Title

An Integrated Approach To Analyzing Systems And Their Subsystems

Keywords

Decision support tools; Modeling; Systems integration

Abstract

System integrity and efficiency in which multiple levels of integration are used require the utmost attention towards intelligent modeling and design. This applies more specifically to systems (or networks) where high levels of expectation in performance and reliability exist. Reliability in this context means meeting or preferably exceeding all required specifications repeatedly over time. In order to achieve such results, the system, from the design of the software elements to the interaction of hardware factors, must be engineered such that the whole system success is regarded as more essential than individual component efficiencies. Once all levels of design feature this system-wide approach in their modeling, a maximized point of both integrity and efficiency can be achieved. As the number of levels of system integration increases, the level of system complexity tends to increase exponentially. System complexity is primarily a function of the interdependencies between the individual components and the variability of the behavior of those components. These two factors make system behavior difficult to analyze and predict exactly. However, in order to maximize the level of system performance and reliability, the solution methodology, often a decision support tool, must be intelligent enough to consider system dynamics within these highly-integrative environments. In general, decision support tools (DSTs) are interactive computer-based systems intended to help decision-makers compile useful information from raw data, documents, personal knowledge, and/or mathematical models to support complex decision-making and problem-solving. Classic DST design consists of three primary subcomponents: (1) a database management system with access to internal and external data, information and knowledge, including the consequences of different decision alternatives, given past experience, (2) a model management system that accesses often sophisticated modeling procedures and (3) a user-interface that enables interactive queries and reporting functions. The database management system (DBMS) supports access to all relevant information assets, including legacy and relational data sources, data warehouses, and data marts. The DBMS must possess the ability to support retrieval and storage of information. The model management system (MMS) not only retrieves the relevant data but it also has the ability to cope with inaccurate, incomplete, fuzzy and conflicting data. The MMS then utilizes a modeling methodology or suggests a set of modeling methodologies that presents a solution or set of solutions to the decision-maker. The more common DST modeling methodologies use traditional optimization (e.g., linear programming, dynamic programming, etc.), statistical methods, meta-heuristics, and artificial intelligence. The user-interface design then presents the solution or set of solutions in a concise yet descriptive and informative manner allowing the decision-maker minimal time to interpret the results. Timely communication among these sub-components is critical to the efficiency with which a user makes a decision and the effectiveness of that decision. Corrective measures are one of the many important ideas behind a system-wide approach to monitoring levels of reliability and integrity. If fact, frequent looping back is necessary as more is learned about the problem, for example which solutions succeed and which solutions fail. In addition, the feedback loop enables the DST to adapt to changing problem and system characteristics. There are several industries where the system-wide approach to effective modeling can be applied, including Ihe airline industry, automotive industry, and aerospace industry. In this research, a case study of a successful application of a DST in one such industry is presented - the aerospace industry. In this case study, the specific integrative system environment is described, including the levels at which information is collected and the information flow between the system components at each level. Finally, the architecture of the DST is presented. The authors are not so much advocating a new paradigm for the development of DSTs, but rather proposing a unique approach to apply DSTs to a complex problem in a highly time - and safety-critical industry.

Publication Date

12-1-2004

Publication Title

IIE Annual Conference and Exhibition 2004

Number of Pages

41-

Document Type

Article; Proceedings Paper

Personal Identifier

scopus

Socpus ID

30044450543 (Scopus)

Source API URL

https://api.elsevier.com/content/abstract/scopus_id/30044450543

This document is currently not available here.

Share

COinS