Controlling a manufacturing process using statistical process control: An empirical investigation of problems.

摘要

Statistical process control (SPC) is commonly used in manufacturing operations as an aid to determine if unusual sources of variability are present in a process. Despite the development and application of many types of SPC, usage over the past 70 years has not been unproblematic. This may be due to the relatively narrow focus which has been traditionally placed on the statistical and mathematical aspects and assumptions required by specific SPC charts and not on the SPC system as a whole.; This dissertation finds industrial evidence to support that the SPC task as practiced in industry, is divergent from the SPC task as designed due to: (a) a misalignment of the goals of SPC and other organizational goals and (b) the design of SPC as a decision making aid. Relevant literature was used to both substantiate these findings and to suggest systemic changes in the application and usage of SPC. Laboratory experimentation aimed at evaluating the efficacy of several suggested changes was then conducted to determine their effect on operator performance.; This examination of SPC has resulted in a greater understanding of the human factor element in the SPC system and has spurred the development of a novel display of process data called Process Window Monitoring. The Process Window Monitoring methodology, (developed and discussed herein), is yet untested but is included as part of this research as a proposed alternative to traditional SPC.