| 摘要: | 半導體廠因製程上的需求，機械設備一直以來皆具有潛在之危害，尤其是有許多火災案列的發生，更顯得自動滅火裝置的重要性。因此，藉由自動滅火裝置的定期保養測試，以達到在真實狀況發生時，功能可正常運作，迅速滅火。在目前現實的工作環境之下，機台測試保養的工作大多由外包廠商負責，派遣的狀況常常依據經驗來安排，並依機台數量來報價，未曾考慮到工時、測試機台之難易度皆與費用有關。因此決策者要在廠內現有的機台狀況下進行適當調度派遣並且降低成本時為重要之課題。本研究係探討機台測試人員指派，並在各種特殊限制條件下，求付給外包廠商最少費用為目標，進行指派與調度，以提供決策者派遣時之參考，除此之外，更依據各種參數進行敏感度變化之趨勢分析，能夠提早發現各種無法預知的結果。
為證實本模式合理性與實用性，本研究以竹科某半導體廠為範例，使用廠內的運作條件及基本資料訂定參數，利用整數規劃方法建構一套數學最佳化模式，以Microsoft Excel 2010軟體資料輸入LINGO套裝軟體求解，即可得機台自動滅火裝置測試人員派遣總成本最佳化結果。除此之外，可以讓最佳化模式之派遣結果與現實廠內實務經驗指派結果進行分析比對，結果發現所需派遣總成本比人工經驗指派來的低且效率更好。故顯示本研究模式確實可靈活運用於半導體廠機台自動滅火系統測試人員派遣問題，亦可當作重要參考之依據。
;Equipment and machinery required in semiconductor manufacturing have always exposed semiconductor plants to potential hazards and risk. With a long history of fire accidents, the importance of automatic fire extinguishing equipment in semiconductor plants cannot be overstated. Therefore, regular maintenance and testing of automatic fire extinguishers can ensure normal function and, in the event of a real fire, swift elimination of the fire hazard. Under current working environments and practices, maintenance and testing of semiconductor equipment are contracted to outside vendors. The scale and state of this outsourcing is usually determined by prior experience and also by price quotations for the number of equipment to be serviced. Little attention is paid to the fact that working hours and difficulty in testing of the equipment are all relevant to overall cost. As a result, how decision-makers coordinate appropriate outsourcing operations for equipment maintenance while lowering cost is an important topic of discussion. This study attempts to investigate the assignment and dispatch of semiconductor equipment testing personnel. Another objective is to determine the minimum fiscal funds that should be paid to outside contractors, subject to certain limitations and conditions, for contract work. This may provide a reference for decision-makers making their outsourcing decisions. Furthermore, this paper seeks to conduct a trend analysis of sensitivity variables through a host of available parameters, so as to discover any unexpected result at the earliest possible time.
To verify the applicability and rationality of this model, this study uses a semiconductor plant in Hsinchu Science Park as a case study, and set parameters based on its operating conditions and basic information. By building a mathematical optimization model through integer programming, and utilizing Microsoft Excel 2010 to integrate data into LINGO software, computational results show optimized overall cost for outsourcing the testing of automatic fire extinguishing equipment to outside contractors. Furthermore, results from the optimized model can be contrasted and analyzed to the current model used by the plant, which is based on prior practical experience. Results show that the optimized model offers lower overall cost and higher efficiency as compared to the previous experience-based model. Therefore, we can conclude that the research model applied in this study can not only act as an important basis for reference, but can also be flexibly utilized to address the issue of outsourcing testing personnel for automatic fire extinguishing systems of semiconductor equipment. |
