Stephen W. Eisenhawer

Progress Toward Future Runway Management

The runway is universally acknowledged as a constraining factor to capacity in the National Airspace System (NAS). It follows that investigation of the effective use of runways, both in terms of selection and assignment, is paramount to the efficiency of future NAS operations. The need to address runway management is not a new idea; however, as the complexities of factors affecting runway selection and usage increase, the need for effective research in this area correspondingly increases. Under the National Aeronautics and Space Administrations Airspace Systems Program, runway management is a key research area. To address a future NAS which promises to be a complex landscape of factors and competing interests among users and operators, effective runway management strategies and capabilities are required. This effort has evolved from an assessment of current practices, an understanding of research activities addressing surface and airspace operations, traffic flow management enhancements, among others. This work has yielded significant progress. Systems analysis work indicates that the value of System Oriented Runway Management tools is significantly increased in the metroplex environment over that of the single airport case. Algorithms have been developed to provide runway configuration recommendations for a single airport with multiple runways. A benefits analysis has been conducted that indicates the SORM benefits include supporting traffic growth, cost reduction as a result of system efficiency, NAS optimization from metroplex operations, fairness in aircraft operations, and rational decision making.

A New Graphical Tool for Building Logic-Gate Trees

Tree structures that use logic gates to model system behavior have proven very useful in safety and reliability studies. In particular process trees are the basic structure used in a decision analysis methodology developed at Los Alamos called Logic Evolved Decision modeling (LED). LED TOOLS is the initial attempt to provide LED-based decision analysis tools in a state of the art software package. The initial release of the software, Version 2.0, addresses the first step in LED — determination of the possibilities. LED TOOLS is an object-oriented application written in Visual Basic for Windows NT based operating systems. It provides an innovative graphical user interface that was designed to emphasize the visual characteristics of logic trees and to make their development efficient and accessible to the subject matter experts who possess the detailed knowledge incorporated in the process trees. This eliminates the need for the current interface between subject matter experts and logic modeling experts. This paper provides an introduction to LED TOOLS. We begin with a description of the programming environment. The construction of a process tree is described and the simplicity and efficiency of the approach incorporated in the software is discussed. We consider the nature of the logical equations that the tree represents and show how solution of the equations yield natural language “paths.” Finally we discuss the planned improvements to the software.Copyright

An approximate reasoning-based method for screening high-level-waste tanks for flammable gas

The in situ retention of flammable gas produced by radiolysis and thermal decomposition in high-level waste can pose a safety problem if the gases are released episodically into the dome space of a storage tank. Screening efforts at the Hanford site have been directed at identifying tanks in which this situation could exist. Problems encountered in screening motivated an effort to develop an improved screening methodology. Approximate reasoning (AR) is a formalism designed to emulate the kinds of complex judgments made by subject matter experts. It uses inductive logic structures to build a sequence of forward-chaining inferences about a subject. Approximate-reasoning models incorporate natural language expressions known as linguistic variables to represent evidence. The use of fuzzy sets to represent these variables mathematically makes it practical to evaluate quantitative and qualitative information consistently. In a pilot study to investigate the utility of AR for flammable gas screening, the effort to implement such a model was found to be acceptable, and computational requirements were found to be reasonable. The preliminary results showed that important judgments about the validity of observational data and the predictive power of models could be made. These results give new insights into the problems observed in previous screening efforts.

Risk assessment when malevolent actions are involved

This paper discusses extending systems analysis techniques developed for analyzing the risk of random or unintentional events to the analysis of the risk arising from malevolent actions. Methods for developing sets of attack scenarios useful for risk analysis using logic-gate trees are described. The importance of using a largely qualitative understanding of adversary decision processes to assess the relative attractiveness of different attack scenarios to an adversary is addressed and potential methods for rigorously developing this relative ordering are discussed. The importance of including an evaluation of the effectiveness of protective features implemented to reduce risk is discussed and an analytical approach described. The importance of dependency among elements of an attack scenario is discussed and a method for the efficient handling of such dependencies is described

A Logic Model Approach to Conceptual Design of a Scientific/Industrial Complex

We demonstrate how the use of logic models and approximate reasoning can lead to more effective conceptual design of a large-scale scientific/industrial complex. A deductive model is used to describe the relationships between the mission of a complex, the tasks that are the concrete expressions of the mission, and the capabilities and resources that allow the mission to be carried out. This deductive logic model provides the formal basis for integrating these relationships into a conceptual design and provides a framework to express design options that are an essential part of the description of the complex. A decision model for choosing among the design options is constructed using approximate reasoning. This AR model uses forward-chaining inference models to emulate the types of reasoning used by experts. This approach allows the analyst to use both numerical data when available and the qualitative knowledge that forms much of the information base used in conceptual designs of new and innovative complexes.Copyright

A Structured Approach to Resource Allocation

Allocating limited resources among competing candidates is an important problem in management. In this paper, we describe a structured and flexible approach to resource allocation using logic-evolved decision (LED) analysis. LED analysis uses logic models to generate an exhaustive set of competing alternatives and the inferential model that is used for preference ordering of these alternatives. The inferential models can use data in numerical, linguistic, or mixed forms; uncertainty in the evaluation results can be expressed using probabilistic- or linguistic-based methods. We illustrate the use of LED analysis for an allocation problem with numerical input data and for an allocation problem with only linguistic input data.Copyright

Evaluation of high-level waste pretreatment processes with an approximate reasoning model

The development of an approximate-reasoning (AR)-based model to analyze pretreatment options for high-level waste is presented. AR methods are used to emulate the processes used by experts in arriving at a judgment. In this paper, the authors first consider two specific issues in applying AR to the analysis of pretreatment options. They examine how to combine quantitative and qualitative evidence to infer the acceptability of a process result using the example of cesium content in low-level waste. They then demonstrate the use of simple physical models to structure expert elicitation and to produce inferences consistent with a problem involving waste particle size effects.