Wortham, A. William (1988, ASQC) A. Willaim Wortham & Associates, Bryan, TX
Artificial Intelligence (AI) is a relatively new science with heavy emphasis on its role in the use of our modern computers. In fact, it should be associated with the early development of symbolic logic in mathematics and the decision sciences in statistics. It has been defined as "the subfield of computer science concerned with developing intelligent computer programs." This is most likely a definition presented by a computer scientist. A broader definition will include the transformation of human and data base intelligence into a system of mathematical logic for the development of knowledge and decision making. It is in this broader context that it becomes important in the field of Statistical Process Control.
Statistical Process Control (SPC) had its beginning with the Shewhart quality control charts. It is not difficult to conceive of them as an early form of artificial intelligence. Based on an assessment of data base information, they do, in fact, give us the knowledge and decision making rule for saying if a process is "in control." Thus, given a computer program to help with the computations and logical processes, they are an artificial intelligence effort whatever the definition. Clearly, with today's computers and microcomputer and with the number of SPC programs already on the market, SPC is perhaps the most important and most frequent use of Artificial Intelligence.
The purpose of this paper is to review the criterion and a system of logic which permits the more generalized use of AI in SPC. The system presented is a weighted logic system, assigning weights to and combining the results of several statistical tests.
Electronic data systems