AN INDIRECT WORKFORCE (RE)ALLOCATION MODEL FOR SEMICONDUCTOR MANUFACTURING Chen-Fu Chien Wen-Chih Chen Shao-Chung Hsu Department of Industrial Engineering and Engineering Management Department of Industrial Engineering and Management Department of Industrial Engineering and Engineering Management National Tsing Hua University National Chiao Tung University National Tsing Hua University Hsinchu 30013, TAIWAN, ROC Hsinchu 30013, TAIWAN, ROC Hsinchu 30013, TAIWAN, ROC ABSTRACT Semiconductor industry is a capital intensive and knowledge intensive industry, in which human resource management and human capital enhancement is increasingly important. To maintain competitive human resource, it is critical to develop a decision framework for headcount planning and workforce allocation for indirect labors. Motivated by the needs in real setting, this study aims to develop a model for allocating indirect workforce among semiconductor fabrication facilities to meet expected outputs and labor productivity improvement. Workforce allocation and reallocation based on the overall corporate workforce level is essential so that the shortage or exceed workforce will be balanced among different production sites. The key to achieve this purpose is the proper understanding of real requirements of each production site according to its corresponding tasks assigned. Non-parametric activity analysis approach is used for the workforce requirement estimation given delegated tasks. The estimation is based on the best performance from the past with adjustments reflecting the expected productivity growth. 1 INTRODUCTION Driven by Moore’s law, semiconductor industry is knowledge and capital intensive. Thus, human capital enhancement and human resource management is getting important nowadays (Chien and Chen 2007; Leachman et al. 2007). In particular, workforce planning and headcount allocation have become important issues for both research and practice in semiconductor manufacturing. Research has been done on workforce planning decisions including staff scheduling or rostering, which determines work timetables for staff so that the demand can be satisfied while optimizing certain criteria. For example, Thompson and Goodale (2006) present a staff scheduling method for the cases where workforces have different productivity levels. Staff scheduling problem is an extension of conventional scheduling problems and comprehensive reviews can be found in (Aykin 2000; Burke et al. 2004; Ernst et al. 2004). These decisions are typically operational and all detailed information is assumed to be obtainable. Another research is related to job assignment or reallocation that determines the staff-job assignment. Assigning workforces to jobs can be modified and modeled as the classic assignment problems (e.g., Holder 2005) or by other means such as simulation (e.g., Zulch et al. 2004). In addition, long-term staffing deals with determine the optimal workforce requirements of each category hired at each period in light of production ramping and technology migration. A stream of this research is workforce planning optimization under deterministic conditions (e.g. Mundschenk and Drexl 2007). Another stream considers stochastic nature of the problem such as learning curve and turnover, which can be modeled as Markov decision processes and then solved using different techniques (e.g., Gans and Zhou 2002; Ahnet et al. 2005). The applications of workforce planning have been studied in various industries. For example, Mundschenk and Drexl (2007) propose an integer programming model for long-run staffing particularly for printing industry, and Pesch and Tetzlaff (2005) study the interactions between staffing and scheduling decisions in the automotive industry. Bard et al. (2007) investigate the workforces planning for United States Postal Service mail processing and distribution centers. Holder (2005) studies the optimizing process of assigning sailors to jobs for US Navy in an attempt to increase sailor satisfaction. However, most of the existing studies on workforce decisions were based on simplified assumptions and thus can hardly solve real problems especially in knowledge intensive high-tech industries. This research was motivated by a semiconductor company in real setting in Taiwan. Firstly, with the scale and involved automation of semiconductor fab increases, the amounts and percentage of knowledge workers also raise hugely. Firms are facing the situation that the cost of automation and manpower is ascending year by year, and the engineers and technical operators play more and more important roles in the factories. Secondly, semiconductor companies in Taiwan used to attract and retain talent by issuing stock dividends rather than giving high salaries. Beginning in 2008, the new accounting rules request expensing employee bonuses, which will affect the financial reports of most high-tech companies. Therefore, maintaining proper level of workforce and enhance people productivity become critical. Thirdly, equipments are long-term investments that are lack of flexibility to any adjustments in response to demand variation once the decisions are made. On the 2201 978-1-4244-2708-6/08/$25.00 ©2008 IEEE Proceedings of the 2008 Winter Simulation Conference S. J. Mason, R. R. Hill, L. Mönch, O. Rose, T. Jefferson, J. W. Fowler eds.