Search Results

This paper presents an application of Bayesian Belief Networks for modeling the uncertainty in aircraft safety diagnostics. Belief networks or influence diagrams represent possible means to efficiently model uncertain causal relationships among components of a system. HUGIN is a software for the construction of knowledge based systems based on Bayesian networks. A HUGIN prototype is dicussed to illustrate how a Bayesian approach could be used to support the decision search routine of aircraft safety inspectors when diagnosing equipment of subsystem malfunctions. The example focuses on diagnostic procedures for assessing aircraft tire condition.

The On-line Project Information System (OPIS) is a LAMP-based (Linux, Apache, MySQL, PHP) system being developed at NASA Ames Research Center to improve Agency information transfer and data availability, largely for improvement of system analysis and engineering. The tool will enable users to investigate NASA technology development efforts, connect with experts, and access technology development data. OPIS is currently being developed for NASA's Exploration Life Support (ELS) Project. Within OPIS, NASA ELS Managers assign projects to Principal Investigators (PI), track responsible individuals and institutions, and designate reporting assignments. Each PI populates a “Project Page” with a project overview, team member information, files, citations, and images. PI's may also delegate on-line report viewing and editing privileges to specific team members. Users can browse or search for project and member information.

This paper presents a probabilistic approach for using the model-based reasoning of Bayesian Belief Networks (BBNs) to perform risk assessments of new aviation safety products. Sponsored by NASA's Aviation Safety and Security Program [1], the author is leading a research team at Rutgers University in the creation of aircraft accident models in order to assess the projected relative risk reductions of an aeronautics technology portfolio. The modeling approach uses elements from a case study architecture, inductive reasoning and analytic generalization. Aspects of the modeling approach, including knowledge capture and sensitivity analyses are emphasized and preliminary results discussed.

In the field of automotive electronics, a quick and accurate technique to predict the useful life of electronic modules is considered a critical tool to assure future, and existing, product designs are successful in meeting long-term life requirements. This need is particularly true for harsh environment electronics, such as transmission and engine control modules. These modules are confronted with many difficulties such as a changing thermal environment and reductions in component packaging size that are not significant problems for most other electronic products. Because these modules are designed for ten or more years of use, accelerated life testing methods are necessary for predicting product life for new designs. However, there is a very limited understanding of the relationship between accelerated life testing and product field requirements. A better understanding of this relationship is critical to future product designs.

The wide applications of automatic sensing devices and data acquisition systems in automotive manufacturing have resulted in a data-rich environment, which demands new data mining methodologies for effective data fusion and information integration to support decision making. This paper presents a new methodology for developing a diagnostic system using manufacturing system data for high-value assets in automotive manufacturing. The proposed method extends the basic attributes control charts with the following key elements: optimal feature subset selection considering multiple features and correlation structure, balancing the type I and type II errors in decision making, on-line process monitoring using adaptive modeling with control charts, and diagnostic performance assessment using shift and trend detection. The performance of the developed diagnostic system can be continuously improved as the knowledge of machine faults is automatically accumulated during production.