Dept of Systems Engineering and Management

The TIMIT and KING databases are used to compare proven spectral processing techinques to an auditory neural representation for speaker identification. The feature sets compared are linear prediction coding (LPC) cepstral coefficients and auditory nerve firing rates using the Payton model (1988). Two clustering algorithms, one statistically based and the other a neural approach, are used to generate speaker-specific codebook vectors. These algorithms are the Linde-Buzo-Gray algorithm and a Kohonen self-organizing feature map. The resulting vector-quantized distortion-based classification indicates the auditory model performs statistically equal to the LPC cepstral representation in clean environments and outperforms the LPC cepstral in noisy environments and in test data recorded over multiple sessions (greater intra-speaker distortions)

While classification has been practiced within the fields of biology and chemistry for hundreds of years and has enjoyed mathematical definition in the form of numerical taxonomy for approximately 50 years, this paper presents an inaugural method and application of quantitative classification of systems to the discipline of systems engineering. This paper establishes, by theory and application, that a quantitative, business process-constrained system classification enjoys a host of systems engineering uses—most notably system design and evaluation. A survey of 50 years of published material in the systems science and systems engineering disciplines on qualitative system classifications is provided to ground the discussion of this paper's quantitative system classification method. Future research topics, such as interoperability measurement, branching from the work described in this paper are also proposed. © 2008 Wiley Periodicals, Inc. Syst Eng

In this paper, we distill the most significant current systems engineering challenges by examining acquisition oversight reports, aircraft mishap investigations, and current systems engineering literature. We then discuss how to better meet those challenges in light of the domains of human systems integration; specifically, manpower, personnel, training, human factors, health, safety, habitability, survivability, and environment. The findings in this paper indicate that the system engineering technical processes are fundamentally sound, but that the application of those processes is frequently inconsistent and ineffective. Causes for this include a need for sound systems engineering earlier in system development, more quantitative methods and tools to support the application of technical processes, and more effective management of interfaces. This is observed in both the commercial and government sectors, and the consequences have proven to be very costly and dangerous.

Security is often treated, whether intentionally or otherwise, as something which can easily be added after a system is built. This has proven in many cases to cause substantial embarrassment and cost. Security must be designed into a system or enterprise, and integrated across all subsystems from the beginning. However, the tools and methodologies to perform securityrelated design are not well developed, defined, standardized, nor integrated. There are many possible ways that security can be modeled. Part of the problem is that security is sometimes considered a nonfunctional or performance system requirement, sometimes a functional system requirement, and sometimes simply an operational mission requirement. This paper explores ways to model each of these using current methodologies, and suggests a unifying theme for modeling security as a whole. Descriptions of how security can be modeled in DoDAF using structured and object oriented techniques are presented, how it is modeling in other frameworks and how the use of Colored Petri-Nets can be a potential candidate for security architectures.

The DoD has frequently demonstrated its ability to procure phenomenal systems; however, these accomplishments are often tarnished by substantial cost and schedule overruns. While defense acquisition policies are continually being revised to address these perennial problems, many believe that a more fundamental source of these overruns is the lack of flexibility in the systems being developed, which tend to preclude effective responses to unexpected events. However, providing justification to invest in flexibility is a tough sell when the measure of value is a military capability or political outcome, as there is no extant method to demonstrate the potential return on investment. This paper introduces a decision tool for valuing the inherent ability of different systems or designs to respond to uncertainty. The proposed tool is essentially a modification of the current life cycle cost model and is premised on the notion that the need for capability changes in a system arises in a stochastic manner that can be incorporated into a continually updated, expected value model presented in terms of total life cycle cost. The cost-based decision tool presented here quantifies the ability of competing designs to respond to these capability changes via a cumulative distribution function (CDF). The design with the most favorable CDF (i.e., the one that is most likely to meet the most likely set of requirements at the lowest expected value curve of life cycle cost) is deemed to be the "best" design.

For decades, the DoD has employed numerous reporting and monitoring tools for characterizing the acquisition cost of its major programs. These tools have resulted in dozens of studies thoroughly documenting the magnitude and extent of DoD acquisition cost growth. Curiously, though, there have been extremely few studies regarding the behavior of the other cost component of a system's life cycle: Operating and Support (O&S) costs. This is particularly strange considering that O&S costs tend to dominate the total life cycle cost (LCC) of a program, and that LCCs are widely regarded as the preferred metric for assessing actual program value. The upshot for not examining such costs is that the DoD has little knowledge of how LCC estimates behave over time, and virtually no insights regarding their accuracy. In recent years, however, enough quality LCC data has amassed to conduct a study to address these deficiencies. This paper describes a method for conducting such a study, and represents (to the authors' knowledge) the first broad-based attempt to do so. The results not only promise insights into the nature of current LCC estimates, but also suggest the possibility of improving the accuracy of DoD LCC estimates via a stochasticallybased model.

As the United States Department of Defense continues to increase the number of Remotely Piloted Aircraft (RPA) operations overseas, improved Human Systems Integration becomes increasingly important. Manpower limitations have motivated the investigation of Multiple Aircraft Control (MAC)

The future of space-based defense weather systems is uncertain due to the cancellation of the National Polar-orbiting Operational Environmental Satellite System (NPOESS) and Defense Weather Satellite System (DWSS) program. The U.S. Department of Defense (DOD) Weather Satellite Follow-on (WSF) program has been identified as the follow-on program and a potential pathfinder for the concept of disaggregation. Disaggregation of satellites may reduce costs, improve performance, and increase system resiliency. However, current research literature does not address a systematic method for performing conceptual trades between large multi-function and small (functionally-cohesive) spacecraft. An multifunction/multi-orbit disaggregated space system optimization methodology is proposed and applied to the WSF conceptual architecture problem. Real-coded mixed integer optimization techniques are implemented to identify, assess, and compare alternative space-based weather system conceptual architectures. Resulting cost and performance figures of merit are analyzed to assess disaggregated space-based weather constellation cost effectiveness.

The DOD routinely demonstrates its capability to develop phenomenal systems; however, these accomplishments are often tarnished by substantial cost and schedule overruns. While defense policies are continually being revised to address these problems, many believe that a more fundamental source of these overruns is the lack of flexibility in the systems being developed. But providing justification to invest in flexibility is a tough sell, as stakeholders struggle to quantitatively demonstrate the potential return on investment. Therefore, this paper introduces an alternative methodology for valuing the inherent ability of a system or design to accommodate change. The proposed methodology is essentially a modification of the current life cycle model and is premised on the notion that the need for capability changes in a system arise in a stochastic manner that can be incorporated into a continually updated, expected value model presented in terms of total life cycle cost.

Despite its ubiquity in the systems engineering literature, flexibility remains an ambiguous concept. There exist a multitude of definitions, which vary not only by domain, but within domains as well. Furthermore, these definitions often conflict with one another, making it difficult to discern the intended meaning in a given study or to form generalizations across studies. Complicating matters, there is a plethora of related terminology that is often used carelessly and/or interchangeably with flexibility. In this paper, we employ a novel ontological framework for clarifying salient aspects of extant flexibility-related terminology. While it was not possible to distill consensus definitions from the literature, we did identify certain dominant characteristics that enabled us to formulate a set of democratic definitions for flexibility, adaptability, and robustness, as well as recommended definitions for agility and versatility. We believe that the proposed definitions of these key system design principles may provide a baseline for improving analysis and communication among systems engineering practitioners and academics.

This research, conducted in the Systems Engineering Research Center (SERC), examined systems engineering and engineering management practices for military rapid capability and urgent needs programs. Lifecycle of urgent needs programs is driven by "time to market" as opposed to complete satisfaction of static requirements, with delivery expected in months versus years/decades. The processes and practices applied to urgent needs must add value and not require an excessive bureaucratic oversight to implement, while at the same time address, understand, and manage risk such that programs can understand better where to include, truncate, eliminate, tailor, or scale systems engineering practices and processes. Focusing on aspects of the product, process, and people of military rapid organizations, the analysis showed that these organizations have the right team, develop innovative conceptual solutions, quickly prune the design space, and identify appropriate designs that can deliver warfighting capability expeditiously. While these observations may not seem new, they provide the foundation for a broader framework of rapid development, which is the subject of ongoing research.