| This research was one of few projects by which traditional work scheduling rules were investigated for Flexible Manufacturing Systems using a physical simulation. A physical FMS model using Fischer Technik components was utilized to evaluate traditional work scheduling rules. The model was a closed-loop system controlled by two microcomputers: a Commodore CBM 8032 and a Radio Shack TRS-80 Model III. It consisted of four major components including eight storage areas, an automatic material handling system, six machine centers and a turning cell. The model processed seven different part families in which five identical parts were included.;Seven traditional work scheduling rules including RANDOM, FSFS, SPT, DDATE, SLACK, S/PT and VALUE were used for selection of a part from one of storage areas. Four traditional scheduling rules including RANDOM, FMFS, NINQ and WINQ were used to select a machine center among six identical machine centers. The total number of decision rule sets evaluated in this research was twenty-eight because one part selection rule was combined with a machine center selection rule to make a decision rule set.;Each decision rule set was simulated with unplanned failures of the major system components for 140 hours in real time. The input data for the physical simulation were based on the actual data and adjusted by a scaling factor for the FMS model.;Six different performance criteria including system effectivity, traveling time, production output, manufacturing throughput time, work-in-process inventory and production lateness were utilized to evaluate the best decision rule sets in the FMS model.;The SPT/WINQ rule set was the best performer under failure effects due to its high fetch rate and high processing rate for parts with short processing times. This decision rule set illustrated the SPT rule's main characteristic in which parts with the short processing time were fetched most frequently. Other decision rule sets were also discussed and evaluated. |
