Date of Award

Spring 1993

Document Type

Thesis - Restricted

Degree Name

Master of Science (MS)

Department

Mechanical Engineering

Abstract

As the complexity of the manufacturing processes grow, it becomes increasingly difficult to develop mathematical models of their behavior. As a result, control of these highly complex processes is accomplished, in part, through the use of heuristics or 'rules of thumb' developed with those charged with maintaining the processes. This approach has disadvantages in that: 1) the rules are usually not documented; 2) no effort has been made to insure that the rules are consistent among all those manipulating the processes; and 3) the rules are not available to everyone involved, resulting in specialities in certain problems. One method of addressing these problems is the application of an expert system, which incorporates the knowledge of all the workers controlling the process. The expert system acts as a consultant, providing advice on analyzing and reacting to problems through operator interaction with the terminal. This expert system project was initiated after observation of actual problems in industry. A commendable portion of quality control manager's time in a production system is spent in routine and complex decision making process. These decisions may have significant impact on average outgoing quality, quality improvement, and quality cost. The quality control manager must decide among various statistical process control methods and sampling plans for each part (or characteristic). These decisions are usually based on how critical a part is, historical information about the quality of the parts, and other factors. Many of these factors require subjective judgments by the quality control manager. For a production facility with an inventory system of thousands of different parts, determination of feasible sampling plans and process control charts is a time consuming and difficult task. The problem is to select appropriate sampling plans. Because of the complex nature of the selection process and the large number of variables which are involved, human (expert) is a must. But only recently, managers have realized the dependence of organizations upon their experts (here quality control managers), mainly where the experts informal power might have disturbed the power hierarchy. And it is only recently that artificial intelligence software technologies have matured to a state where expert knowledge can be captured and made available wherever and whenever needed. Expert systems, in this view are the carriers of second revolution in automation of knowledge ("white collar robot"). This thesis demonstrates the design and development of an expert system to tabulate the selection of an appropriate sampling plan for each part and focuses on the practical problems associated with constructing such a system for use on the factory floor. It is based on the experience of building an expert system to aid in the selection of sampling plans in a typical manufacturing environment. The thesis emphasizes issues associated with developing a knowledge acquisition strategy which overcomes the problems unique to factory floor application and facilitates user acceptance of the final product. Also included is a brief comparison of some software available for microcomputers based expert system.

Download

Share

COinS

Restricted Access Item

Having trouble?