The bulk of the C code develops the logic associated with both OOM and WINQ assignment rules.
A MT will be assigned to the part based upon the current assignment rule, OOM or WINQ. A SLAM global variable emulates a toggle switch so that the selection of OOM as the current assignment rule prevents the computer program from executing the WINQ subroutine and vice versa.
Time in system is less when parts were assigned to machine using the WINQ rule rather than OOM.
Results show that the WINQ rule performs significantly better than OOM regardless of how machine commonality and number of AGV are set.
Low machine commonality High Machine Commonality OOM WINQ OOM WINQ 3 AGV 42.35 36.20 42.55 29.47 5 AGV 39.26 24.26 40.74 23.88 7 AGV 40.60 25.52 41.77 25.25 The two best performers combine WINQ and 5 AGVs at either high or low level of machine commonality.
Results show there are significant differences in mean time in system for a given sequencing rule depending on whether OOM or WINQ is employed to assign machines.
OOM SPT/TOT SPT MOPP 40.75 40.81 42.07 WINQ 27.35 27.40 27.53 Analysis of time in system grouped by S and V shows the lowest times in system are associated with 5 AGVs regardless of the scheduling rule, as shown in Table VII.
AGV SPT/TOT SPT MOPP 3 37.15 37.28 38.51 5 31.87 31.95 32.29 7 33.14 33.10 33.61 The major surprise of the results is the superior performance of WINQ when compared to OOM.
The analysis of the average machine utilization controlled for assignment rule, number of AGVs, and machine commonality shows that for every treatment combination, the WINQ maximum machine utilization were lower than those generated by OOM.