Scheduling analysis using discrete event simulation
SS '96 Proceedings of the 29th Annual Simulation Symposium (SS '96)
| Hi-index | 0.00 |
This paper presents an application of combined discrete and network simulation modelling to determine and validate an appropriate sequencing technique for a modified flow shop. The research was based upon a study of a modified flow shop at the International Business Machines, Federal Systems Division (IBM/FSD), Manassas, Virginia, manufacturing Facility. The company's concerns were directed towards enhancing the real-time scheduling of a man and machine dependent flow shop where meeting customer due dates was vital. In this manufacturing facility it is necessary to rework all parts that do not initially meet stringent quality control specifications until those parts do meet those quality limits. Therefore another reason for analyzing different sequencing rules was the necessity to better control the rework activity. The sequencing technique currently used at this facility is based on Earliest Due Date scheduling. With the cooperation of the production control organization at IBM/FSD, Manassas, a simulation study was performed in order to determine if an enhancement to the current system could be found. The objective of the research was to determine which due date based sequencing technique would best meet the overriding production control criteria of the IBM/FSD flow shop. The company's production control objectives were to minimize number of tardy jobs, total amount of job tardiness, and total amount of in--process inventory. The flow shop was modelled using SLAM simulation language. The flow shop was both machine and man dependent requiring both entities to be modelled. A time-consuming task encountered in the development of this model, and with the development of many other simulation models which attempt to represent real world systems, was the task of obtaining data in the proper format to analytically determine the control parameters for the model. This paper presents and discusses some of the difficulties encountered with the data interpretation. Also included in the paper is a discussion of the due date based sequencing techniques studied and the usefulness of simulation to determine and validate the appropriate sequencing technique.