This study deals with a measurement of manufacturing flexibility in manufacturing systems. This study begins with the attempt to find out the problems. From the review of vast literature, the reasons why manufacturing flexibility is difficult to be operationalized are followings: 1) ambiguous definitions of flexibility, 2) interdependency among various types of flexibility, and 3) mixed use of available and realized flexibility.
In this study, to solve the previous problems, we have introduced a new concept of flexibility, Comprehensive Flexibility, not with respect to how to be adapted to them, but with respect to the degree to be adapted to environmental changes as a whole and with respect to the concept of realized flexibility. It is defined as the ability to keep up the efficiency of manufacturing under various environmental changes.
Then, it is measured a unified performance measure, Performance Measure of Comprehensive Flexibility (PMCF), to cover over various types of flexibility in a manufacturing system and to prevent the duplicate measurement of various types of manufacturing flexibility in the process of absorbing environmental changes. PMCF is measured from efficiency in a manufacturing system under the consideration of the status of its machines and parts. Partial performance measures, Performance Measure of Comprehensive Flexibility in machines ($PMCF_M$) and Performance Measure of Comprehensive Flexibility in parts ($PMCF_P$) are calculated by efficiency of the machines and parts in the manufacturing system. Then, PMCF is developed by weighted average between $PMCF_M$ and $PMCF_P$. The weighting values representing the degrees of importance of $PMCF_M$ and $PMCF_P$ can be calculated through the concept of their opportunity costs. As a supplementary measure, Redundancy Measure (RM) represents the degree of remaining capacity of a manufacturing system after manufacturing a given product planning with certain environmental changes.
A simulation study is also presented to illustrate the process of quantifying the proposed measures, to interpret the meaning of the measures, and to examine the relationships between them. The simulation model is written in borland C++, a widely useful programming language, and implemented on a personal computer with an 80486 processor. Three types of manufacturing systems are evaluated using this simulation model with encouraging results concerning the trends of their flexibility according to the environmental changes.
