outperform current traps in terms of their ability to attract a wide variety of important haematophagous species.

TDA RESEARCH, INC.
12345 W. 52nd Ave.
Wheat Ridge, CO 80033
Phone:
PI:
Topic#:
(303) 940-2355
Dr. William L. Bell
OSD 03-H01       Awarded: 31DEC03
Title:Improved Arthropod Trap System
Abstract:Biting insects and ticks represent a very substantial threat to human health. U.S. military personnel may be at greatly increased risk from insect-borne disease as they operate at locations around the world. Any program to evaluate health risks and control disease-carrying arthropods relies on a system to trap the arthropods for evaluation. Improvements to the current arthropod capture technology are required to improve its efficiency and make it more readily transportable. In this project TDA Research, Inc. (TDA) proposes to develop an improved system to trap dangerous arthropods. In Phase I we will focus on an improved mosquito trap, because mosquitoes represent both a serious threat and a challenge to trapping technology. We will first identify specific methods and formulations to increase the capture effectiveness of standard traps. We will test the units in enclosures, modifying as necessary to improve performance. We will conduct a field test to compare our improved unit with standard technology. We will conduct an engineering analysis to assess the cost of the improved system in production and assure that it can be safely manufactured, transported used and disposed of after use. In Phase II we will complete development with additional field testing.

UTD, INC.
8350 Alban Road, Suite 700
Springfield, VA 22150
Phone:
PI:
Topic#:
(703) 440-8834
Dr. Charles A. Bass
OSD 03-H01       Awarded: 01JAN04
Title:Optimized Off-the-shelf Portable Trap for Haematophagous Arthropods
Abstract:Blood-sucking or haematophagous arthropods such as mosquitoes, ticks, and sandflies act as vectors of some extremely deadly and debilitating diseases such as malaria, dengue, yellow fever, sandfly fever, Japanese encephalitis, Q-fever, and tularemia. The current kit used by preventative medical teams contains a light trap based on design developed in 1962. Many mosquito species cannot be monitored by light traps. Samples collected are frequently inadequate to assess human health risk. While light traps show more success when supplemented with carbon dioxide, usual methods for CO2 baiting - bottled gas or dry ice - lack of universal availability and create heavy logistical demands. UTD proposes a design concept that emphasizes portability, storability, and ease of use. It will use a combination of attractants including CO2, heat, moisture, UV and white light, and selected synthetic scents. The trap will require no batteries, and will function on a variety of logistic fuels including JP8. UTD brings a wealth of experience in the application of design analysis, simulation, and innovative engineering prototyping to this problem. We have assembled a talented development team and have secured the services of a medial entomologist who is experienced in both academic research and military applications of vector surveillance.

ALPHATECH, INC.
6 New England Executive Park
Burlington, MA 01803
Phone:
PI:
Topic#:
(858) 812-2994
Dr. Gary L. Jahns
OSD 03-H02       Awarded: 22DEC03
Title:Encapsulating Generalized Principal Components and Support Vector Machines in a Nonlinear Statistical Pattern Recognition Toxicology Tool
Abstract:Identifying toxic-substance exposure at low, subtoxic concentrations requires interpretation of complex, time-related changes in gene, protein, and metabolite expression patterns. We propose a statistical pattern recognition software tool based on flexible and adaptable modules. In Phase I, we will produce the tool framework and Principal Components Analysis (PCA) and Support Vector Machine (SVM) modules. The tool framework preprocesses genomic, proteomic, and metabonomic data sets from clinical sources. PCA linearly transforms each type of data to an orthogonal space of significantly reduced dimension in which expression of like toxins are clustered. Through Statistical Learning Theory, the SVM adapts and estimates a nonlinear mapping function from the expression-data input space to a decision feature space using data for which ground truth has been independently established. A similarity kernel in feature space induces a metric on the input space by selecting key feature components and produces a nonlinear decision boundary. Selection of the kernel can range from a simple distance metric to neural networks (multi-layer perceptrons, radial/elliptical basis function networks, etc) to fuzzy membership functions. By changing the kernel, the performance of the SVM decision boundaries can be optimized over a range of kernel similarity metrics, feature mappings, and feature selection. Other modules may be added in the future to implement a wider suite of solutions in a user-friendly analysis environment.

DELTA SEARCH LABS, INC.
400 Technology Square
Cambridge, MA 02139
Phone:
PI:
Topic#:
(617) 551-4603
Dr. John Zhang
OSD 03-H02       Awarded: 05DEC03
Title:An Expert System for the Pattern Recognition of Biological Effects
Abstract:The objective of this project is to develop an expert system capable of measuring and interpreting complex, time-related, global changes in gene, protein expression, and metabolite pattern profiles. The expert system will identify toxic substances exposure at very low, sub-symptomatic concentrations to military personnel. The project will meet three challenges to the development of the system: (a) identify the toxic substances and quantify the low exposure level with estimated probabilistic confidence; (b) predict the future evolution of the biological effects utilizing available time-dependent measurement records of pattern profiles; and (c) integrate the usage of all available pattern profiles and optimize the synergized capability of interpreting and predicting the biological effects. In Phase I, we will develop a system concept by applying proven multivariate statistical methods for signal processing widely used in electrical engineering and geophysical sciences. The feasibility and extendibility of the concept in terms of accuracy, efficiency, and applicability will be demonstrated with a few practical examples.

GENOMATICA, INC.
5405 Morehouse Drive, Suite 210
San Diego, CA 92121
Phone:
PI:
Topic#:
(858) 362-8550
Dr. Tom Fahland
OSD 03-H02       Awarded: 09DEC03
Title:Biological Effects Pattern Recognition Tool Using Multivariate Statistical Data Reduction
Abstract:The application of modern high throughput genomic, proteomic, and metabonomic technologies to the field of toxicology will provide significant breakthroughs and advances. By simultaneously looking at data on the gene, protein, and metabolite concentrations, a more complete picture of cellular behavior can be determined. Using statistical and mathematical algorithms applied to high throughput data, patterns consistent with low level toxic exposure can be determined. In the current proposal, we intend to apply multivariate linear and non-linear statistical and mathematical techniques to find complex patterns in the data that are consistent with validated samples of low-level toxic exposure. By using supervised machine learning techniques, training data with solid cross-validated measurements will be used to distinguish between different levels of toxic exposure. By seamlessly applying these techniques to all the different data types, small correlations and patterns can be best measured and identified. Upon completion of this SBIR project, we expect to have developed a prototype software tool that integrates the multivariate supervised methods into a useable software package.

INSIGHTFUL, INC.
1700 Westlake Ave N, Suite 500
Seattle, WA 98109
Phone:
PI:
Topic#:
(206) 283-8802
Dr. Giovanni Marchisio
OSD 03-H02       Awarded: 08DEC03
Title:Biological Effects Pattern Recognition Tool Using Multivariate Statistical Data Reduction
Abstract:We propose multivariate data reduction, interpretation and pattern matching techniques for classifying and testing genomic, proteomic, and metabonomic profiles against patterns of toxicity. For each type of toxic exposure, these analyses will perform: 1) data normalization and validation; 2) data reduction; 4) class discovery; 5) class prediction. The innovation is in the application of hierarchical clustering and classification methodologies that can produce a list of most diagnostic genes, proteins or biomarkers for each class or type of exposure. We employ powerful literature mining techniques to relate these diagnostic entities to actual networks and mechanisms of action.

INTELLIGENT AUTOMATION CORP.
13029 Danielson Street, Suite 200
Poway, CA 92064
Phone:
PI:
Topic#:
(858) 679-4140
Dr. Dariusz Wroblewski
OSD 03-H02       Awarded: 05DEC03
Title:Toolbox for Analysis of Molecular Expression Data
Abstract:Exposure to toxic chemicals and pathogens, and disease process cause sequential changes in gene and protein expression patterns, and in the metabolic state of the affected cells and tissues. Detection of changes in molecular expressions patterns in the blood or, other easily accessible tissue, should provide means for early identification of the chemical hazards and application of preventive measures. This long-term goal must be supported by development of analysis tools for evaluation and mining of large data sets produced by the molecular expression studies. IAC proposes to develop a comprehensive set of numerical tools for early identification of disease induced by exposure to toxic chemicals, with special emphasis on chemicals producing long-term adverse effects. In Phase I feasibility study, we will address the analysis of relevant gene expression data. The standard methods for biomarker identification in micro-array data will be employed and their performance compared with a number of newly proposed approaches. The molecular expression analysis toolbox will be implemented in the later phases of the proposed development and will also address the needs of proteomic and metabonomic studies. The toolbox will be implemented within open, Web enabled, collaborative environment based on Web services technology

INTELLIGENT OPTICAL SYSTEMS, INC.
2520 W. 237th Street
Torrance, CA 90505
Phone:
PI:
Topic#:
(310) 530-7130
Dr. Roland Suri
OSD 03-H02       Awarded: 08DEC03
Title:Bioinformatic Tool for Assessing Low-Level Toxicity
Abstract:Toxic contaminants are a serious danger to military personnel, with effects that may only appear after weeks, or months. Traditional diagnostic tools focus on levels of specific markers for certain diseases or toxins. When these specific markers do not show high levels, it is difficult to identify exposure to several toxins, each of which may not be toxic alone, but in combination may be extremely dangerous, to genetically predisposed people. Several biotechnology techniques are being developed that gather huge amounts of data in a single measurement, such as genomics (DNA microarrays), proteomics, and metabonomics. Analysis of whole-system data with bioinformatic tools ("systeomics") will provide early warnings of the exposure of personnel to a variety of conditions that may pose future health risks. Intelligent Optical Systems, Inc. (IOS) proposes to develop bioinformatics software that will immediately detect influences of low-level toxicity on soldiers from such data. These bioinformatic methods will provide immediate warnings when personnel have been exposed to low-level toxins, before they experience any symptoms, enabling immediate countermeasures. In addition to military applications, this software will have civilian applications for the early detection of exposure to low-level toxicity, and the detection of drug side effects.

SILICO INSIGHTS, INC.
400 West Cummings Park, Suite 2850
Woburn, MA 01801
Phone:
PI:
Topic#:
(781) 938-3829
Dr. Christos Hatzis
OSD 03-H02       Awarded: 05DEC03
Title:Biological Effects Pattern Recognition Tool Using Multivariate Statistical Data Reduction
Abstract:Application of advanced technologies to facilitate the investigation of mechanisms and pathways associated with toxicant action has recently defined the new discipline of toxicogenomics. The main premise of toxicogenomics is that each class of toxicants elicits a specific and sufficiently distinct and reproducible pattern of global transcriptional response, which corresponds to a specific mechanism of action. Therefore, such a "genetic fingerprint" can be used to discriminate between different classes of toxicants or to predict the toxicity profiles of new suspected toxicants or drug candidates. Protein-level measurements can provide additional information on the toxic response - the term `toxicoproteomics' has been coined to describe this discipline. Furthermore, metabonomic profiling has emerged as a powerful technology to provide time-related quantitative behavior of key cellular pathways. Systems biology provides a comprehensive approach to combining experimental research with a strong computational focus based on network circuitry. Several groups have presented multivariate techniques to delineate a time-dependent physiological response. These analytical techniques can be difficult to interpret and to apply in real-time. Current literature emphasizes the challenge of data compilation into a manageable framework and the need for effective analytical techniques that allow normalization of the multitude of physiological responses.

CHARLES RIVER ANALYTICS, INC.
625 Mount Auburn Street
Cambridge, MA 02138
Phone:
PI:
Topic#:
(617) 491-3474
Mr. Paul G. Gonsalves
OSD 03-H03       Awarded: 08DEC03
Title:Automated Toxicity Hazard Assessment Decision-aiding System (ATHADS)
Abstract:A key stage in military planning is the intelligence preparation of the battlespace (IPB). IPB is a continuous process that defines the battlespace environment and its effects and determines and evaluates likely enemy actions. Typically, characterization of the battlespace and its effects has focused on performing terrain and spatial analysis (delineation of areas of operation and interest, avenues of approach, cover and concealment, etc.), determining effects of weather on military operations (e.g. on mobility), and use of demographic information (population, cultural effects) to guide rules of engagement. However, a critical need exists to integrate the potential effects hazardous chemicals and material may have on mission plans. The presence of industrial toxic chemicals and materials can have a dramatic impact on mission success and is seen as a key impediment to force protection. Here we propose an Automated Toxicity Hazard Assessment Decision-aiding System (ATHADS) to support enhanced mission planning. Our approach leverages three synergistic technologies: intelligent agent-based information retrieval of toxicology data; a rule-based mechanism for event and alert generation based on retrieved information; and a Bayesian belief network approach to environmental hazard assessment and prediction. We see considerable potential for this approach in enhancing mission planning systems for operational forces.

ENTERPRISE WEB MACHINES CORP.
1400 Main Street, Suite 1206
Columbia, SC 29201
Phone:
PI:
Topic#:
(888) 737-1752
Dr. John C. Blackburn
OSD 03-H03       Awarded: 05DEC03
Title:Advanced Force Protection Tools (AFPROT) for Integrated Mission Planning
Abstract:The goal of this project is to develop an XML-driven platform that will rapidly determine the hazards associated with a collection of chemicals at a given location. Beginning with a query consisting of location, date, and time, the CHAIN (Chemical Hazard Assessment Information Network) system will prioritize the materials at the location to provide a ranked list by hazard along with recommended protective measures. This will allow mission planners or emergency response personnel to make decisions based on current materials knowledge and provide an estimate of the associated hazards and priority of those hazards. The system will access site-specific inventory data and individual material hazard information to develop the hazard assessment. In addition, access to GIS and weather data will allow for further modelling of site-site interactions and atmospheric transport of hazardous materials. The system will be developed using using XML to allow ready interchange of data into and out of the model. The formatting of the CHAIN system output will be selectable by the user to provide needed level of material detail, recommended response, and speed in sorting through complicated collections of hazardous materials.

INTELLIGENT AUTOMATION CORP.
13029 Danielson Street, Suite 200
Poway, CA 92064
Phone:
PI:
Topic#:
(858) 679-4140
Dr. Dariuz Wroblewski
OSD 03-H03       Awarded: 05DEC03
Title:Advanced Force Protection Tools
Abstract:Modern battleground includes heavily industrialized areas with heavy concentrations of chemical substances that may potentially be detrimental to the health of military personnel. Currently, only rough evaluations of the health risks can be made and a computerized, near real-time system for risk assessment is needed for advanced and contingency planning of the military operations. The system should have the ability to produce an expert assessment of the health risks to the military personnel based on sound toxicological data and models, and up-to-date operational and environmental information. IAC proposes to develop the Advanced Force Protection Tools (AFPROT) system that uses modular, distributed design based on Web services technology, with standardized (XML-based) data interfaces. The system will integrate toxicological data and models with chemical hazard intelligence, and operational data. Additionally, it will provide the required flexibility in the system configuration, means for integration of legacy software and models, and means for integration with control and command systems employed by the military. In Phase I, we will provide an initial demonstration of the AFPROT platform, which will include prototypes of essential computational modules and integration with existing databases of toxicological data.

BEVILACQUA RESEARCH CORP.
4040 South Memorial Parkway, Suite B2
Huntsville, AL 35802
Phone:
PI:
Topic#:
(256) 882-6229
Dr. David Skipper
OSD 03-H04       Awarded: 31DEC03
Title:Deployable Simulation Training for Operational Medical Personnel & Emergency Responders
Abstract:The goals of this Phase I program will be to investigate the issues associated with the development of a cognitive reasoning system that allows the rapid development and modification of NBC-related medical training scenarios and computer-based simulations support for NBC threat assessments, predictive analysis, and decision-making in both training and operational environments. The heart of the teams approach is the use of an advanced conceptual modeling approach for development of human reasoning software called a Bounded Neural Network (BNN). This specialized intelligent agent architecture and its associated research-level toolset/editor, called CORE, were created by Bevilacqua Research Corporation (BRC), as a quick and efficient way to add complex cognitive processing capabilities to software systems. The team proposes to use Conceptual Graphs as the storage mechanism for the scenario knowledge necessary to create a reasoning and learning environment for medical personnel. The team will define requirements, research issues and prepare a demonstration in Phase I that can be used as the basis of a go-ahead decision for Phase II.

DIAMOND VISIONICS LLC
400 Plaza Drive, Suite-A, PO Box 1276
Vestal, NY 13851
Phone:
PI:
Topic#:
(607) 729-8526
Dr. James P. Pendrys
OSD 03-H04       Awarded: 29DEC03
Title:Deployable Simulation Training for Operational Medical Personnel & Emergency Responders
Abstract:Large numbers of nurses, emergency medical technicians, emergency physicians, Army medics, and medical personnel from different backgrounds need to be trained quickly for chemical , biological and radiological threat detection, assessment and response . The solution to such a complex training problem lies in deploying high-technology based training solutions. A high fidelity intelligent simulation system is proposed here, which will take advantage of the limited amount of expertise existing in the field of CBR threat detection, assessment and management. High cognitive-fidelity realistic simulations will provide realistic trauma training to medical personnel. Simulated case scenarios will train medical trainees at different levels of beginning knowledge to become proficient in CBR threat management. The proposed intelligent simulation system could be rapidly deployed to provide just-in-time training. This unique system will be scalable to allow various types of deployable medical training.

JXT APPLICATIONS, INC.
2673 Commons Blvd, Suite 20
Dayton, OH 45431
Phone:
PI:
Topic#:
(937) 306-5003
Ms. Roberta S. Gearhardt
OSD 03-H04       Awarded: 01JAN04
Title:Deployable Simulation Training for Operational Medical Personnel & Emergency Responders
Abstract:Currently, almost all training for deployable medical personnel occurs in the field through on-the-job training and focuses on the acquisition of procedural knowledge. This severely limits the capability of training organizations to rapidly produce "mission ready" personnel for the field, and hinders mission performance for deploying forces, both ground based medical personnel and air evacuation medical personnel. Medical professionals need a means of accelerating the acquisition of expertise in decision-making and team coordination that underlies responses to chemical, biological and radiological (CBR) threats. This Phase I effort involves the use of cognitive approaches to conduct analyses of current and potential future threat environments where ground based medical crews and aerospace medical personnel are expected to operate. The use of simulation-based training will provide medical personnel with realistic, high fidelity, mission-oriented training in critical medical skills, decision-making and team coordination for emergency response and rapid deployment. The simulations developed will be hosted in multiple delivery media to facilitate their use at the home duty station, on transport aircraft en route to deployment, at ground bases, and in theater. A conceptual high-level design will be presented in a proof-of-concept demonstration of proposed training technologies and simulation content for Phase II development.

MDINFORMATICS, LLC
629 Aloha Rd.
Salt Lake City, UT 84103
Phone:
PI:
Topic#:
(801) 718-4181
Dr. Gregory Thompson
OSD 03-H04       Awarded: 29DEC03
Title:Deployable Simulation Training for Operational Medical Personnel & Emergency Responders
Abstract:Military and civilian medicine currently lack a capability to develop and deploy realistic, mission-oriented training to personnel. Non-physician medical personnel are at increased risk of unpreparedness due to abbreviated training for similar mission-oriented responsibilities as physicians. In this Phase I effort MDInformatics (MDI) proposes to design a Disaster Medicine Core Curriculum (DMCC), a customized dual-pedagogy (didactic presentations + case-based instruction) e-learning platform, and a course content development process to be used in Phase II in order to complete development of the full Disaster Medicine courseware. This project is targeting the > 8 million U.S. medical first responders with a new Web-based instructional platform designed to promote long-term knowledge retention and richer clinical thought processes through the incorporation of analogic case simulations and basic science learning surround. The e-learning platform will include a customized learning management system (LMS) to track user performance. Deliverables for Phase I will include fully detailed investigational study designs to measure the impact of the new system on learner knowledge retention and ability to relate clinical problems to the pathophysiological basis of disease. The new system will represent a major step forward in e-learning technology incorporating important instructional psychological principles and a powerful "learn-and-practice" approach.

ONTAR CORP.
9 Village Way
North Andover, MA 01845
Phone:
PI:
Topic#:
(978) 689-9622
Dr. John Schroeder
OSD 03-H04       Awarded: 07JAN04
Title:Deployable Simulation Training for Operational Medical Personnel & Emergency Responders
Abstract:There is a need in the modern military to train medical personnel rapidly and efficiently to respond to new challenges. The SBIR solicitation noted the need for "advanced simulation capability for use in training the high level decision-making associated with assessment and response to CBR [chemical, biological and radiological] threats" with the goal of providing "medical forces with a deployable capability that can be used at the home duty station, on transport aircraft in route to deployment, for ground bases, and within theater." Our goals are twofold. First, we will develop further technology already produced by Ontar so it can be used as a more effective training tool. Second, we will investigate a specific use of this tool, namely Mental Imagery Simulation-and-Correction (MISC) method. The MISC approach is based on research in cognitive science, and has its roots in the classic TOTE theory of Miller, Galanter and Pribram. Specifically, the system we will develop has the following components: At the conclusion of Phase I we will deliver to OSD a prototype Phase I local wireless ethernet hardware/software system to demonstrate the team training approach.

STOTTLER HENKE ASSOC., INC.
951 Mariner''''s Island Blvd., STE 360
San Mateo, CA 94404
Phone:
PI:
Topic#:
(781) 643-1444
Dr. Eric A. Domeshek
OSD 03-H04       Awarded: 01JAN04
Title:Medical Emergency Team Tutored Learning Environment
Abstract:The U.S. military and civil society both face the increasing prospect that enemies will employ horrific weapons to cause mass casualties. Any effective response to large-scale medical emergencies will hinge on the decision-making and teamwork of medical and emergency professionals, which in turn is tied to the quality of the training those professionals have received. However, beyond relatively low frequency live exercises, this community-including the military medical community-has little effective training for the professional judgement and coordination required in such circumstances. Deployable computer-based simulations with embedded intelligent tutors can address this gap in emergency preparedness. In this Phase I project, Stottler Henke will work with experts on medical emergency response to combine, adapt, and extend selected simulation and Intelligent Tutoring System technologies to develop and prove feasibility for a Medical Emergency Team Tutored Learning Environment (METTLE). METTLE will simulate large-scale medical emergencies-e.g. Chemical, Biological, or Radiological incidents-and tutor individual professionals on the decisions and team interactions appropriate to their roles. To enable easy deployment, standalone operation, and focus on the individual student, METTLE will simulate other team members, team dialogs, and necessary supporting tools. To emphasize professional-level decision-making, METTLE will exploit Socratic-style dialogs with the tutor.

ONTAR CORP.
9 Village Way
North Andover, MA 01845
Phone:
PI:
Topic#:
(978) 689-9622
Dr. John Schroeder
OSD 03-H05       Awarded: 11DEC03
Title:Training Military Medical Personnel for Respond for Directed Energy Weapons and Technology Casualties on Humanitarian Missions
Abstract:Directed energy devices on the battlefield and on humanitarian missions create an increasingly important hazard for our war fighters and for non-combatants. The appropriate education of our war fighters will include the pathophysiology of the effects of directed energy sources with tissue, the sources of exposure (both intentional and accidental) the methods for evaluating injuries, and their medical management. The goal of medical management will be to provide combat medics with the knowledge to rapidly diagnose and treat laser eye injuries on the battlefield to minimize vision loss, to recognize and treat corneal burns, and to treat skin burns due to directed energy sources. The proposed program centers around developing a centralized Directed Energy Combat Casualty Care training program and information process. The training itself will enable casualty care personnel to identify hazards; assess health and psychological effects of directed energy devices; know and be able to use appropriate triage and treatment protocols; be able to apply appropriate treatment (including acute care management and psychological care); and ensure standardization and interoperability of Combat Casualty Care training readiness programs.

SAPIEN SYSTEMS
3510 Ingraham Street Suite 200
San Diego, CA 92109
Phone:
PI:
Topic#:
(858) 775-9755
Ms. Jessica Baldis
OSD 03-H05       Awarded: 10DEC03
Title:Electronic Directed Energy Training System (EDETS)
Abstract:Directed Energy (DE) refers to a wide range of technologies involving concentrated electromagnetic energy or atomic or subatomic particles. DE technologies are rapidly advancing, and practical application of DE technologies is resulting in Directed Energy Weapons (DEWs) and other DE devices (radar, range finders, laser designators, etc.) for military use. As military use of DE technologies increases the likelihood of DE related injuries increases, and combat casualty care (CCC) providers need to be trained effectively. Because the usage of DE technologies is rapidly changing, the amount and availability of medical information and treatment protocols widely differs for different DE injuries. Currently, CCC providers receive minimal, if any, medical training on DE related injuries. This creates a unique challenge for training CCC providers. Sapien Systems proposes to develop a proof-of-concept Electronic Directed Energy Training System (EDETS) for combat casualty care. The EDETS will be an innovative, distributed Computer Based Training (CBT) system which provides DE CCC training through the integration of conceptual, procedural, and case-based learning techniques. EDETS will include a comprehensive DE training curriculum and will be accessible to a wide (Tri-Service) audience with varying levels of expertise. Additionally, EDETS training will be easily modifiable and updatable.

STS INTERNATIONAL
204 Sand Mine Road, PO Box 10
Berkeley Springs, WV 25411
Phone:
PI:
Topic#:
(703) 575-5180
Dr. Charles Stewart
OSD 03-H05       Awarded: 12DEC03
Title:Training Military Medical Personnel for Respond for Directed Energy Weapons and Technology Casualties on Humanitarian Missions
Abstract:STS International proposes to develop combat casualty standards of care for assessing injuries, identifying treatment protocols, and developing triage and care processes related to exposure to Directed Energy (DE) weapons for the purpose of reducing casualties and enhancing unit combat enhancements. STS will also demonstrate the feasibility of converting the training doctrine into high fidelity, mission oriented, multi-media based training simulations for response to Directed Energy Weapon threats. The primary objective of these training simulations will be to teach medical care providers the critical decision making skills needed to effectively detect the threat, assess the situation, and treat casualties injured in a DE event. . In order to properly respond to a DE threat, medical professionals must possess an array of knowledge, skills and experience such that the proper critical decisions can be made when faced with varying risks and courses of action. This situational awareness is a complex cognitive activity that cannot be acquired through training that focuses on the acquisition of procedural knowledge. Simulation training allows students to be engaged in situations that challenge their cognitive, psychomotor and affective domains thus facilitating the development of both experience and judgment.

BARRON ASSOC., INC.
1160 Pepsi Place, Suite 300
Charlottesville, VA 22901
Phone:
PI:
Topic#:
(434) 973-1215
Dr. B. E. Parker,
OSD 03-H06       Awarded: 31DEC03
Title:Predicting the Cognitive Performance of Individuals and Teams
Abstract:The objectives of this SBIR Phase I research program are to design and implement an unobtrusive methodology that can be used to rapidly and easily predict individual and team cognitive readiness to perform critical missions. The methodology proposed herein provides a rich decisional framework for cognitive readiness assessment in complex environments. Among the attributes offered by the approach are design flexibility (e.g., in choosing the measures that correlate with cognitive performance), dynamic updating, automatic and proper handling of inconsistency and uncertainty, and the ability to facilitate team decision making and understanding. As part of the Phase I effort, a prototype Self Team Assessment Tool (STAT) for assessing individual and team cognitive readiness will be developed and tested. In conducting the R&D program, Barron Associates, Inc., a company with a strong track record in the proposed decisional methodology, will collaborate with Dr. Stephanie Guerlain of the University of Virginia (UVA), an expert in cognitive engineering and team evaluation methodology. Testing of the Phase I prototype STAT tool will be conducted on teams of UVA Naval Reserve Officer Training Corps (NROTC) midshipmen as subjects.

JXT APPLICATIONS, INC.
2673 Commons Blvd, Suite 20
Dayton, OH 45431
Phone:
PI:
Topic#:
(937) 306-5003
Mr. Terrell W. Fulbright
OSD 03-H06       Awarded: 01JAN04
Title:Predicting Cognitive Performance of Deploying Health Teams
Abstract:The deployment of geographically dispersed military medical teams requires personnel to be prepared to manage all threats and employ all contingencies. Currently, these medical teams and their missions are comprised of individuals who differ in competency skill levels within the field of medicine as well as allied fields of intelligence, communication, and security across domains of air, land, sea and space. These rapidly generated teams may originate from different geographical locations and varied operational missions. Further, individual team members are likely unfamiliar with each other, have limited knowledge of the deployment location, and are not trained in the specifics of the tasks to be performed. It is essential that cognitive assessment of mission readiness begin at initiation of the assignment and continue through arrival in the theater of operation. This effort is to design and develop unobtrusive techniques and methodologies to predict individual and team readiness to deploy and perform critical mission objectives. We will investigate methodologies and techniques that hold promise to non-invasively measure the cognitive capability of an individual to comprehend and carry out medical mission requirements under varying threats and contingencies, and use these measures to predict the cognitive readiness of individuals and teams for deployment.

ONTAR CORP.
9 Village Way
North Andover, MA 01845
Phone:
PI:
Topic#:
(978) 689-9622
Dr. John Schroeder
OSD 03-H06       Awarded: 07JAN04
Title:Predicting Cognitive Performance of Deploying Health Teams
Abstract:The goal of the proposed research is precisely as described in the SBIR solicitation, namely to develop "methodologies and techniques to non-invasively measure the cognitive capability of an individual to comprehend and carry out medical mission requirements under varying threats and contingencies and use these measures to predict cognitive readiness of individuals and teams for medical readiness." Such methodologies and techniques are increasingly needed in the military, as well as many other walks of life. Human error cuts across multiple domains and results in an incalculable loss of productivity and even human life every year. A recent article by Sandra I. Erwin (National Defense, October 2002) estimates the financial cost of aviation mishaps in the Navy at $4.3 billion over five years, with an additional $20 to $30 billion in indirect costs such as litigation, investigation and program delays. According to Circadian Technologies, Inc. (Lexington, MA), the total cost of human performance decline due to shiftwork fatigue in the U.S. is $77 billion (2001 estimate). This breaks down to $64.5 billion in lost productivity and $12.5 billion in health-related costs (see ). We will design a portable system (the Performance Assessment Laboratory or PAL) that monitors both psychophysiological indices (heart rate, blood pressure and skin conductance) and cognitive focus (a short executive function and attention task) in order to provide an "early warning alert," indicating when a worker is suffering from stress-related deficits, such as fatigue, that may affect performance. The "cognitive focus" (CF) task will rely on auditory stimuli that can be processed in any setting, via a discrete ear plug. One strength of our system is that it will combine predictive power from all of these measures. Another advantage is that the assessment is quick. The information can be used by the workers themselves, either alone or as part of a team, or by a supervisor. In all cases, workers can be warned to pay additional attention and take extra care, or better yet, take a break, consume food or caffeine, or even take a nap. In the initial phase of our research, we focus on one example of mental stress: sleep deprivation. This is an important and common source of human error and in many cases individuals suffering from sleep-deprivation are not aware of their cognitive deficits. PAL will assess when cognitive performance is in fact significantly impaired by sleep deprivation, and will be capable of doing so even if individuals claim to be "just fine". The ability to test individuals before they initiate a complex task may prevent serious mistakes as well as expensive delays due to slower reaction time. This approach can also be applied to improving interactions among members of a team. Other factors, such as hunger, anxiety, exposure to noise, or high intake of caffeine or other substances may also affect both cognitive performance and psychophysiological responses - and our system should be effective in predicting decrements in task performance that arise from a wide variety of stressors. In Phase I of this research, we will test 20 participants under conditions of sleep deprivation versus being fully rested, and use PAL to predict performance decrement on a criterion task (a stylized medical problem-solving task). In future phases, we will extend our research to include greater numbers of participants, other sources of stress, and other criterion tasks, such as group-problem solving; we will test PAL in a real-world situation as well.

SA TECHNOLOGIES, INC.
4731 East Forest Peak
Marietta, GA 30066
Phone:
PI:
Topic#:
(410) 838-3748
Dr. Cheryl Bolstad
OSD 03-H06       Awarded: 31DEC03
Title:Predicting Cognitive Performance of Deploying Health Teams
Abstract:Determining if a soldier is cognitively ready to perform his or her job poses a considerable challenge to the research community. This need is especially apparent in the military medical community where soldiers must be ready and able to handle a multitude of conditions and problems at any one moment. This SBIR seeks to employ state of the art techniques in Cognitive Task Analysis, Situation Awareness and Instructional Design to identify and eventually develop comprehensive cognitive readiness measures for military medical teams. These measures will have far reaching implications and can ultimately be used to not only evaluate cognitive readiness, but can also provide comprehensive diagnostic information and structured feedback to the soldiers and their superiors. This structured feedback can be used in training applications or can be used to determine when personnel rotation is called for.

CLINICALTOOLS, INC.
431 W. Franklin St. #24
Chapel Hill, NC 27516
Phone:
PI:
Topic#:
(919) 960-8118
Dr. T. B. Tanner
OSD 03-H07       Selected for Award
Title:Distributed Medical Training for Force Mobilization and Disaster Response
Abstract:This SBIR application proposes to develop, evaluate, and implement Internet-based continuing education in trauma care to train physicians and other health professionals without specific trauma expertise. The courses will be available via the Internet and will utilize interactive clinical scenarios to disseminate knowledge. Online information resources and discussion capability will accompany each course. The curriculum development will be based on reviews of existing trauma care curriculum guidelines, surveys of the skills and needs of physicians in non-trauma specialties, and revised by trauma specialists and military physicians. This curriculum will provide physicians and other health professionals with training that will enable them to transfer the skills and knowledge they acquire from our interactive training scenarios to new situations. The course design will support this goal by applying the principles of cognitive flexibility theory, which seeks to address issues of advanced knowledge acquisition. Cognitive flexibility theory is especially useful when the knowledge to be acquired is nonlinear, complex, and integrated, such as medical training.

SAPIEN SYSTEMS
3510 Ingraham Street Suite 200
San Diego, CA 92109
Phone:
PI:
Topic#:
(858) 775-9755
Ms. Jessica Baldis
OSD 03-H07       Selected for Award
Title:Interactive Integrated Training System (IITS)
Abstract:Mobilization of physicians and other medical personnel for military missions and operations creates a substantial training challenge. Large numbers of medical personnel with varying specialties are rapidly mobilized. These personnel often encounter medical problems outside their area of expertise with which they have little experience. This gap in existing knowledge and skills and the knowledge and skills demanded can result in inappropriate or inadequate medical care. A training system needs to be developed that can be rapidly deployed to a large number of users to augment their existing medical knowledge and skills allowing them to successfully complete their jobs once mobilized for missions and operations. Sapien Systems proposes to develop a proof-of-concept Interactive Integrated Training System (IITS). IITS will be an innovative, distributed Computer Based Training system to augment existing medical knowledge and skills through the integration of conceptual, procedural, and case-based instruction. An innovative approach will be applied to the case-based instruction through the addition of "simulation elements" and interactive problem solving designed to pinpoint and address flaws in the student's knowledge or decision-making.

SONALYSTS, INC.
215 Parkway North, P.O. Box 280
Waterford, CT 06385
Phone:
PI:
Topic#:
(860) 326-3621
Ms. Margaret Bailey
OSD 03-H07       Selected for Award
Title:Distributed Medical Training for Force Mobilization and Disaster Response
Abstract:Training is critically needed to prepare physicians of various specialties to perform trauma care when mobilized on military missions or operations. The goal of Sonalysts' and the University of Maryland's Medical Center is to design a scenario-driven, real-time, simulation-based training system to close the gap between prior experience and likely future demands on physicians' skills. Our research will focus specifically on training rapid evaluation of potential abdominal trauma by conducting a Focused Assessment with Sonography for Trauma (FAST) examination. We will examine the feasibility of integrating ultrasound simulations with existing medical, case-based, electronic tutoring software. We will conduct a task analysis of FAST medical diagnosis and design a training system that incorporates the latest in instructional methods. We will foster apprenticeship learning through a learn-by-doing approach, with training scenarios presented in a narrative format to provide engaging and authentic context. An online coach will provide analogical comparisons, automatically generated hints and immediate remedial feedback. Identifiable features on the ultrasound images will contain active links, and normal and abnormal patient ultrasounds will be viewable for comparison. Students may select virtual patient views from any of the predetermined FAST viewpoints. Self critiques may be compared with computer-generated intelligent critiques during opportunities for reflection.

THINKSHARP, INC.
539 Rock Spring Road
Bel Air, MD 21014
Phone:
PI:
Topic#:
(410) 893-5338
Mr. William Sacco
OSD 03-H07       Selected for Award
Title:Distributed Medical Training for Force Mobilization and Disaster Response
Abstract:Non trauma surgeons mobilized to support military operations are often ill-prepared for the challenges of trauma surgery. Training to date has not bridged the gap between their experience and abilities and the rigor of their assignment. The objective of this research is to develop a training system that utilizes best known instructional practices and exploits advanced, distributed training technology to deliver comprehensive trauma surgery preparation. The components of this system are in place, and this research will focus on their integration. A comprehensive set of almost 200 trauma decision trees, developed for the Navy (Contract N66001-85-C-0297), will provide the foundation for the training content. An existing, robust online training delivery system that imbeds best teaching practices will provide the technology for delivery. The research will result in a case study based, online learning system that provides experiential learning to mobilized non trauma surgeons. Case studies will simulate combat trauma and disaster scenarios. Interactive and self paced instruction will provide challenges to surgeon trainees to make medical treatment decisions and to react to nuances in treatment response. Implementation would cost effectively prepare mobilized non trauma surgeons to face trauma scenarios, and to utilize proper treatment decision making.

ONTAR CORP.
9 Village Way
North Andover, MA 01845
Phone:
PI:
Topic#:
(978) 689-9622
Dr. John Schroeder
OSD 03-H08       Awarded: 07JAN04
Title:Facilitating Post-Deployment Data Mining to Evaluate War-Related
Abstract:There is a need in the modern military to optimize the post-deployment evaluation of war-related illness of unclear etiology. The phenomenon of war-related "medically unexplained symptoms" (MUS, ICD-9 code 799.6) has been a persistent problem throughout the history of combat. Elucidation of the time course (onset, changes over time) and etiology (cause/effect) in these cases has been hampered by the lack of a longitudinal patient record spanning the pre-, during, and post-deployment continuum. Retrospective analysis should rely on real longitude medical data. The military healthcare administration also needs to determine the required fields in military electronic Data Theater in order to force input of those data to benefit the evaluation and also to update the Defense Medical Surveillance System (DMSS) database. MUS modeling will contain three types of information: patient medical history, physical and psychosocial information. The patient medical history should only contain information related to diagnosis of MUS; physical information should include epidemiology, immunization information about the soldier; and as well as patient psychosocial information should have psychosocial evaluation results of the soldier. Combination of three areas data together will help to evaluate MUS. We will have a spreadsheet of minimal required MUS data elements by the end of Phase I. In order to collect those data, two interfaces have to be developed to retrieve data from CHCS II and DMSS (Defense Medial Surveillance System). We are going to build a prototype of web service interfaces to access those two data repositories to demonstrate the proposed program on Phase I. The combination of telemedicine and medical data mining technology will be a powerful tool for evaluation of MUS. We will accomplish this by utilizing two technologies that we have in place at Ontar. The first is called IMD2S (Cognitive Integrated Medical Data Display System) which is a data mining software built on top of EMRs. The second system is called ComPResS (Computerized Patient Records System). ComPResS assists physician by providing EMRs on handheld devices in a wireless environment.

SILICO INSIGHTS, INC.
400 West Cummings Park, Suite 2850
Woburn, MA 01801
Phone:
PI:
Topic#:
(781) 938-3844
Dr. Christos Hatzis
OSD 03-H08       Awarded: 15JAN04
Title:Data Mining Infrastructure to Characterize Medically Unexplained Symptoms
Abstract:War-related syndromes have existed throughout modern wars; however, studies to-date have not provided evidence to point to a specific group of symptoms and their causes. Nevertheless, these studies have confirmed increased rates of reported symptoms and general deterioration of perceived health by war veterans. Military health care systems today offer unique opportunities in monitoring and predicting health-related events of military personnel. A closer investigation of war-related symptoms is warranted to evaluate the adequacy of the information collected and the analytical approaches employed. A data mining infrastructure can facilitate (i) collection of specific fields not previously gathered, and (ii) avoid forced entry of data fields from field personnel that are not relevant. Existing biosurveillance systems are well suited for counts-based analyses and efficient data collection. They do not offer analytical techniques to visualize dependencies between known and unexplained symptoms. New approaches in assessing information content will allow examination of syndromes and their associations with medical and non-medical inputs. Further efforts can then refine the architecture and apply a suite of analyses to provide decision management tools. The availability of such a system can not only leverage health care data for disease modeling but also enable timely response management.

INHAND ELECTRONICS, INC.
30 W. Gude Drive, Suite 100
Rockville, MD 20850
Phone:
PI:
Topic#:
(240) 558-2014
Mr. Boris Donskoy
OSD 03-H09       Awarded: 29DEC03
Title:A Modular, Battery-Free, Wireless Information Carrier for Storing Long-Term Medical Records
Abstract:InHand Electronics (Rockville, MD) proposes the development of a state-of-the-art wireless information carrier, a so-called Medical Dogtag or MTAG. The MTAG will meld a combination of technologies - a passive radio frequency identification (RFID) technology and a bus-powered Universal Serial Bus (USB) interface, to create a battery-free flexible storage device capable of storing up to 20 years of textual and graphical medical record data. InHand Electronics will leverage its expertise in developing extremely small, low-power, low-cost electronics platforms in the development of the MTAG. The key features of the MTAG are its low production cost, its ability to derive all required power for basic functions from outside sources and thus avoid the need for batteries, and its modular architecture which facilitates upgrades and improvements as technology advances occur. InHand will perform a complete system, hardware, and packaging design of the MTAG in Phase I, and will perform detailed manufacturability analyses, including cost estimates in low- and high-volume production. The Phase I design will form the launching point for Phase II prototyping and design refinement.

IRVINE SENSORS CORP.
3001 Redhill Avenue, Building #3
Costa Mesa, CA 92626
Phone:
PI:
Topic#:
(714) 435-8920
Mr. John Leon
OSD 03-H09       Awarded: 31DEC03
Title:Development of Wireless Electronic Information Carrier
Abstract:The objective of the Office Of The Secretary Of Defense OSD-DH09 SBIR effort is to develop a secure, encrypted, wirelessly enabled, high capacity electronic information carrier. Trauma care provisioning either in the battlefield or in residential neighborhoods is a time critical process. Multiple personnel may evaluate the same patient giving different prognosis based on the patient's current state and evaluator's point of view. Newer drugs being introduced into the market increases an individual's chance of survival, but an individual may have allergic reactions to these new drugs, which may not be understood by an evaluators general examine. An individual longitudinal medical record, which stores years of an individuals medical history is required. The initial point of such record for an individual is their personal identification card (PIC). Our approach is to incorporate the needed functionality (wireless and security) needed for evaluators (such as a medic in the battlefield) to be incorporated in the size look and field of a "dog tag" by evaluating existing current devices and deriving a solution that grows and scales as technology evolves. Our investigation will lead to the development and implementation of a prototype in Phase II called Irvine Sensor Longitudinal Medical Record (ISCLMR).

PHYSICAL OPTICS CORP.
Information Technologies Division, 20600 Gramercy
Torrance, CA 90501
Phone:
PI:
Topic#:
(310) 320-3088
Dr. Andrew Kostrzewski
OSD 03-H09       Awarded: 31DEC03
Title:Wireless Personal Information Carrier System
Abstract:The Office of the Secretary of Defense is seeking to develop a new type of personal identification card (PIC) for the patientcentric data network and database, containing a complete longitudinal patient record up to 20 years long. Coupled with the Battlefield Medical Information System Tactical (BMIST), individual medical data often needs to be accessed and updated by medical personnel even when real-time connectivity to a database is unavailable. To meet thisthese requirements Physical Optics Corporation (POC) proposes to develop a new Wireless Personal Information Carrier (WPIC) than will exceed these requirements by introducing innovative hybrid power supplies, including both passive (inductive power transfer), and active (efficient battery) systems with minimal data transfer times. The proposed WPIC will have an ultrahigh effective data communication rate, achieved by modifying Bluetooth technology to exceed 10 Mbps. It is based on POC's unique lossless and perceptually lossless data/voice/image compression, which reduces data transfer time by a factor of five compared to existing systems and will support expansion of the WPIC to include high-quality audio, high-resolution images, and video sequences. With efficient power management of hybrid inductive and battery-based power supplies, WPIC can operate for over five years operation without battery change.

TRITON SYSTEMS, INC.
200 TURNPIKE ROAD
Chelmsford, MA 01824
Phone:
PI:
Topic#:
(978) 250-4200
Mr. Bill Bergeron
OSD 03-H09       Awarded: 31DEC03
Title:Development of Wireless Electronic Information Carrier (1000-309)
Abstract:Triton Systems, in partnership with a leading edge information security firm, proposes to develop a small, lightweight, robust, durable, and encrypted wireless Electronic Information Carrier (EIC). In Phase I, Triton proposes to design and test a system. During Phase II, we will produce prototypes for the US military. We anticipate that, prior to the end of the Phase II effort, we will be in commercial production supply for non-military applications. The manufacturing scale-up will allow experience-curve based cost reduction and prototype refinement specific to the needs of the US military during the Phase III effort. This dual-use device will have a similar size and weight of traditional military "Dog Tag", and have capability of securely storing a full 20-year longitudinal medical history including high-resolution graphical images. This write-able system will allow the combat medic or military medical personnel to update the medical history, record current assessment of patient condition, recent care provided, and evacuation information. Additionally the device will allow for patient tracking, and current location information, throughout the evacuation chain, in order to provide continuity of care for the patient.

CONCEPTUAL HEALTH SOLUTIONS
6951 Forest Glade Cr.
Dallas, TX 75230
Phone:
PI:
Topic#:
(507) 284-1551
Dr. Peter L. Elkin
OSD 03-H10       Awarded: 31DEC03
Title:Data Mining Longitudinal Clinical Data to Detect Adverse Drug Events
Abstract:Conceptual Health Solutions (CHS) technology specializes in the conversion of free-text to structured data for medical data mining and automated processing. CHS will adapt its existing technology to develop a core data infrastructure for the automation of the evaluation of longitudinal clinical data from dynamic DoD health information systems to detect adverse events after use of drugs, vaccine, or other medication. The technology involves medical concept-based indexing, "natural language processing" and controlled medical vocabularies including SNOMED-CT, MedDRA and ICD to normalize data in medical records for data mining of ADE indicators. Automated analysis tools will process potential ADEs and compare results with other ADE monitoring sources such as the FDA. CHS will combine its medical practice specific data management technology with Mayo Clinic''''''''''''''''s medical research and ADE subject matter expertise. The resulting data infrastructure will allow DoD to determine which adverse events put patients at significant risk or increased healthcare expenditure and enable DoD to take evidence-based, targeted risk mitigation steps.

LINCOLN TECHNOLOGIES, INC.
40 Washington Street, Suite 220
Wellesley, MA 02481
Phone:
PI:
Topic#:
(781) 237-8722
Mr. David Fram
OSD 03-H10       Awarded: 01JAN04
Title:Mining Longitudinal Data for Known and Unknown Adverse Events
Abstract:Longitudinal clinical data resources (e.g., the Composite Health Care System) are important large-scale repositories of data from real-world medical practice that can provide significant new insights regarding the safety of approved drugs and vaccines, through the application of automated data mining. Such "secondary" use depends, however, on overcoming substantial information-handling obstacles: the original data, collected primarily for patient care and administration, are highly complex and fine-grained and unsuitable for direct analysis. The project objective is to develop innovative software capable of (1) extracting and standardizing relevant information from medical records in military health systems, (2) utilizing temporal abstraction and pattern matching to identify medical record segments that correspond to potential drug-related adverse events, (3) providing alerts to physicians when "labeled" adverse reactions are so identified, and (4) performing aggregate analyses to detect drug-event associations with greater-than-expected frequencies that may signal previously uncharacterized adverse reactions. Phase I focuses on developing temporal abstraction, pattern matching, and disproportionality analysis technologies that show promise in identifying adverse drug events when applied to moderate-scale test data. Phase II focuses on prototype completion, scale-up, and validation, on automating data extraction and on comparing results with those from other data sources (clinical trials, post-marketing surveillance, chart review).

ONTAR CORP.
9 Village Way
North Andover, MA 01845
Phone:
PI:
Topic#:
(978) 689-9622
Dr. John Schroeder
OSD 03-H10       Awarded: 07JAN04
Title:Data Mining Longitudinal Clinical Data to Detect Adverse Drug Events
Abstract:The proposed program will develop a data mining tool to detect both known and unknown safety signals in the CHCS databases. The proposed program will accomplish this through careful, innovative data preparation (code mapping, redundancy removal), advanced screening algorithms, user-friendly interface and cognitive report displays. The overall goal is to provide an active surveillance system that can detect unknown adverse events, monitor known adverse events and report ADEs in a timely fashion so that the physicians can rapidly access information which serves as an alert or reminder in their practice so the soldiers using drugs or vaccines can receive the appropriate risk management and care. At the conclusion of Phase I we will deliver to OSD a prototype system (ADE-Discovery) that demonstrates the feasibility of detecting both known and unknown safety signals in the CHCS database.

STOTTLER HENKE ASSOC., INC.
951 Mariner''''s Island Blvd., STE 360
San Mateo, CA 94404
Phone:
PI:
Topic#:
(206) 545-1478
Ms. Laurie Spencer
OSD 03-H10       Awarded: 01JAN04
Title:Incremental Online Data Mining for Adverse Drug Events
Abstract:Patients are at risk of adverse reactions to and interactions between medical drugs. Though databases exist for reporting of adverse drug events (ADEs), more must be done to speed the discovery of "safety signals" and protect patients and physicians alike. Stottler Henke has devised a novel approach for learning safety signals directly from longitudinal patient data, such as that in the military's Composite Health Care System. Using rules, heuristics, and machine learning techniques, patient data will be cleaned and transformed into ADE-like sequences. Higher than expected associations between drugs and adverse events will be discovered using MGPS, a data mining tool whose development was sponsored by the FDA. Stottler Henke's Safety Signal Miner will extend the MGPS algorithm to work in "online" fashion so that it can process new data as it arrives, saving computing resources and speeding discovery. Preprocessing and safety signal detection will pass several automated, self-validation checkpoints; moreover, Safety Signal Miner will provide validation assistance to the operator via interactive exploration, record drill-down, and summarization by strata and ICD9 codes. Safety signal miner should have broad commercial appeal to hospitals, clinics, and research institutions.