Search

The authors present a strategy for systematically allocating resources to competing independent research and development (IR&D) projects in a company building systems for the US Department of Defense. The approach is less mathematically complex than many techniques reported in the literature, utilizing the judgement of key business and technical elements in the organization. The authors first give a taxonomy characterizing various approaches and briefly review R&D selection methods reported in literature. Then they describe an interactive process that they have developed for selecting IR&D projects which, rather than substituting complex calculations for good technical and material judgement, facilitates better utilization of the expertise of various organizational elements within the context of corporate strategic planning. Procedures are included for idea submittal and evaluation. Experiences are discussed obtained during the application of the process to a corporation with multiple business units. >

One of the challenging issues in a distributed time sensitive information rich environment is how to assist the decision makers to make decisions quickly and effectively. This paper describes a decision model that has been developed for such situations and how using the RCAST agent architecture can assist the humans in dealing with the information challenges. The example demonstrates how the system can be used in a military scenario.

Rapid developments in sensor technology and its applications have energized research efforts towards devising a firm theoretical foundation for sensor management. Ubiquitous sensing, wide bandwidth communications and distributed processing provide both opportunities and challenges for sensor and process control and optimization. Traditional optimization techniques do not have the ability to simultaneously consider the wildly non-commensurate measures involved in sensor management in a single optimization routine. Market-oriented programming provides a valuable and principled paradigm to designing systems to solve this dynamic and distributed resource allocation problem. We have modeled the sensor management scenario as a competitive market, wherein the sensor manager holds a combinatorial auction to sell the various items produced by the sensors and the communication channels. However, standard auction mechanisms have been found not to be directly applicable to the sensor management domain. For this purpose, we have developed a specialized market architecture MASM (Market architecture for Sensor Management). In MASM, the mission manager is responsible for deciding task allocations to the consumers and their corresponding budgets and the sensor manager is responsible for resource allocation to the various consumers. In addition to having a modified combinatorial winner determination algorithm, MASM has specialized sensor network modules that address commensurability issues between consumers and producers in the sensor network domain. A preliminary multi-sensor, multi-target simulation environment has been implemented to test the performance of the proposed system. MASM outperformed the information theoretic sensor manager in meeting the mission objectives in the simulation experiments.

Antiterrorist analysts often need to work in teams with the requirement to analyze voluminous amounts of dynamic information in order to assess potential terrorist threats. Analysts have a high cognitive demand complicated by factors that typically the information has restrict access and requires special expertise for interpretation. The goal of this research is to enhance team performance by modeling and implementing a cognitive agent architecture capable of proactively seeking, linking and sharing information using knowledge and experience distributed among team members. The agent architecture is empowered by a collaborative RPD model - a novel team-based naturalistic decision making process derived from Klein's recognition-primed decision framework.

Recent advances in acoustic sensors, microprocessors, local wireless communications, and low-cost global positioning satellite (GPS) receivers allow the creation of acoustic sensor networks for target detection, tracking, and identification. In particular we describe the concept of robust acoustic sensors that exhibit the ability for self-calibration and self-location. Use of such sensors in a wireless, internetted development provides accurate multi-sensor target tracking and classification. This paper describes the concept and design of such sensors and networks. Particular attention is paid to how the environment affects the sensor performance and system performance. Initial simulator results are presented. >

Fuzzy logic techniques have become popular to address various processes for multisensor data fusion. Examples include: (1) fuzzy membership functions for data association, (2) evaluation of alternative hypotheses in multiple hypothesis trackers, (3) fuzzy-logic-based pattern recognition (e.g., for feature-based object identification), and (4) fuzzy inference schemes for sensor resource allocation. These approaches have been individually successful but are limited to only a single subprocess within a data fusion system. At The Pennsylvania State University, Applied Research Laboratory, a general-purpose fuzzy logic architecture has been developed that provides for control of sensing resources, fusion of data for tracking, automatic object recognition, control of system resources and elements, and automated situation assessment. This general architecture has been applied to implement an autonomous vehicle capable of self-direction, obstacle avoidance, and mission completion. The fuzzy logic architecture provides interpretation and fusion of multisensor data (i.e., perception) as well as logic for process control (action). This paper provides an overview of the fuzzy logic architecture and a discussion of its application to data fusion in the context of the Department of Defense (DoD) Joint Directors of Laboratories (JDL) Data Fusion Process Model. A new, robust, fuzzy calculus is introduced. An example is provided by modeling a component of the perception processing of a bat. >

Multi-sensor data fusion is an emerging technology applied to Department of Defense (DoD) areas such as automated target recognition, battlefield surveillance, and guidance and control of autonomous vehicles, and to non-DoD applications such as monitoring of complex machinery, medical diagnosis, and smart buildings. Techniques for multi-sensor data fusion are drawn from a wide range of areas including artificial intelligence, pattern recognition, statistical estimation, and other areas. This paper provides a tutorial on data fusion, introducing data fusion applications, process models, and identification of applicable techniques. Comments are made on the state-of-the-art in data fusion.

As we approach the millennium, numerous programs and research plans have focused on the use of advanced technology to improve the effectiveness of military forces. Examples include: the Army XXI concepts and the Army After Next (http://cacfs.army.mil/aan.html) concepts for military technology beyond the year 2020. Technologies to improve the effectiveness of soldiers include advanced tactical communications, the Global Positioning Satellite (GPS) system for precise geolocation, fire and forget smart weapons, improved sensors, and multi-sensor data fusion systems. The interest in such technology is exemplified by the recent movie, Starship Troopers, based on a science fiction book by Heinlein. This paper summarizes some of the innovations provided to soldiers to transform them from the era of GI Joe to an anticipated era of the starship trooper. In particular, we analyze information technologies such as advanced sensors, multi-sensor data fusion, and automated reasoning.

Water absorption spectroscopy has been successfully demonstrated as a sensitive and accurate means for in situ determination of temperature and H2O mole fraction in silica (SiO2) particle-forming flames. Frequency modulation of near-infrared emission from a semiconductor diode laser was used to obtain multiple line-shape profiles of H2O rovibrational (v1 + v3) transitions in the 7170-7185-cm(-1) region. Temperature was determined by the relative peak height ratios, and XH2O was determined by use of the line-shape profiles. Measurements were made in the multiphase regions of silane/hydrogen/oxygen/ argon flames to verify the applicability of the diagnostic approach to combustion synthesis systems with high particle loadings. A range of equivalence ratios was studied (phi = 0.47 - 2.15). The results were compared with flames where no silane was present and with adiabatic equilibrium calculations. The spectroscopic results for temperature were in good agreement with thermocouple measurements, and the qualitative trends as a function of the equivalence ratio were in good agreement with the equilibrium predictions. The determinations for water mole fraction were in good agreement with theoretical predictions but were sensitive to the spectroscopic model parameters used to describe collisional broadening. Water absorption spectroscopy has substantial potential as a valuable and practical technology for both research and production combustion synthesis facilities.

In recent years, numerous prototypical systems have been developed for multisensor data fusion (DF). A paper by Hall and co-workers1 describes more than 50 such systems developed for Department of Defense (DoD) applications some 15 years ago. Such systems have become ever more sophisticated. Indeed, many of the prototypical systems summarized by Hall and co-workers1 use advanced identi cation techniques such as knowledge-based or expert systems, Dempster-Shafer interface techniques, adaptive neural networks, and sophisticated tracking algorithms.

Multisensor data fusion systems seek to combine information from multiple sources and sensors in order to achieve inferences that cannot be achieved with a single sensor or source. Applications of data fusion for Department of Defense (DoD) applications include automatic target recognition (ATR), identification-friend-foe-neutral (IFFN), and battlefield surveillance and situation assessment. The use of data fusion for these applications is appealing. Conceptually, the use of a broad spectrum of sensors should improve system accuracy, decrease uncertainty, and make these systems more robust to changes in the targets and environmental conditions. Techniques for data fusion are drawn from a diverse set of disciplines including signal and image processing, pattern recognition, statistical estimation, and artificial intelligence. Many of these techniques have an extensive history, ranging from Bayesian inference (first published in 1793) to fuzzy logic (originating in the 1920s) to neural nets (developed in the 1940s). In the past two decades an enormous amount of DoD funds have been expended to develop data fusion systems. While there are many successes, there are still a number of challenges and limitations. Indeed, critics of data fusion argue that data fusion technology is disappointing and ask, "why is it that when all is said and done (in data fusion), there is so much more said than done?" This paper presents a summary of the current state and limitations of data fusion. Key issues are identified that limit the ability to implement a successful data fusion system.

Developed a formalization of CBM technology and engineering. Established test beds, tools, and processes for CBM. Created new methods for observing, detecting, and characterizing mechanical failures. new architectures and methods for automated reasoning. Developed new mathematical formulations and models for failure prediction. Demonstrated viability of hierarchical machinery prognostics. Established enabling technology for CBM transition and implementation using a open system architecture. S&T base of data, models, tools, and training material available to CBM community.

The ONR Initiatives grant is aimed at supporting basic research initiatives to the U. S. Navy. These initiatives encourage interaction between Penn State's Applied Research Laboratory (ARL) researchers and Penn State faculty members, and they provide opportunities for student research. During 1998, the ONR Initiatives program included five research projects and a High School Student Intern Program. The Objectives of the FY 98 ONR Initiatives Projects are as follows: Turbulent Effects on Sound Propagation -- This effort investigated electromagnetic (EM) wave propagation through a realistic, turbulent atmosphere that includes both large-scale deterministic features (.1 approximately equal 1000 km) as well as smaller scale, stochastic turbulent features (1 100 m); Turbulent Flows over Rough Walls -- This continues research in modeling and experimental data collection related to fluid flow near rough surfaces (for realistic Reynolds numbers); Micro-scale Heat Exchangers and Pumps -- This continues a project initiated in FY 97 to design, model, and fabricate a microscale power source; Coordination Science -- This continues a project initiated in FY 97 to develop fundamental techniques and associated computer technology to support greatly improved coordination in simulation-based design and acquisition of complex systems; Nonlethal Technologies -- This is a new project to investigate key technologies related to nonlethal defense, such as the capability for remote sensing and activation using coordinated semi-autonomous mobile devices (e.g., robots, aircraft, or underwater vehicles); and High School Intern Program -- this is a Summer enrichment program for high school students to encourage an interest in science and engineering.

One of the top-level processing functions defined in the Joint Directors of Laboratories data fusion processing mode is Level 4 processing. This is a meta-process that monitors the on-going data fusion process to optimize the fusion of data. To date, two basic approaches have been used for Level 4 processing. The first approach uses classic constrained optimization methods to optimize a measure of performance of the data fusion process. The second approach treats Level 4 processing as a classical control problem, invovling dynamic sensor tasking, to meet performance objectives. Both of these approaches generally ignore the human-in-the-loop user, and treat sensor as relatively unintelligent devices. Recent development in smart sensor, improved models of dynamic sensor performance, and advances in cognitive psychology suggests that a new perspective on Level 4 processing is needed. In this paper, the concept of Level 4 processing is extended to exploit these rapid evolutions, resulting in increased performance of data fusion systems.

In a previous paper (Hansen et al., 1995), a conceptual framework for developing a true prognostic or predictive diagnostic capability was described. The current paper expands on this framework by describing micro-mechanical and dynamic models, sensors and data fusion, signal processing, approximate reasoning, distributed architecture, and human factors research and development being conducted to provide such a capability for a broad range of applications. These include both autonomous and man-in-the-loop decision making about maintenance actions and local and geographically distributed monitoring and data analysis architectures.

In recent years, numerous multisensor data fusion systems have been developed for a wide variety of applications. Defense related applications include; automatic target recognition systems, identification-friend-foe-neutral, automated situation assessment and threat assessment systems, and systems for smart weapons. Non-defense applications include; robotics, condition-based maintenance, environmental monitoring, and medical diagnostics. For each of these applications, multiple sensor data are combined to achieve inferences which are not generally possible using only a single sensor. Implementation of these data fusion systems often involves a significant amount of effort. In particular, software must be developed for components such as data base access, human computer interfaces and displays, communication software, and data fusion algorithms. While commercial software packages exist to assist development of data bases, communications, and human computer interfaces, there are no general purpose packages available to support the implementation of the data fusion algorithms. This paper describes a visual programming tool developed to assist in rapid prototyping of data fusion systems. This toolkit is modeled after the popular tool, Khoros, used by the image processing community. The tool described here is written in visual C, and provides the capability to rapidly implement and apply data fusion algorithms. An application to condition based maintenance is described.© (1996) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Terfenol-D is a `giant' magnetostrictive material offering mechanical strains on the order of 1000 X 10-6 m/m. Dynamic transducers constructed using Terfenol-D as the motion source offer displacements based on approximately +/- 500 (mu) strain. These transducers are known to be nonlinear; however, in some applications they can be treated as linear systems and their behavior approximated by the classic pair of linear transduction equations. This approximation is made in this study to facilitate analysis. Several simple analog and digital controllers are used with acceleration feedback to improve transducer linearity. A nonlinearity of particular importance when using Terfenol-D transducers is wave form distortion. The distortion is a result of nonlinear strains vs. magnetization relationships and the magnetic hysteresis occurring within the Terfenol-D. The net result is the presence of integer harmonics in the transducer voltage, current, and output velocity. Assuming the harmonics to be disturbances, an expression is developed for predicting the change in harmonic amplitudes of acceleration as a function of frequency and parameters for the controller, load, and transducer. Experimental measurements comparing controlled and uncontrolled output accelerations are presented to validate the approach. A significant extension of the linear range of transducer behavior, due to feedback control, is also demonstrated.© (1996) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

An experimental approach is used to identify Terfenol-D material properties under magnetic bias and mechanical prestress conditions typical of transducer applications for the magnetostrictive material Terfenol-D. The approach is based on low signal transduction models, a mechanical model of the test transducer, and on the theory of vector impedance and admittance analysis. The material properties being investigated, measured directly or calculated from measured quantities, are: two Young's moduli, magnetomechanical coupling factor (or 'figure of merit' of the material), linear coupling coefficient or axial strain coefficient (d33) and two magnetic permeabilities. Electrical impedance and admittance and acceleration per unit current complex functions are measured at frequencies from 100 Hz to 5 kHz using a swept sine excitation at five applied magnetic field levels. Ten Terfenol-D samples are used in randomized performance tests to obtain the preliminary material property information that is presented. Of these ten rods, two were used in randomized performance repeatability tests for which data is also presented. These data sets demonstrate the dependence of material properties on applied magnetic field levels, and provide a preliminary assessment of the variability in material properties given known operating conditions.© (1996) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

We are working to understand Terfenol-D's behavior in dynamic applications. Experimental results demonstrate that the amplitudes of harmonics present in Terfenol-D transducer output displacement spectra increase with increasing drive-current amplitude. An empirical mathematical model which predicts harmonics is presented. The model approximates both the nonlinear strain versus magnetic flux density behavior (c vs B), and the magnetic hysteresis (B vs H) of the material. Trends are shown for displacement versus drive-current amplitude as a function of frequency. Generally, displacements increase with increasing drive-current amplitude, and decrease with increasing drive frequency (when operated away from mechanical resonance). Single frequency drive-current results are compared with the trends predicted by the mathematical model.

Multi-sensor data fusion technology is rapidly maturing. In recent years, numerous algorithms and techniques have been introduced or applied ranging from classic estimation and statistical methods (e.g., Bayesian inference), and pattern recognition methods, to heuristic techniques such as expert systems and templating techniques. Because of the range and variety of techniques, Hall and Linn (1990) developed a taxonomy which maps data fusion functions to classes of algorithms and to specific techniques. This paper describes a survey of commercial software tools applicable to multi-sensor data fusion. The paper identifies specific software programs, summarizes the required computing resources (hardware and software environment), and identifies the source of the software. The survey maps the computer software to the data fusion taxonomy, establishing the relationships between data fusion levels, functions, and algorithms. The intent of this paper is to allow the data fusion community to readily access these software building blocks without reinvention. Thus, building new data fusion systems may be performed with significant use of commercial off the shelf software.© (1993) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Multisensor data fusion integrates data from multiple sensors (and types of sensors) to perform inferences which are more accurate and specific than those from processing single-sensor data. Levels of inference range from target detection and identification to higher level situation assessment and threat assessment. This paper provides a survey of more than 50 data fusion systems and summarizes their application, development environment, system status and key techniques. The techniques are mapped to a taxonomy previously developed by Hall and Linn (1990); these include positional fusion techniques, such as association and estimation, and identity fusion methods, including statistical methods, nonparametric methods, and cognitive techniques (e.g. templating, knowledge-based systems, and fuzzy reasoning). An assessment of the state of fusion system development is provided.© (1991) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

A table of asymptotic relative efficiencies of a large variety of R-estimators is given and some of its uses discussed.The family of distributions used includes both those common in the robustness literature and those that correspondent to various rank tests for location.

Simulated environments are suitable for test and validation of certain classes of expert systems. This paper provides an introduction to the problems involved in the expert system test and validation process and describes the use of simulations in this process. Particular attention is given to mapping results obtained in simulated environments into real-world expectations. Simulation testing of a signal analysis expert system provides a source of examples.

There has been considerable recent discussion as to what location estimator one should use. Traditionally, the sample mean has been used, often in conjunction with formal or informal outlier rejection rules. However, there may be considerable loss in efficiency if the mean is used on data for which it is not appropriate. This expository article seeks to provide insight into the estimation problem. It focuses on the all pervasive importance of f'/f in efficient estimation of location, where f is the density function of the distribution generating the data and f' is its derivative. Connections are discussed among the three major classes of location estimators dominant in the robustness literature: M estimators (maximum likelihood-like), R (rank) estimators, and L estimators (linear combinations of order statistics). In all three classes f'/f is seen to be the key to efficient estimation. A heuristic explanation is given as to why f'/f is an intuitively reasonable quantity on which to base estimatio...