Labor Allocation Optimization Modeling And Study For Manufacturing System

In ATM manufactory, the HC is manually allocated basing on MMR by monthly. This process cause a lot of manual work loading, furthermore, the HC forecast process will introduce low accuracy because of the demand variation and worker absenteeism! Here, we will introduce a new method using work loading allocation with the Goal Programming Model by shiftily, which improves the accuracy by eliminate the impact of the demand variation and worker absenteeism. The difficulties to implement this tool to daily operation in the semi-conductor industry are how to make this tool more close to line real situation with considering about the interference of machine, available HC, machine availability, ect. This paper will introduce an integrated automatic tool, which allocate the shiftily HC base on the work loading. This solution can optimize the one area HC, even one factory HC, as a cell to full utilize the cross-training and SJF resource. This tool works out the solution by goal programming with a strong line programming solver provided by the iLOG CPLEX? software. As a computer integrated tool, it can get the result by shiftily with low manual work loading, furthermore, the shiftily simulation can also eliminate the impact on HC allocation from the WIP fluctuation and machine readiness, which make the HC arrangement more reasonable by micro-management. As for man-machine interference problem, the solution given by tool has been verified by a self-adjustment model, which is developed by the factory physics queuing theory. The model will verify the machine capacity lose base on the HC allocation and check if this capacity lose is affordable by current situation. When machine capacity is not enough to support this shift loading, and system will increase the HC and do a simulation again. This loop will work out the probable solution again and again until we get the doable result. A friendly GUI is developed to end user. GUI help the end user easily input the current shift information such as the available HC, current shift goal, machine conversion plan, ect. This help solver get the real-time information and workout the more reliable solution. As a result, the tool will output a matrix to advise supervisors which labor allocates to which specific work station directly. It has been demonstrated that the tool can work out the HC allocation solution within 2 min, and this solution can help shift to allocate the labor more effectively and improve the efficiency of the SJF and cross-module supporting. The work loading allocation method is not only suitable for semi-conductor industry, but also workable in other manufactory model to help the factory management HC more efficiently.