R&D Contracting Directorate, Bldg 7, 2530 C Street, WPAFB, OH 45433-7607

A -- SENSOR TECHNOLOGIES INTEGRATION LABORATORY SOL PRDA No. 97-18-AAK POC Contact Alan Struckman, Contract Negotiator, 937-255-2902 or Vicki A. Fry, Contracting Officer, 937-255-2902 WEB: click here to view the supplemental package, www.wl.wpafb.af.mil/contract/hp.htm. E-MAIL: click here to contact the contract negotator, struckac@qq.wpafb.af.mil. INTRODUCTION: Wright Laboratory (WL/AAKR) is interested in receiving proposals (technical and cost) on the research effort described below. Technical and cost proposals in response to this Program Research and Development Announcement (PRDA) shall be received by 08 Dec 97, 1500 Eastern Time, addressed to the attention of Mr. Alan Struckman, WL/AAKR, Building 7, 2530 C Street, Wright-Patterson AFB, OH 45433-7607. This is an unrestricted PRDA. Small businesses are encouraged to propose on this PRDA. Proposals submitted shall be in accordance with this announcement. Proposal receipt after the cutoff date and time specified herein shall be treated in accordance with restrictions of FAR 52.215-10; a copy of this provision may be obtained from the contracting point of contact. There will be no other solicitation issued in regard to this requirement. Offerors should be alert for any PRDA amendments that may permit subsequent submission of proposal dates. Information incorporated in a supplemental package is necessary for proposal preparation. The Wright Laboratory Guide titled "PRDA and BAA Guide for Industry," dated November 1992, also provides information specifically designed to assist offerors in understanding the PRDA proposal process. The "PRDA and BAA Guide for Industry" and the supplemental package of proposal information are available through either of the following methods: (1) Download from the Internet -- The "PRDA and BAA Guide for Industry" is available on the Wright Laboratory R&D Contracting Home Page (www.wl.wpafb.af.mil/contract/hp.htm) under "A Contracting Toolbox," "PRDA/BAA Indexed Guide"). The supplemental package is available on the same home page. If using this method to obtain the information, please confirm receipt by providing your name, address and telephone number via e-mail or fax to Alan Struckman, Contract Negotiator. Fax number is (937) 255-3985. E-mail address is struckac@aa.wpafb.af.mil. (2) You may submit a request for the packages in writing if you are unable to access or download the information from the home page. Written requests should be addressed to Mr. Struckman, WL/AAKR, Bldg 7, 2530 C Street, Wright-Patterson AFB OH 45433-7607, telephone (937) 255-2902. B -- REQUIREMENTS: (1) Technical Description: Wright Laboratory intends to award one or more Indefinite Delivery/Indefinite Quantity (ID/IQ) contracts to design solutions for Automatic Target Recognition (ATR) and Sensor Fusion (SF) technology domains for the Wright Laboratory Combat Information Technology Division (WL/AAC) and any successor organization. This program, called Sensor Technologies Integration Laboratory (STIL), has the objectives to: Design, develop, improve, analyze, model, test, evaluate and demonstrate sensor databases, development and evaluation environments, systems, algorithms, and technologies; Apply advanced engineering and computer technologies to improve capability, reliability, maintainability, and affordability throughout the ATR and SF hardware/software life-cycle. Research and development shall be performed in the following seven topic areas: (a) Data Management -- Topic Area 01: WL/AAC requires the development of innovative techniques for data collection planning and execution, ground truthing, massive data storage, data basing and retrieval, and data distribution to meet the requirement for an on-site heterogeneous Automatic Target Recognition (ATR)/Sensor Fusion (SF) database and database tools capable of providing data to WL and other DoD programs performing ATR and SF research. The contents of this database to consider, but not be limited to, include: Imagery Intelligence (IMINT) and Signal Intelligence (SIGINT) products; synthetic signatures; on and off board sensor data; performance results; ground truth; and collateral data (i.e. DMA, LANDSAT, models.) In order for the database to fulfill data requirements of an ATR/SF R&D effort, a full understanding of the multiple data types and data collecting sensors (EO, IR, hyperspectral, multi-frequency RF) is necessary. Cost-effective and intelligent strategies for collecting ATR and SF developer/evaluators requirements and data, that adequately promotes ATR development, training, testing and evaluation is of interest. Participation in data collection planning to enhance single and multiple ATR and sensor fusion application programs will result in increased utility of collected data. Paramount in such planning is execution flexibility in the face of equipment and logistical problems and delays. Collection episodes can be single sensor or include multiple sensor phenomenologies collected either in series or simultaneously. Such planning will include overall, end-to-end life cycle cost for the complete data management activity. Innovative techniques will be required to obtain ground and image truth information, as well as screening, characterizing, and databasing data collected from highly dynamic (moving target), time and space varying episodes using multiple sensors over highly disparate collection geometries and time scales. Such datais especially difficult to correlate and associate as to content in space and time. Understanding sensor and data phenomenology provides the necessary knowledge to develop image metric tools for identifying image level artifacts and anomalies, and to inform ATR developers and evaluators of these sensor anomalies and data artifacts within the imagery prior to algorithm developers/evaluators receiving the data. Data characterization tools are also required to show evaluators what data exists in the database in order to develop sequestered data sets for independent evaluation and for assessment of the kinds of evaluations to be performed with data. With the ability to accurately collect ground and image truth of time/space variant and dynamic targets, the need exists to continuously enhance the current heterogeneous ATR and SF database to handle temporal and geo-registered data from multiple sensors, and fused data from advanced systems. Engineering judgment is required to determine the optimal data base structure given the data characteristics, such as relational, object oriented, or flat files or other structure or hybrid combinations thereof. Knowledge of existing environments (i.e. Khoros, CORBA, C, Matlab) is required for data tool or toolkit insertion to promote ATR development and evaluation. In order to enhance ATR and SF Research, design and development of intelligent techniques for data access and distribution are required. These intelligent techniques include but are not limited to: visually organizing large amounts of data to quickly identify pertinent data before disseminating it; facilitating this visualization process by utilizing the latest data visualization and human factors research into the design; and developing intelligent product agents to pull together related relevant data that may be useful to the user. Intelligent product agents should include but not be limited to: techniques for either intelligent data pulling, where the user interactively works with the visualization and product agent tools; and intelligent data pushing, where pertinent dataset(s) are identified, gathered and disseminated by the database automatically up to the data recipient's appropriate security level. In order for intelligent push and pull technology to be successful, users must be able to remotely access or query the WL/AAC database. Therefore, techniques are required to disseminate data at the appropriate security level via the World Wide Web up to the highest classified network available, or appropriate electronic media. Additional requirements for the data management effort are to field technical questions from data recipients, as many users tend to be narrow experts in a related technology and lack the insight into the data collection and generation process that would allow them to state their requirements in a succinct and efficient manner, provide hardcopy documentation when appropriate and to update data and dataset description as applicable via the WWW. (b) Laboratory Environments -- Topic Area 02: WL/AAC has a requirement for an onsite world-class ATR and SF research and development laboratory environment. The contractor shall develop a laboratory architecture compatible with but not limited to: the Common Object Request Broker Architecture (CORBA); the Defense Model and Simulation Office's (DMSO) High Level Architecture (HLA); the Joint Modeling and Simulation System (J-MASS), and Khoros. CORBA will allow network services and applications to be viewed as objects that share common interfaces, rather than treated as separate entities that have to be linked directly in fixed, hard-coded relationships. Khoros is an integrated software environment that couples strong scientific data visualization and processing with software development tools and application programming interfaces (APIs) to allow easy user extensibility. The architecture shall consider information distribution, as STIL will contain large databases that will distribute data electronically, and information security considerations, as STIL will contain data from unclassified public releasable, to International Traffic in Arms Regulation (ITAR) data, to top secret/SCI requiring division-wide sharing, contractor/government interchange, and outside world access. These requirements will drive design issues regarding World Wide Web encryption and design, and firewall concerns, including integration of secure networking through the firewall, and demilitarized zones. In addition, viewing network traffic by applications rather than protocols, segmenting the network into zones, going with virtual LANs where appropriate, and exploring VPDNs (virtual private data networks) are actions to consider. Optimization of supercomputer architecture and interconnects with workstations for a variety of computer systems ranging from desktop workstations to state-of-the-art parallel and multiple processor systems are of high interest. Packages such as CORBA, X-Windows and Khoros will be available to develop software graphical user interface tools. Connectivity is required between STIL and: High Performance Computing (HPC) sites; the World Wide Web; the Office of Secretary of Defense's (OSD's) Virtual Distributed Laboratory (VDL); and classified networks (i.e. Secret Internet Protocol Routed Network (SIPRNET), Joint Worldwide Intelligence Communication System (JWICS)). In considering network design, the Distributed Object Model (DOM) paradigm applies the well-established benefits of object-oriented programming, including rapid development, reusability, and built-in security to the enterprise. DOM allows distributed objects to interact without knowing anything about their location. The environment contractor shall be responsible for integrating tools, algorithms/systems, models and databases into the STIL environment, working with other contractors awarded under this PRDA via an associate contractor clause. This will involve designing common wrappers or interfaces to allow algorithms to communicate successfully with other environment objects. Ultimate goals include the integration ofthe STIL environment into other modeling and simulation environments such as the Avionics Collaborative Engineering Environment (CEE). This might entail a common infrastructure for distributed loosely coupled applications. Multiple Object Request Brokers (ORBs) may be required because of the heterogeneous nature of computing environments. Long term goals of a real-time environment are important. (c) ATR/SF Evaluation -- Topic Area 03: The contractor shall perform tasks in the actual evaluation of air-to-air and/or air-to-ground target recognition algorithms and/or sensor fusion algorithms. The evaluation methodology should encompass ATR/SF algorithms, single and multi-sensor data inputs, correlation and tracking functions, real-time SF systems, and performance estimation of ATR/SF systems. Evaluation tasks include but are not limited to: experiment design, considering performance requirements, availability of measured and synthetic data, and algorithm functionality; data collection activities, including establishing data requirements, selecting targets and backgrounds, and documenting ground and air truth; development of performance metrics, considering program requirements and standard community metrics and developing innovative metrics as required; execution of tests, including accessing data from the WL database or other sources, running algorithms on that data, and scoring, analyzing, and reporting the results; and performance estimation, including extrapolating results through system modeling. Single and multi-sensor algorithms will employ data from a variety of sensors, including synthetic aperture radar (SAR), high range resolution radar (HRR), forward looking infrared (FLIR), electro-optic (EO) multi-spectral, and signal intelligence. An evaluation methodology is required which addresses the differences in phenomenology, acquisition geometry, and sensor design while maximizing standardization in tool development, performance metrics, and results presentation. Multi-sensor applications will pose challenges to the algorithm evaluator in the areas of experiment design and understanding the impact of registration errors. An approach is required which addresses difficulties in assembling multi-sensor data sets and considers approaches to developing measured, synthetic, or hybrid data sets. An approach is required to quantify the effect of reference based and image/signal based misregistration on algorithm performance. Test procedures will be adapted to specific targeting missions, including adaptation of software models for weapon delivery constraints and time lines, as well as, aircraft survivability, as required to predict increased mission effectiveness provided by the ATR/SF systems. (d) Target Modeling -- Topic Area 04: Innovative solutions in the areas of computational electromagnetic (CEM) predictions for complex targets, integration and application of target phenomenology and signature exploitation technology, computer aided design (CAD) model development and construction technologies, and validation technologies are sought. In the area of CEM predictions for complex targets, expansion of the current state of the art for prediction techniques involving complex targets is desired. Innovative solutions and computer codes that employ asymptotic methods, frequency domain methods, and time domain methods should be investigated, as well as efficient and accurate methods of combining low and high frequency techniques into hybrid prediction codes demonstrating a major increase in CEM capabilities. The contractor also should consider ways to hybridize measured data with predicted data to form a more robust overall solution. CEM techniques that predict radiation, as well as scattering phenomena, involving whole targets or components should be investigated. The CEM techniques developed in all cases should strive for rapid generation of solutions as a function of frequency and incident and/or observation angle. Stochastic and perturbation techniques that address the technical challenge of augmenting static deterministic synthetic data to reflect the dynamic nature of ATR scenarios, target configuration variation, and target intra-class variability where appropriate should be examined. An additional area of interest is the investigation and demonstration of methods to synthesize scattering from ground clutter in both rural and urban environments. The contractor should demonstrate performance of CEM codes on multiple high performance computing (HPC) platforms and workstations, and provide and demonstrate technologies which stress workstation to HPC platform connectivity and functionality. It is strongly encouraged that prediction codes be written using an object oriented design and C++ programming language. Target phenomenology and signature exploitation technology that can be integrated into promising identification algorithms should be identified. The goal of phenomenology/signature exploitation is to develop additional robust sources of target information for use in ATR development. CAD model construction techniques and technologies which improve their accuracy or fidelity, increase the level of automation, and decrease the time required to build CAD geometry models for complex targets are sought. Efficient and accurate techniques of sampling target geometry and materials from a variety of sources including photographs, line drawings, blue prints, scale models and the actual target body should be considered. These construction techniques will be demonstrated via building multiple CAD geometry models. A variety of CAD development software should be used to demonstrate these advanced construction techniques including but not limited to ACAD, BRL-CAD, and Euclid. In the validation technologies area the contractor should investigate, develop, and demonstrate efficient, computer based, validation methods that perform quantitative comparisons of truth data to synthetic data. This truth data can take many forms such as radar cross section measurements; measured antenna radiation patterns; physical target dimensions, details, andmeasurements; or observed or measured data resulting from target interrogation by active or passive means. Synthetic data is the obvious predicted or computer generated counterpart to truth data (e.g. radar cross section predictions or CAD target models). The comparison metrics should allow for rapid determination of error margins. Technology which indicates means to reduce these error margins should also be developed and demonstrated. The contractor should investigate, develop, and demonstrate methods to incorporate CEM advancements, signature exploitation techniques, CAD model geometries, and validation technologies into 1-on-1 and few-on-few modeling and simulation (M&S) tools. These enhanced M&S tools will be used to demonstrate the value that improved phenomenology modeling provides to the overall simulation results. (e) Radar Algorithm Development -- Topic Area 05: The objective of this thrust area is the development of automated detection, tracking, discrimination, classification and identification algorithms using radar sensors. The scope includes both air and ground targets and includes both high frequency (10 GHz and above) and low frequency (100 MHz -- 1 GHz). The signatures of interest include wide band signatures such as high range resolution signatures and resonance signatures, imaging signatures such as SAR and ISAR signatures, and multi-modal signatures such as full polarimetric, wide angle, and multi-aperture (e.g. interferometric SAR). A key element of the algorithm development will include the use of EM modeling to understand the phenomenology basis of the signature characteristics, to motivate sound algorithms based on fundamental EM principles, and to generate synthetic signatures for use, along with measured data, in algorithm development. Both model-based and learning-based algorithms are desired with particular interest in hybrid algorithms that leverage the respective strengths of each basic approach. Algorithm approaches that scale to difficult problems (problems that havelarge combinatorics -- e.g., target type x configuration x articulation state x obscuration x ) are of particular interest. Algorithm approaches should scale favorably in the following dimensions: the amount of measured data required for training, the amount of memory required to store target representations, and the amount of computation required to recognize the target. Sublinear growth with problem complexity is desired for each of these dimensions. END OF PART 1 (0294)

Loren Data Corp. http://www.ld.com (SYN# 0005 19971023\A-0005.SOL)