Decision Support for Operator Staffing and Assignment in Semi-Conductor Manufacturing

Yap Shi Li

Abstract

Staffing Models (SM) are very important for semiconductor manufacturing. As the machines are very expensive, a good SM able to improve work efficiency and manufacturing productivity thus saving cost and increasing profit. In this paper, we propose a set of algorithm in managing manpower needs in the fab using AHP, linear and multi-integer programming. Productive and non-productive tasks are taken into consideration, and profile of each operator will be evaluated with the proposed techniques.

For many years the semiconductor industry has a traditional method of managing its headcount by grouping the operators into four typical shift groups that take turns to cover day and night shifts every fortnight. In this paper, an operator scheduling method is used to hope to eliminate the restriction of the operators into those shift groups to allow more flexibility in the planning as well as increasing the work efficiency and more time management for the operators.

The model is formulated in such a way that decision makers can easily understand and set parameters for the objective function. Due to the multi-criteria nature of the problem, the model is built flexible enough to handle most of the cases. For the purposes of illustration and comparison, a numerical example is given with real data. We show that by tuning the weights for different targets, the model is able to reflect the decision makers' preferences. Lastly, an Excel template is designed for the model Expert Choice and ILOG OPL, which are readily available online, are used to solve the problem. This feature further enhances the practical appeal of our formulation.