Computer integrated flexible manufacturing systems (FMSs) open up a new horizon for manufacturing industries. An FMS combines the elements of machining centers and robots with computerized parts transfer to form a flexible automated factory. This kind of advanced manufacturing technology is capable of producing a large variety of products with a large variety of volumes. As such, the capital investment in establishing an FMS is much heavier than any other type of manufacturing system. It is therefore of paramount importance to conduct pre-installation system design for an FMS.
In general, poor design which results in the development and installation of an FMS that does not perform as intended or that does not operate at a level that has made use of its full capacity, originates from a lack of proper design methodology that enables the estimate and evaluation of the dynamic impact of the FMS activities on its overall system performance.
This research aims at establishing an effective design methodology for FMS. It is also intended to explore and to identify factors that affect the performance of FMS, and to generate a set of guidelines governing the final stage of design of FMS.
After a thorough literature survey and study of the FMS activities and related technology, discrete-event computer simulation has been identified as an effective tool for the design of FMS. Exhaustive searching and evaluation of about three dozen manufacturing simulation packages were performed before the MAST simulation package was selected as a design tool for this research. A new computer model named Floor.Mod was developed to test the validity of MAST before conducting the simulation runs.
In addition to MAST, many other computer software packages such as SPSS, AutoCad, Microsoft word, Symphony, and Chart Master, were used as supplementary tools for this research.
In order to conduct the simulation runs, field work has been done in Japan to observe and collect industrial data pertaining to the design of FMS. Three sets of data related to the metal manufacturing industries using FMS technology have been collected as input for the computer simulation runs. Though the three Japanese FMSs have been in existence they have never been simulated to study their potential. Various simulation runs were designed and conducted in the ways that the major system parameters that may upgrade the output performance of an FMS were identified and tested.
In the final stage of the research, two large scale FMSs namely the Ohjasi and the Combin Systems were developed with many of their system alternatives being evaluated to verify the results derived from the Japanese FMSs and to explore the feasibility of improved system flexibility.
Eleven thousand and eighty (11,080) simulation runs (Appendix 5) have been conducted with simulation experience and research findings that have not been found in any of the previous academic work, and they can be applied directly to industries that intend to establish FMSs.
In this research, discrete-event computer simulation has been proved as an effective design tool to gain insight into the complexity of an FMS by the comparison of results obtained by the SPAR analysis based on queueing model and by MAST computer simulation and animation. Other major findings of this research include improving system performance through pallet design and the feasibility of improving system flexibility through the development of composite FMS.
In an FMS, a pallet is configured to be able to carry a number of parts at one time. The number of pallets used and the number of parts per pallet are two system parameters that have very significant effect on system performance. An increase in the number of pallets/part-type has very little effect on system output rate but contributes to system congestion and eventually causes blockage of the system while a proper design of pallets capable of carrying more parts will upgrade the output rates to a very large extent without increasing the number of workstations that requires rigid capital involvement.
Flexibility of FMS in terms of the variety of product-types and the variety of product-volumes can be improved through the development of composite FMSs with reduced number of workstations as compared with each individual subsystems. It has also been concluded that there is no significant effect of increasing the number and the speed of AGVs, and a well-designed FMS configuration is not necessarily the one with the largest number of workstations.
The need for simulation studies of the compatibility of the system configuration and the control algorithm has also been highlighted, and a set of general guidelines for the design of FMS has been generated.