Rome Laboratory/PKPX, 26 Electronic Parkway, Rome NY 13441-4514

A -- SPECIAL NOTICE: ROME LABORATORY'S FY 97 SBIR TOPICS SOL SBIR TOPIC 20 POC Joetta A. Bernhard, Contracting Officer, A/C 315-330-2308; Margot Ashcroft, SBIR Program Manager, 315-330-1793. PART 5 OF 6. ROME LABORATORY'S FY 97 SBIR TOPIC. ROME LABORATORY IS PLEASED TO MAKE AVAILABLE THE FOLLOWING SMALL BUSINESS INNOVATIVE RESEARCH (SBIR) PROGRAM TOPICS, AS FOLLOWS: SBIR TOPIC (20) - TECHNICAL POINT OF CONTACT: Albert A. Jamberdino, RL/IRA, 315-330-4581. TITLE: Information Storage and Retrieval - Optical Memories. OBJECTIVE: To design and experimentally validate concepts for volumetric optical storage that demonstrate an optical advantage, i.e. massively parallel I/O, storage capacities of 10E10-10E12, and access times of milliseconds. DESCRIPTION: Rome Laboratory is investigating the use of photonic technology in advancing the state of the art in data storage. Optical memories show promise in many areas of the data storage hierarchy. Applications include: archival storage, random access memory, read only memory, cache memory, and associative (content addressable) memory. Three dimensional optical memory offers the potential of terabit storage in volumes on the order of a cubic centimeter. High data transfer will be crucial for military applications as well as civilian uses. This initiative is directed towards exploiting the ''Optical Advantage'' of storing digital data in the form of optical volume or 3-dimensionally. Concepts such as content addressable memory either numerical, textural or image identification techniques can be implemented in memory, results isolated, and effectively provide acceleration of output speed, and access time. Correlation, auto-correlation, and change detection concepts within the memory itself should also be exploited.- PHASE I: Identify and characterize candidate media, lens architecture's, or beam steering concepts to provide storage capacities of 10E10-10E12 bits per cubic centimeter, or at least 10E3-10E6 discrete locations per centimeter. POTENTIAL COMMERCIAL MARKET: Image exploitation would be greatly enhanced by the development of faster storage devices not to mention the benefits of terabit of data accessible at any instant. Medical data will benefit from the advancement of these technologies as well. Imagine your entire medical history available to a physician in another town should medical attention be necessary away from home. A library of X-ray files stored digitally that not only are available on demand, but now that images are stored digitally, a computer would assist the doctor in detecting tumors earlier than would have been possible before. The development of the ''Information Superhighway'' will hinge on the development of memory systems capable of storing more data than ever before as well as transferring that data faster than ever before. SBIR TOPIC (21) TITLE: Intelligent Desktop Assistant. TECHNICAL POINT OF CONTACT: Patrick McCabe, RL/IRDS, 315-330-3222. OBJECTIVE: Develop an intelligent desktop computer assistant to autonomously access, evaluate, retrieve, and fuse information from the growing number of on-line information sources available. DESCRIPTION: The intelligent desktop assistant (IDA) will use machine learning techniques to familiarize itself with user styles, techniques, preferences and interests. IDA will be able to guide users through the process of on-line information source selection, utilization, and interaction management (i.e. cost tracking, query refinement, etc.). Documentation will not be required as IDA adjusts user interface characteristics to reflect the changing experience of the user. IDA will schedule and execute multiple information retrieval tasks in accordance with user priorities, deadlines, and preferences. The software will be capable of evolving the processes by which it interacts with other systems, learning the characteristics of their interfaces and languages.- PHASE I: Prototype user and system interfaces and identify the learning algorithms required to support both. Develop a mechanism to specify deadlines, so that tasks can be completed in accordance with user defined priorities. POTENTIAL COMMERCIAL MARKET: The potential commercial market is literally the size of the personal computer market. This capability would be highly useful to any individual with a computer that is connected to any network. SBIR TOPIC (22) TITLE: Single-Channel Spectral Characterization. TECHNICAL POINT OF CONTACT: Andrew Noga, PhD, RL/IRA, 315-330-4581. OBJECTIVE: The objective of this effort is to further develop the Adjustable Bandwidth Concept Signal Energy Detector (U.S. Patent 5,257,211) for enhancement to existing and well known spectrum analyzer designs. DESCRIPTION: Rome Laboratory, through a series of contractual and in-house efforts, has developed a variety of approaches to characterize signal activity in user selectable RF band segments. In particular, proof-of-concept, FFT (Fast Fourier Transform)- based, single channel signal energy detectors (i.e., spectrum analyzers) have been implemented, which are capable of detecting and ''grouping'' dispersed signal energy to form estimates of the composite bandwidth, center frequency, and signal-to-noise ratios. While multi-channel approaches exist and have many desirable characteristics, these techniques likewise require orders of magnitude increases in implementation complexity, size, maintainability and cost, over that of single channel techniques. Similarly, Time-Frequency distribution approaches, while useful in many scenarios, also have rather complex processing requirements. Currently available commercial spectrum/signal analyzers, including those that employ FFT-based Fourier analysis, lack the composite signal grouping and parameter estimation reporting functionalities. With such capabilities, the enhanced spectrum analyzer becomes an extremely useful RF spectrum characterization device. The ''Adjustable Bandwidth Concept (ABC) Signal Energy Detector'', shows particular promise as a technique for spectral analysis enhancement, and is the focus of this research effort. Essentially, this technique allows for the grouping and parameter report generation of signals with various bandwidths within the RF segment analyzed. Specifically, the ABC detector allows for averaging narrowband signals more over time and less over frequency, while simultaneously averaging wideband signals less over time and more over frequency, for improved detection, grouping and parameter estimate performance.- PHASE I: In this phase of the effort, a proof-of-concept implementation of the ABC Signal Energy Detector will be developed (e.g., in the MATLAB language) for development purposes. At the same time, a theoretical analysis will be performed to optimize the ABC detection algorithm performance. The signal parameter report contents and format shall be developed in this phase. Basic user definable device settings along with appropriate operator interface considerations will be addressed. The intent of this phase is to lead into the design and development of a real-time implementation in phase II, which leverages currently existing spectrum analyzer and associated equipment designs. DUAL-USE COMMERCIALIZATION POTENTIAL: In recent years the private sector has witnessed a burgeoning growth in the sales and service of personal communications equipment, including such devices as the cellular telephone and Global Positioning System (GPS) receivers. This is in addition to the plethora of devices already common in the private sector, with purposeful and/or unintentional RF generation abilities. Both the communication engineer and the electromagnetic compatibility engineer, designing to meet FCC regulations, have come to rely upon various laboratory bench tools to properly develop and test their designs. In particular, the traditional spectrum analyzer is indispensable. The enhancements to the traditional spectrum analyzer as identified in this research have the potential to decrease communication equipment development time and cost, via superior signal analysis capability and automated specification testing. Regulatory agencies can likewise benefit from the development of the enhanced spectral analysis device. SBIR TOPIC (23) TITLE: Automated Information Extraction Tools. TECHNICAL POINT OF CONTACT: Carrie Pine, RL/IRA, 315-330-4518. OBJECTIVE: To develop Natural Language Understanding (NLU) tools that automate the extraction of information from unformatted text, and the porting of these capabilities to new domains. DESCRIPTION: The Information Age has brought with it the need for tools to help people exploit the overwhelming volumes of textual information now available to them. Unformatted text is a rich source of potential information. Tools that automatically find and extract simple data from unformatted text could be of enormous value to any industry or organization that deals with large volumes of textual information. Data that can feasibly be extracted from text includes ''shallow'' information, such as the names of people, places, locations, organizations, and equipment, quantities, and dates. The ability to extract such data and to put it into a structured form would enables a multitude of powerful applications, including automatic document indexing, automatic data base generation, and data visualization (for analytical purposes). Tools would also be developed to make the toolset portable to new domains, so it could easily learn about and recognize previously unseen information (e.g., people's names for a new country, equipment names related to a new application area, etc.)- PHASE I: Research the area of information extraction. Develop and refine the requirements for ''shallow'' information extraction tools, tools capable of recognizing and extracting simple (i.e., feasible) data from text, such as names, places, locations, organizations, and equipment, quantities, and dates. Include requirements for making the tools portable to new application domains. DUAL-USE COMMERCIALIZATION: Tools for shallow information extraction would be useful in any industry or organization that processes large volumes of unformatted text. This includes financial institutions (e.g., tracking competitors/business intelligence analysis), law enforcement agencies (e.g., automatic data base generation from police reports, visualization of data from unformatted text to help investigators analyze crime data), and the publishing industry (e.g., automatic document indexing). SBIR TOPIC (24) TITLE: Exploitation of GPS Controlled Imagery TECHNICAL POINT OF CONTACT: Mark Rosiek, RL/IRRE, 315-330-7787,. CRITICAL TECHNOLOGY: The Department of Defense (DOD) developed the Global Positioning System (GPS) to aid navigation. This system is fully operational and in use for both military and civilian applications. One GPS application is in the acquisition of airborne imagery. Air vehicles are equipped with GPS and imaging sensors. When airborne imagery is acquired GPS measurements are recorded to capture the exterior orientation information of the imaging sensor. OBJECTIVE: This effort will develop new and innovative methods, techniques or products that take advantage of GPS controlled imagery. DESCRIPTION: Technology developed under this effort will replace traditional techniques for estimating airborne imagery exterior orientation parameters before exploiting the imagery. This will result in improved techniques for producing current imagery products. GPS controlled imagery will also support the development of new imagery products that were not feasible with current techniques. GPS controlled imagery provides an estimate of the airborne imagery exterior orientation parameters. Technology developed under this effort will take advantage of the GPS information. These techniques will assist in working with blocks or strips of airborne imagery, imagery acquired at different times or from different sensors, and metric information available with GPS controlled imagery. The new products produced will take advantage of GPS controlled imagery to produce image mosaics, change detection, new methods to catalog imagery or obtain metric information.- Phase I should focus on defining, developing a specification and demonstrating the method, technique or product that is proposed. This should result in showing the required input information, an estimate of the processing time required and the accuracy of the output information. The contractor should supply any GPS controlled imagery required to support any proposed demonstration. POTENTIAL COMMERCIAL MARKET: This topic has high Dual Use Commercialization Potential. Both the military and the commercial sectors have sources for GPS controlled imagery. Although the sources may differ, methods to exploit the imagery can be similar. The commercial sector can utilize image mosaics, image catalogues, for resource management and the military can use image mosaics, image catalogs for wide area search applications. Imagery acquired at different times or from different sensors can be exploited for change detection by the commercial sector for detecting storm damage or environmental health. The military can use change detection for detecting changes in military activity. (0229)

Loren Data Corp. http://www.ld.com (SYN# 0002 19960819\A-0002.SOL)